WO2025217542A1 - System, method, and computer accessible medium for development of a fast-dose biodosimeter for rapid assessment of radiation exposure in human blood - Google Patents

Abstract

Exemplary systems, methods, and computer-accessible medium can be provided for detecting radiation exposure. Thus, the exemplary systems, methods, and computer-accessible medium are provided that input at least one intracellular or blood plasma protein information into a machine learning model and generate, by the machine learning model, exposure data for the radiation exposure based on the input at least one protein information. The exposure data can be or include a prediction as to exposure and/or amount of exposure that can be used to guide medical treatment. Exemplary systems, methods, and computer-accessible medium can further direct treatments based on these predictions.

Description

SYSTEM , METHOD, AND COMPUTER ACCESSIBLE MEDIUM FOR DEVELOPMENT OF A FAST-DOSE BIODOSIMETER FOR RAPID ASSESSMENT OF RADIATION EXPOSURE IN HUMAN BLOOD

STATEMENT REGARDING FEDERALLY FUNDED RESEARCH

[0001] This invention was made with government support under U01 AI148309 and U19 AI067773 awarded by the National Institutes of Health. The government has certain rights in the invention.

CROSS REFERENCE TO RELATED APPLICATION(S)

[0002] This application relates to U.S. Provisional Patent Application Serial Nos. 63/633,434, filed on April 12, 2024, and 63/662,549, filed on June 21, 2024, the entire disclosures of which are incorporated herein by reference.

FIELD OF THE DISCLOSURE

[0003] The present disclosure relates generally to systems, methods and computer-accessible medium for facilitating a biodosimetry for an accelerated and accurate assessment of a radiation absorbed dose following a radiati on/nucl ear incident, and to an exemplary bioassay for radiation triage and dose categorization to reconstruct dose based on BAX, DDB2, FDXR, ACTN1, TSPYL2, (and p53) lymphocyte (intracellular) protein expression levels, blood plasma biomarkers (CD20, CD5, MCP-1, FLT3-L), cell counts/viability and/or biomarker ratios.

BACKGROUND INFORMATION

[0003] Radiation exposure is a major threat to public health whether in the form of an improvised nuclear device (IND), nuclear reactor accidents caused by natural calamities such as what happened in Fukushima or the loss of radioactive sources. Following a mass-casualty radiological/nuclear incident, there would likely be a critical need to rapidly evaluate potentially exposed individuals for clinical triage and medical interventions. Modelling studies performed by Lawrence Livermore National Laboratory for a 10 KT (kiloton) nuclear detonation (see, e.g., Ref. 1) and prompt radiation exposures within an urban environment suggest that up to one million individuals could be subjected to triage based on the prediction that there will be significant infrastructure damage from ground zero to approximately a 2 miles radius (see, e.g., Refs. 2-5). About 0.75 to 1 Gy is the threshold level of exposure that induces mild radiation illness (See, e.g., US EPA 2013). Although this level of exposure is not anticipated to pose an immediate danger to life, individuals receiving this dose of radiation may still need medical management and treatment for symptoms or a follow-up evaluation. Individuals exposed to more than 2 Gy are at risk of suffering large damage to hematological, gastrointestinal, cutaneous, cardiovascular, and central nervous systems and will likely develop the acute radiation syndrome (ARS). These individuals can benefit considerably from timely medical attention. (See, e.g., Refs. 6 and 7).

[0004] Preparedness and response for such a catastrophic event can require emergency responders to be able to discriminate between different levels of radiation exposure quickly and accurately. Currently, no radiation biodosimetry methods have been cleared or approved by the U.S. Food and Drug Administration (FDA). (See, e.g., Ref. 8). The ability to rapidly triage patients and characterize exposure level can be essential to enable effective prioritization of scarce medical resources. Unfortunately, the lack of effective high throughput laboratory -based assays for biodosimetry triage tools are a recognized deficiency in national preparedness. Computer-based software diagnostic tools are available to health-care providers, such as the EAST tool (see, e.g., Ref. 9), BAT (see, e.g., Ref. 10) and HemoDose (see, e.g., Ref. 11), which largely rely on data from complete blood counts, lymphocyte depletion kinetics and clinical symptoms (e.g., nausea and vomiting) from the suspected radiation exposed subject to assist in identifying individuals for radiological triage.

[0005] Presently, several biodosimetry tools are in development that use cytogenetics (see, e.g., Ref. 12), proteomics (see, e.g., Refs. 13 and 14), and genomics (see, e.g., Ref. 15) endpoints. The CytoRADx system developed by ASELL™ employs a high throughput process to perform a standardized micronucleus assay that has been validated in human and NHP blood samples in ex-vivo and in-vivo models up to 8 Gy dose. (See, e.g., Ref. 12). SRI International is apparently developing a lateral flow immunoassay that has been used to quantify cytokine markers SAA, FLT3-L and MCP-1 in blood plasma of NHPs up to 10 Gy and up to 2 weeks post irradiation. (See, e.g., Ref. 16). Recent advancements in genomic mRNA markers have led to the development of a PCR-based high throughput ARad biodosimetry test by Arizona State University with Midwest Research Institute global (MRI) to estimate absorbed dose between 0 and 10 Gy from blood sampled 1 to 7 days post exposure. (See, e.g., Ref. 17). Additionally, the REDI-Dx Biodosimetry test system developed by DxTerity, can measure RNA expression in blood using the DxDirect genomic platform to classify absorbed dose at 2 thresholds above 2 Gy and above 6 Gy. (See, e.g., Ref. 15).

[0006] While these systems in development are beneficial, there remains a need for technological innovations that improve the diagnostic throughput, accuracy, window, and speed for radiation exposure in mass exposure events. Accordingly, it can be beneficial to provide a FAST-DOSE (Fluorescent Automated Screening Tool for Dosimetry) bioassay, designed to rapidly quantify radio-responsive intracellular proteins in blood leukocytes by imaging flow cytometry (IFC) for retrospective dose reconstruction up to at least a week after exposure to ionizing radiation as previously provided (see, e.g., Refs. 13 and 18-22), which can overcome at least some of the deficiencies described herein above.

SUMMARY OF EXEMPLARY EMBODIMENTS

[0007] The following is intended to be a brief summary of the exemplary embodiments of the present disclosure, and is not intended to limit the scope of the exemplary embodiments described herein.

[0008] In some exemplary embodiments of the present disclosure, the exemplary systems, methods, and computer accessible medium can be provided for detecting radiation exposure which can include inputting at least one blood lymphocyte or plasma protein information into a machine learning model and generating, by the machine learning model, individual radiation exposure information based on the input expression level of at least one protein information from a single blood sample. The exposure information can comprise information for triage to identify individuals who have been exposed or unexposed to ionizing irradiation and semi-quantitative measurements for radiation absorbed dose (categories 0 to < 2 Gy, 2-4 Gy, 4-6 Gy and > 6 Gy) that can be used to score the severity of radiation exposure, guide treatment planning and monitoring patient outcomes. The exposure prediction for triage can be based on a single blood lymphocyte or plasma protein information whereas the dose prediction (categorization) can be based on a combination of at least two protein information. The accuracy of the generated radiation exposure information can be directly correlated with a number of lymphocyte and plasma protein information input into the machine learning model. Additionally, the accuracy of the generated radiation exposure information can be directly correlated with a type of protein biomarker information input into the machine learning model. The exemplary systems, methods, and computer accessible medium can generate the radiation dose prediction up to 7-14 days after exposure to ionizing radiation. Also, the exemplary systems, methods, and computer accessible medium can generate by the machine learning model, a specific medical treatment guidance based on the generated radiation exposure information.

[0009] These and other objects, features and advantages of the exemplary embodiments of the present disclosure will become apparent upon reading the following detailed description of the exemplary embodiments of the present disclosure, when taken in conjunction with the appended claims.

BRIEF DESCRIPTION OF THE DRAWINGS

[0010] Further objects, features and advantages of the present disclosure will become apparent from the following detailed description taken in conjunction with the accompanying Figures showing illustrative embodiments of the present disclosure, in which:

[0011] Figure 1A is an exemplary graph of a percentage of lymphocyte viability versus dose for an average human according to an exemplary embodiment of the present disclosure;

[0012] Figure IB is an exemplary graph of a percentage of lymphocyte viability versus dose for NHP according to an exemplary embodiment of the present disclosure;

[0013] Figure 2A is an exemplary graph of a BAX concentration (after 1 or 2 days of cell culture) versus dose according to an exemplary embodiment of the present disclosure;

[0014] Figure 2B is an exemplary plot of the BAX concentration presented by sex versus dose according to an exemplary embodiment of the present disclosure;

[0015] Figure 3A is an exemplary graph of a DDB2 concentration (after 1 or 2 days of cell culture) versus dose according to an exemplary embodiment of the present disclosure;

[0016] Figure 3B is an exemplary plot of the DDB2 concentration presented by sex versus dose according to an exemplary embodiment of the present disclosure;

[0017] Figure 4A is an exemplary graph of dose response curves for BAX in NHP peripheral blood samples exposed to X rays ex vivo according to an exemplary embodiment of the present disclosure; [0018] Figure 4B is an exemplary graph of dose response curves for DDB2 in NHP peripheral blood samples exposed to X rays ex vivo according to an exemplary embodiment of the present disclosure;

[0019] Figure 5A is an exemplary illustration of the performance of the stacking ensemble for classifying samples as irradiated or not on human data according to an exemplary embodiment of the present disclosure;

[0020] Figure 5B is an exemplary illustration of the performance of the stacking ensemble for classifying samples as irradiated or not on NHP data according to an exemplary embodiment of the present disclosure;

[0021] Figure 6A is a set of exemplary plots illustrating performance of the stacking ensemble for reconstructing dose quantitatively for human data according to an exemplary embodiment of the present disclosure;

[0022] Figure 6B is a set of exemplary plots illustrating performance of the stacking ensemble for reconstructing dose quantitatively for NHP data according to an exemplary embodiment of the present disclosure;

[0023] Figure 7A is an exemplary plot of Concentration of BAX measured in NHPs exposed in-vivo to whole-body irradiation of 2.5 Gy according to an exemplary embodiment of the present disclosure;

[0024] Figure 7B is an exemplary plot of Concentration of DDB2 measured in NHPs exposed in-vivo to whole-body irradiation of 2.5 Gy according to an exemplary embodiment of the present disclosure;

[0025] Figure 8A is an exemplary plot of a reference standard curve for BAX according to an exemplary embodiment of the present disclosure;

[0026] Figure 8B is an exemplary plot of a reference standard curve for DDB2 according to an exemplary embodiment of the present disclosure;

[0027] Figure 9 is a block diagram of an exemplary embodiment of a system according to the present disclosure; and

[0028] Figure 10 is an exemplary table of intracellular biomarkers BAX and DDB2, lymphocyte cell counts and viability measurements for both human and NHP samples according to an exemplary embodiment of the present disclosure. [0029] Figure. 1 1 is a diagram showing an exemplary workflow of a FAST-DOSE bioassay which consists of four modules used to rapidly generate dose predictions from a single peripheral blood sample within 3.5 to 4 hours according to an exemplary embodiment of the present disclosure.

[0030] Figure 12 is a diagram of an exemplary workflow of a machine learning (ML) platform for training models for radiation dose prediction - the RDP/Module 4 of the FASTDOSE bioassay according to an exemplary embodiment of the present disclosure.

[0031] Figure 13 is an exemplary plot of longitudinal measurements in NHPs exposed to acute dose total body irradiation (e.g., 0 to 10 Gy) where dose/time kinetics of specific blood lymphocyte and plasma biomarkers and differential blood counts are measured at specific time points up to a week after radiation exposure according to an exemplary embodiment of the present disclosure.

[0032] Figure 14 is an exemplary table for fold change measurements of biomarker expression measured at each time/dose datapoint according to an exemplary embodiment of the present disclosure.

[0033] Figure 15 is an exemplary table for biomarker data used for input into a machine learning platform of a FAST-DOSE radiation dose predict model according to an exemplary embodiment of the present disclosure.

[0034] Figure 16 is an exemplary table for diagnostic performance of a FAST-DOSE biodosimetry tool for radiological triage according to an exemplary embodiment of the present disclosure.

[0035] Figure 17 is an exemplary table for diagnostic performance of a FAST-DOSE biodosimetry tool for dose categorization (e.g., 3- and 4-classes) using a combination of 1 or 2 biomarkers up to a week after radiation exposure according to an exemplary embodiment of the present disclosure.

[0036] Figure 18 is an exemplary table for diagnostic performance of a FAST-DOSE biodosimetry tool for dose categorization (e.g., 3- and 4-classes) using a combination of multiple biomarkers up to a week after radiation exposure according to an exemplary embodiment of the present disclosure.

[0037] Figure 19 is an exemplary table for diagnostic performance of a FAST-DOSE biodosimetry tool for dose categorization (e.g., 3- and 4-classes) using a combination of at least two biomarkers up to 14 days after radiation exposure according to an exemplary embodiment of the present disclosure.

[0038] Figure 20 is an exemplary plot for biomarker response in human cancer patients (n = 5 patients) after 3 fractions of total body irradiation according to an exemplary embodiment of the present disclosure.

[0039] Throughout the drawings, the same reference numerals and characters, unless otherwise stated, are used to denote like features, elements, components or portions of the illustrated embodiments. Moreover, while the present disclosure will now be described in detail with reference to the figures, it is done so in connection with the illustrative embodiments and is not limited by the particular embodiments illustrated in the figures and the appended claims.

DETAILED DESCRIPTION OF EXAMPLARY EMBODIMENTS

[0040] The following description of exemplary embodiments provides non-limiting representative examples referencing numerals to particularly describe features and teachings of different exemplary aspects and exemplary embodiments of the present disclosure. The exemplary embodiments described should be recognized as capable of implementation separately, or in combination, with other exemplary embodiments from the description of the exemplary embodiments. A person of ordinary skill in the art reviewing the description of the exemplary embodiments should be able to learn and understand the different described aspects of the present disclosure. The description of the exemplary embodiments should facilitate understanding of the exemplary embodiments of the present disclosure to such an extent that other implementations, not specifically covered but within the knowledge of a person of skill in the art having read the description of embodiments, would be understood to be consistent with an application of the exemplary embodiments of the present disclosure.

[0041] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure relate to and/or utilize a FAST-DOSE (Fluorescent Automated Screening Tool for Dosimetry) bioassay system, which can be configured to rapidly quantify radio-responsive intracellular proteins in blood leukocytes by imaging flow cytometry (IFC) or for retrospective dose reconstruction up to at least a week after exposure to ionizing radiation - as previously described. (See, e.g., Refs. 13, 18-22). [0042] The bioassay of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be designed to estimate radiation exposure level in individuals suspected of being exposed to ionizing radiation in a high- risk environment (e.g., detonation of radiological dispersal device, improvised nuclear device explosion; power plant accident or military combat) and support medical triage and resource allocation in emergency settings by identifying individuals who require urgent medical treatments (e.g., cytokine therapy, hospitalization) versus those who can be monitored remotely. Treatment decisions may not just be tied to exposure level but can also correlate with clinical signs and symptoms (e.g., emesis, complete blood counts). The FAST-DOSE biodosimetry tool of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can provide a high throughput device with a time-to-result in < 4 hours. The blood test can be performed at a centralized laboratory and can be used across the general population.

[0043] There are currently no FDA-cleared biodosimetry tools for rapid and accurate assessment of radiation absorbed dose following a radiation/nuclear incident. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can provide an ELISA-based bioassay system for radiation biodosimetry. The prediction accuracy of the bioassay according to exemplary embodiments of the present disclosure for exposure classification and dose reconstruction can be determined by combining BAX and DDB2 (and FDXR, ACTN1, TSPYL2, (and p53) intracellular) protein expression levels protein expression levels and cell counts/viability in adult human and non-human primate (NHP; Rhesus macaques) leukocytes, irradiated ex vivo with 0 to 5 Gy X rays using machine learning methods. The bioassay according to exemplary embodiments can show a 97.92% and 96.15% accuracy in classifying the human and NHP in-vitro samples up to 48 h after exposure, respectively and an adequate correlation between reconstructed and actual dose in the human samples (R² = 0.79, RMSE = 0.80 Gy, and MAE = 0.63 Gy) and NHP (R² = 0.80, RMSE = 0.78 Gy, and MAE = 0.61 Gy). Biomarker measurements in vivo from four NHPs exposed to a single 2.5 Gy total body dose can indicate a persistent upregulation in blood samples collected on days 2 and 5 after irradiation. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be used to indicate that using a combined approach of targeted proteins can increase bioassay sensitivity and provide a more accurate dose prediction.

[0044] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can identify a panel of top-candidate intracellular protein biomarkers (DDB2, BAX, FDXR, TSPYL2 and ACTN1) using shotgun proteomics to assess proteome-wide changes in human CD45+ blood leukocytes in X-irradiated humanized mice (See, e.g., Ref. 18). For example, currently, the exemplary embodiments of the present disclosure have evaluated the performance of the FAST-DOSE bioassay across different models and species that include the human blood ex-vivo model (see, e.g., Ref. 19), non-human primates (NHP) (see, e.g., Ref. 13), humanized and C57BL/6 mice (see, e.g., Refs. 13, 20 and 21, 22).

[0045] One of the objectives of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be to transition and integrate two of the top-performing biomarkers BAX (BCL2 associated X, a regulator of apoptosis - see, e.g., Refs. 23, 24) and DDB2 (DNA damage specific binding protein, a protein which binds to DNA as part of the cellular response to DNA damage - see, e.g., Ref. 25) into an ELISA-based platform with the goal to simplify the assay to increase speed and reduce the time- to-result.

[0046] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can use, e.g., a human and NHP blood ex-vivo model and expose four NHPs in vivo to validate the two protein biomarkers by exposing blood samples to acute radiation and culturing for up to 48 hours to measure the biomarker dose response. By applying machine learning models, The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can predict (e.g., reconstruct) radiation exposure level and the absorbed dose based on the dosedependent response of the lymphocyte and plasma protein biomarkers, leukocyte cell counts and viability, a recognized indicator of radiation injury. (See, e.g., Refs. 26-28).

[0047] Table 1 summarizes an exemplary performance of high throughput FAST-DOSE bioassay which have been tested across the two different platforms (IFC and ELISA). The performance metrics for each are described (AUROC = Area Under the ROC curve; RMSE: Root Mean Squared Error; MAE: Mean Absolute Error). Sample _ Exposure

Platform Model Collection . Classification Dose Reconstruction Metrices Reference

(Days) ^,t,y) (AUROC)

Table 1. Summary of exemplary performance of FAST-DOSE bioassay

[0048] To extend the FAST-DOSE protein panel, the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can integrate four established radio-responsive blood plasma proteins, MCP-1 (monocyte chemotactic protein 1), FLT3-L (Fms-related tyrosine kinase 3 ligand) and free surface CD20 (B-cell) and CD5 (T-cell) antigen which are present in the blood plasma blood leukocyte, as part of a multi-parameter approach to improve dose reconstruction. Previous work indicates that biomarkers can show a dose dependent upregulation (MCP-1 and FLT3-L) in blood samples collected from human cancer patients, NHPs, baboons and mice up to a week after radiation exposure (See, e.g., Refs. 29-32). CD20 and CD5 reflect the dose dependent depletion of circulating free surface B and T cell antigen which are present in the blood plasma (see, e.g., Ref. 33) observed after radiation exposure.

Exemplary Design of the FAST-DOSE Biodosimetry Tool

[0049] Figure 11 illustrates a workflow of exemplary FAST-DOSE bioassay which can include a number (e.g., four) modules: 1) The Protein Extraction (PrEx) module (90 min) 1110 can extract plasma and leukocyte proteins from fractionated blood that can serve as the sample input for the biodosimetry tool. 2) The protein sample can then be input into the Protein Detection (Prd) module (90 min) 1120 which can use a sandwich ELISA-based immunoassay detection panel to bind and immobilize target protein biomarkers. 3) The Biomarker Quantification (BioQ) module (15 min) 1130 can measure biomarker absorbance using e.g., validated BioTek Gen5 software (including 21 CFR Part 11 Compliance), and can generate a biomarker signature (e.g., a comprehensive profile of the quantified proteins). 4) The Radiation Dose Prediction (RDP) module (< 1 min per sample) 1140 can use a custom -trained ML model to generate predictions of absorbed dose based on the biomarker signature. Overall, this biodosimetry tool of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can generate dose predictions from a single peripheral blood sample within e.g., 3.5 to 4 hours. ELISA is a highly sensitive tool for in vitro diagnostics, thus permitting its use in clinical and CLIA certified laboratories. (See, e.g., Refs. 34, 35). As ELISA protocols can utilize 96 and 384 plate formats and be automated, the anticipated sample throughput of exemplary biodosimetry bioassay could be thousands per day, which can be critical following a mass-casualty R/N emergency.

[0050] It is possible using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure to provide a protein biomarker signature that can be used for absorbed dose estimates in peripheral blood samples across the general population.

Exemplary Multiparametric modeling of FAST-DOSE biomarker signals for dose prediction using the radiation dose predict (RDP) module:

[0051] Figure 12 shows the machine learning (ML) platform to rank and select the most important features which can predict absorbed dose and combine them to produce a robust ensemble model for accurate dose prediction. At a high level, Figure 12 shows ingestion of all features at procedure 1210, then at procedure 1220, features can be selected from all features. At procedure 1230, retained features can be passed on to a model stacking module 1240. Testing dataset 1250 can be used to arrive at predicted dose 1260 and performance metrics 1270.

[0052] The robust testing of biomarker signals across various experimental designs can generate dose response curve data associated with many predictor variables, also called features (i.e. individual biomarkers [leukocyte/plasma/cell counts/ratios], as well as for tested biological variables [e.g. sex, age, immune status, chronic disease] and time). An exemplary custom- designed machine learning (ML) platform can be provided for the seamless testing and integration of biomarker dose response features, to produce multiparametric biodosimetry outputs (see, e.g., Refs, in Table 1). Dose response curve data can be split into training and testing datasets, 1202 and 1250 respectively. It is possible to use the training set 1202 on the ML platform (see, e.g., Fig. 12) to rank and select the most important features which predict absorbed dose and combine them to produce a robust ensemble model for accurate dose prediction. [0053] Feature Selection Module 1220'. To improve model effectiveness, non-informative or redundant features from training dataset can be removed before training the final dose prediction model. The “Boruta Shap” wrapper algorithm can be used for feature selection at, which can use SHapley Additive exPlanations -SHAP (36) values to evaluate the strength of the features that contribute to the model and prediction outputs. SHAP values consider all features at once, account for joint effects, and handle interactions. Using a set of real (original) features 1224 and shadow features 1225 (duplicated features that have been shuffled into noise), a base model 1226 (e.g., random forest) can generate mean absolute SHAP values for each feature 1222, and the features can be ranked and converted to Z-scores by Boruta to compare each feature against the top shadow Z-score (p value cutoff = 0.05) 1223. This combined exemplary approach can evaluate:

1) whether each feature significantly outperforms random noise as a predictor of dose,

2) how much each feature contributes to the model’s dose prediction outputs,

3) whether interacting features produce a combined effect/enhanced contribution on dose prediction, and

4) whether a set of correlated features can be considered redundant, and one feature can be discarded.

[0054] This exemplary approach based on BorutaShap can reduce or remove the complexity and potential overfitting of the model, while also identifying which features may contribute the greatest impact of the predictions generated by the model, thus providing information about which biomarkers are useful to keep in the panel design.

[0055] Model Stacking Module 1240'. The ML platform shown in Figure 12 can utilize important exemplary retained features 1230 identified by the BorutaShap procedure to perform model stacking to comprehensively combine the predictions (and harness the capabilities) from multiple well-performing base models (e.g. parametric regression, random forest (see, e.g., Ref. 37), XGBoost (see, e.g., Ref. 38), LGBM (see, e.g., Ref. 39), CatBoost (see, e.g., Ref. 40), and SVM (see, e.g., Ref. 41). These exemplary predictions can be used to train a meta-model 1242 to improve prediction accuracy, reduce variance, and increase robustness to noise and outliers. To reduce overfitting, the base model-fitting procedure 1244 can be performed multiple times with repeated cross validation to generate many exemplary predictions 1246. Each time, part of the dataset can randomly be designated as training and the rest as testing. Performance metrics relevant to each dose prediction can be calculated (root mean squared error (RMSE), mean absolute error (MAE) and coefficient of determination (R²) for regression tasks, and total and balanced accuracy for classification) can be calculated or otherwise determined. A distribution of performance metric values across runs can provide an estimate of the ensemble model’s mean performance and variability.

[0056] Dose Prediction Module 1260 and Performance Evaluation Module 1270 Using the exemplary stacked ensemble model, absorbed dose can be predicted in each dataset 1250 in the following exemplary ways: 1) Binary: models can be trained to differentiate irradiated versus unirradiated individuals, and 2) Categorical Dose: to improve potential applications of the modeling to triage decisions, it is possible to also convert the dose data to categorical classes that correspond with treatment decisions using biologically defined cutoff values for acute radiation syndrome. Regression models can be trained, e.g., to predict Dose as a categorical variable, binned into clinically relevant categories of exposure levels: low (e.g., < 2 Gy), medium (e.g., 2- 4 Gy) and high (e.g., 4-6 Gy and > 6 Gy). Exemplary performance metrics 1270 of total accuracy (or balanced accuracy, which mitigates the impact of potential imbalance of data across the categories), can be measured. To assess the consistency of dose predictions across the entire dose spectrum, it is possible to examine model performance within different dose categories. If the ML performance is suboptimal for a specific dose category, it is possible to apply a weighted loss function during the classification step to prioritize the underperforming category during training.

Exemplary cell viability and correlation with dose

[0057] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, lymphocyte viability based on PI/ AO staining measured on days 1 and 2 post-irradiation of both human and NHP blood samples in- vitro can be indicated to be, e.g., the percentage of surviving lymphocytes decreased as the dose of radiation exposure increased, (see, e.g., Ref. 26). For example, Figure 1A (human) shows an exemplary graph showing the percentage of lymphocyte viability versus dose for the average human, plotted for each day, and measured by AO/PI staining on days 1 and 2 post-irradiation. This figure illustrates that the average viability on day 1 was 97.7% ± 0.4% in the control group compared to 86.4% ± 2% in samples exposed to 5 Gy X rays (n = 10), whereas by day 2, 92.9% ± 2.7% of the control group lymphocytes and 74.5% ± 4.2% in cells exposed to 5 Gy X rays (n = 10) remained viable. Figure IB shows an exemplary graph showing the percentage of lymphocyte viability versus dose for NHP, plotted for each day, and measured by AO/PI staining on days 1 and 2 post-irradiation. This figure illustrates that overall lymphocyte cell viability can be lower compared to the human samples which can be due to the fact that the NHP blood was shipped fresh overnight from Wake Forest. On days 1 and 2, control samples showed an average viability of 86.2% ± 1.9% and 83.1% ± 2% and in the 5 Gy-irradiated blood samples viability was 52.9% ± 2.6% (n = 13) and 49.4% ± 5% (n = 12), respectively.

[0058] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can study viability as a function of radiation dose at each timepoint using Pearson’s correlation coefficient: The exemplary human samples (see, e.g., Figure 1A), show a correlation between dose and lymphocyte viability on days 1 (r = 1, p = 0.003) and 2 (r = 0.955, p = 0.003); however, the relationship in this example is non-linear at either timepoint (R² > 0.24). Further there is no discernible difference in lymphocyte viability based on sex on either day (p > 0.35). In the exemplary NHP samples (see, e.g., Figure IB), a correlation can be observed between dose and lymphocyte viability on day 1 (r = 1, p = 0.006) and day 2 (r = 0.935, p = 0.006); the curves are non-linear (R² > 0.373). Overall, this exemplary data according to exemplary embodiments of the present disclosure indicates that a high percentage of the unirradiated lymphocytes survived the isolation and 2-day culture, whereas cells exposed to x-ray doses exhibited reduced cell viability and apparent radiation-induced cell death.

Exemplary quantification of BAX and DDB2 in X-irradiated human PBMCs

[0059] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be used to determine and/or confirm that BAX and DDB2 levels can be negatively correlated with lymphocyte viability on both days: BAX on day 1 (r = -0.8224, p =0.0445) and day 2 (r = -0.9643, p = 0.0019) and DDB2 on day 1 (r = -0.9535, p = 0.0032) and day 2 (r = -0.9897, p = 0.0002). Figure 2A shows an exemplary graph providing an exemplary dose response of BAX concentration in human lymphocyte cell lysates at 24 h (n = 11) and 48 h (n = 10) after exposure, where all data points are plotted with lines connecting mean concentration for each dose on each day. According to the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, there can be a significant positive correlation between dose and BAX response on day 1 (r = 1.000, p = 0.002) and day 2 (r = 0.961, p = 0.002) which could not be fitted to a linear regression curve (R² > 0.33). A significant difference can be found between doses 0 and 1 Gy on day 1 (p = 0.0002) and day 2 (p = 0.0003). Differences between the two time points can indicate that BAX yields are significantly (p < 0.025) higher in the day 1 samples exposed to 1-3 Gy. Figure 2B shows an exemplary plot of the BAX concentration presented by sex versus dose according to an exemplary embodiment of the present disclosure which indicates that there is no significant difference in BAX concentrations between male and female donors on either day (p > 0.35). The box and whisker plots shown in Figure 2B illustrate, e g., minimum, median, quartiles and maximum BAX concentrations for each sex on each day at each dose. [0060] Figure 3A shows an exemplary graph of a DDB2 concentration (after 1 or 2 days of cell culture) versus dose providing an exemplary dose response for DDB2 at 24 h (n = 8) and 48 h (n = 8) after exposure, where all data points are plotted with lines connecting mean concentration for each dose on each day. As illustrated in Figure 3 A, it appears that there is a significant positive correlation between dose and DDB2 response on day 1 (r = 0.9651, p = 0.0018) and day 2 (r = 0.9742, p = 0.0010) which can be fitted reasonably well to a simple linear regression model on days 1 (R² = 0.8303) and 2 (R² = 0.7859). Figure 3B shows an exemplary plot of the DDB2 concentration presented by sex versus dose indicating the DDB2 levels in the male and female donors, and where box and whisker plots show minimum, median, and maximum DDB2 concentration values. Unlike BAX, there appears to be no significant increase (p > 0.55) in DDB2 expression control vs 1 Gy, although there is a significant dose dependent increase on both days, but DDB2 yields are significantly higher (p > 0.013) higher in day 1 samples exposed to 2 and 5 Gy. Between the two days, DDB2 levels are significantly higher (p < 0.013) on day 1 after exposure to 2 and 5 Gy X rays. Figure 3B shows that the mean DDB2 levels are significantly higher (p = 0.0304) in the male samples on day 1 but not on day 2 when compared to female samples.

Exemplary quantification of BAX and DDB2 in X-irradiated NHP PBMCs [0061] According to exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, on both days, BAX and DDB2 levels show a significant negative correlation with lymphocyte viability in the NHP cell cultures using Pearson correlation coefficients: for BAX on day 1 (r = -0.5587, p = 0.2491) and day 2 (r = -0.5536, p = 0.2545); for DDB2 on day 1 (r = -0.939, p = 0.0055) and on day 2 (r = - 0.9850, p = 0.0003). Figures 4A and 4B show exemplary graphs of dose response curves for BAX/DDB2 and in NHP peripheral blood samples exposed to X rays ex vivo indicating. For example, exemplary dose responses for BAX and DDB2 in NHP blood samples irradiated ex vivo up to 5 Gy according to exemplary embodiments of the present disclosure. All NHP samples (n = 13 on both days) were collected from male NHPs, so sex-related differences were not determined.

[0062] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can indicate that there is no significant correlation between BAX concentration and dose on either day 1 (r = 0.6791, p = 0.1380) or day 2 (r = 0.4716, p = 0.3451). Similar to the human samples, there is a significant difference in BAX levels between the control 0 Gy and 1 Gy samples on days 1 (p = 0.0137) and 2 (p = 0.0111) with no measurable dose dependent increase across the irradiated doses. DDB2 shows a significant positive correlation for concentration and dose on both days (day 1, r = 0.9905 and day 2, r = 0.9909; p = 0.0001 on both days). Using the exemplary systems, methods and computer-accessible medium according to exemplary embodiments of the present disclosure, there is a significant difference between the control samples and 1 Gy exposed samples on day 1 (p = 0.0002), whereas at day 2 there is no significant difference (p = 0.0569). BAX does not fit linearly with dose on both the time points (R² < 0.2), whereas the DDB2 response can be reasonably fitted well linearly with dose on day 1 (R² = 0.812) and for day 2 it does not fit well (R² = 0.619).

Exemplary Dose Reconstruction

[0063] According to the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, exemplary dose classification (exposed vs unexposed) and reconstruction (quantitative dose predictions) results for human and NHP data (on the testing data subset) are shown in Figures 5A and 5B and Figures 6A and 6B, respectively, which also indicate the exemplary performance metrics on each testing data set. The combination of the three markers of radiation exposure, including cell count/lymphocyte viability (see, e.g., Figure 10) for all the donors and animals, at both the time points can be measured and intracellular concentration of BAX and DDB2 (and FDXR, ACTN1, TSPYL2, (and p53) intracellular) protein expression levels) can be successfully used to classify samples as exposed or non-exposed (see, e.g., Figures 5 A and 5B). In humans, an exemplary assay and machine learning workflow according to The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be about 97.92% accurate in predicting the exposure status of a sample on testing data, with a true positive rate of 100% and a true negative rate of about 88.89%. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can utilize, e.g., 39 samples exposed to radiation of 1 Gy or more for testing in the binary dose classification model. Of those samples, all 39 can be correctly predicted by The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure to be exposed samples based on their viability, BAX, and DDB 2 data.

[0064] For example, from the 9 non-irradiated samples that can be used for testing in the exemplary model of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, 8 were correctly predicted to be non-exposed, and only 1 was mis-classified as radiation-exposed (See, e.g., Figure 5 A). In NHPs, the assay and analysis utilized by the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be slightly less successful in classifying radiation exposure and similar to humans, the model may struggle more with identifying non-irradiated samples compared to identifying samples that have been irradiated. The exemplary binary classification model of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be 96.15% accurate in NHPs, with a true positive rate of 100% and a true negative rate of 70%. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can classify, e.g., all 68 irradiated samples as such, while 7 of 10 non-irradiated samples used for testing were correctly classified as nonirradiated (See, e.g., Figure 5B). [0065] The exemplary approach according to The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can also achieve an appropriate performance for reconstructing dose in a quantitative manner. In humans, plotting dose reconstruction values generated by modeling against the true dose for each sample in the testing set (See, e.g., Figure 6 A) can produce R² = 0.7914, RMSE (Root Mean Square Error) = 0.8007 Gy, and MAE (Mean absolute difference) = 0.6304 Gy. In this exemplary case, BAX and DDB2 concentration, and lymphocyte cell counts can be used to build the dose reconstruction model of exemplary embodiments of the present disclosure. Lymphocyte viability did not pass Boruta testing and so was not included in the model. Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, for NHPs, the relationship between reconstructed dose and actual dose of the testing set can produce R² = 0.7980, RMSE = 0.7816 Gy, and MAE = 0.6099 Gy (Fig. 6B). Thus, BAX and DDB2 levels, lymphocyte cell counts, and cell viability passed Boruta testing and were used in the dose reconstruction model of exemplary embodiments of the present disclosure.

[0066] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, BAX and DDB2 protein biomarker levels (and FDXR, ACTN1, TSPYL2, (and p53) intracellular) protein expression levels) can also be measured in the blood samples collected from healthy NHPs (n = 4) exposed to a 2.5 Gy totalbody radiation in-vivo with blood samples collected on days 2, 5 and 14. Figure 7A shows an exemplary plot of Concentration of BAX measured in NHPs exposed in-vivo to whole-body irradiation of 2.5 Gy, whereas, using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, intracellular BAX levels significantly can increase on day 2 (p = 0.0206) and day 5 (p = 0.0096) when compared to that of pre-irradiated samples with also a significant increase between days 2 and 5 (p=0.0037). Figure 7B shows an exemplary plot of Concentration of DDB2 measured in NHPs exposed in- vivo to whole-body irradiation of 2.5 Gy, providing a similar response for DDB2 as compared to BAX, whereby protein expression significantly increases on day 2 (p = 0.0148) and day 5 (p = 0.0135) compared to the pre-irradiated samples, and between days 2 and 5 (p = 0.0196).

[0067] Longitudinal measurements in Rhesus Macaque NHPs exposed to TBI doses (0, 2, 4, 6, 8 and 10 Gy): Figure 13 shows the time/dose kinetics of protein biomarker measurements in blood samples collected from NHPs exposed to acute-dose total body irradiation (0 to 10 Gy). Specific blood lymphocyte and plasma biomarkers and differential blood counts (absolute) were measured at specific time points up to a week after radiation exposure. The data highlights that intracellular BAX expression levels can continue to increase and persist up to days 5-7, whereas FLT3-L and MCP-1 can show persistent elevated levels to day 14. Predictably, lymphocytes and white blood and neutrophils are sensitive to radiation and can show a rapid depletion after radiation exposure that can reach nadirs between 2-14 days. Figure 14 shows the fold change for each evaluated biomarker analyte in NHPs exposed to total body irradiation (doses 2 to 10 Gy) at specific time points (days 2, 5 and 7) after exposure. Measurements were normalized (relative) to blood samples collected from NHPs before (either day -1 or day -3) irradiation exposure.

Exemplary performance of FAST-DOSE bioassay for radiological triage and dose categorization

[0068] Radiological triage. Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, 14 datasets can include biomarker measurements from non-human primates (NHPs) exposed to ionizing radiation at doses of 0, 2, 4, 6, 8, or 10 Gy, and can be collected across timepoints ranging from baseline (day 0, prior to exposure) through day 14 post-exposure, as reflected in Figure 15.

[0069] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, for triage-focused analyses, the dataset can be fdtered to include only baseline unexposed animals (0 Gy, day 0) and animals exposed to > 2 Gy at days 2, 5, and 7 post-exposure, as illustrated in Figure 16. This filtering can ensure that the modeling addresses the central objective of distinguishing unexposed animals from those receiving triage-relevant doses within the first week of exposure. A binary classification target variable can be constructed to reflect the Triage Index, where 0 can represent unexposed animals (0 Gy at day 0), and 1 can represent animals exposed to doses >2 Gy at any of days 2, 5, or 7. This index can be used as the outcome variable for all modeling analyses. To robustly evaluate model performance, the dataset can be split into a 50:50 training and testing set using stratified sampling based on the Triage Index, in order to preserve the balance of exposed and unexposed cases in both sets. The larger testing set, compared to earlier versions of the analysis, can ensure more representation of unexposed animals in the held-out set for performance evaluation. This can be important given that unexposed animals are only available at a single timepoint (day 0), and earlier random splits can occasionally result in minimal unexposed representation in the data. [0070] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, to evaluate the diagnostic performance of these biomarkers, Random Forest (RF) models can be applied using e.g., the random Forest package in R, with 100 trees per model and otherwise default parameters. Two modeling strategies can be implemented. First, e.g., all 13 biomarkers can be included as input features in a single RF classifier to assess overall classification performance. Second, a combinatorial approach can be taken to identify minimal biomarker panels capable of robust triage classification. Specifically, e.g., all possible one-biomarker and two-biomarker combinations (n = 13 for single features; n = 78 for two-feature combinations) can be enumerated, and a separate RF model can be trained for each combination using the training data. Each model can be evaluated on the held-out test set, and standard classification metrics can be calculated, including overall accuracy, sensitivity (true positive rate), specificity (true negative rate), and positive predictive value (precision). The confusion matrix for each model can also be examined to understand misclassification patterns. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can compile and rank all results to determine which single biomarkers and two- biomarker combinations can yield the highest classification performance. In the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, perfect classification (accuracy = 1.0) can be achieved by several individual biomarkers, supporting the potential utility of single-biomarker diagnostics for triage in field settings. Two-biomarker panels can further reinforce this performance, with multiple pairs achieving perfect or near-perfect metrics across all evaluation criteria.

[0071] Dose Categorization'. Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, to model dose severity based on biomarker profiles, supervised classification can be performed using two categorical dose groupings. In the 3-class setup (i.e., the number of dose categories), samples were labeled as 0 to < 2 Gy, 2-6 Gy, or > 6 Gy. In the 4-class setup, the mid-range was further divided into 0 to < 2 Gy, 2-4 Gy, 4-6 Gy, and > 6 Gy. These dose categories were derived from the Dose Gy column and applied to data collected on days 2, 5, and 7 post-irradiation. The analysis used data from 14 blood biomarkers described in Figure 15. For each timepoint and classification scheme, we evaluated all possible combinations of 1, 2, 3, and 4 biomarkers as input features.

[0072] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, each model was trained using a random forest classifier with 100 trees. The data were split into 50% training and 50% testing sets, stratified by class. For each biomarker combination, a model was trained on the training set and evaluated on the testing set. Accuracy, precision, recall, and Fl score were calculated for each target class based on confusion matrices. The top-performing combinations were identified for each dose class at each timepoint, and results were saved and ranked according to F 1 score which considers both precision and recall (sensitivity). Figure 17 shows the performance of the FAST-DOSE biodosimetry tool for dose categorization using a 3- and 4- class scheme for a single or two biomarker combination up to a week after radiation exposure. Figures 18 and 19 identify the top-ranking biomarker combinations using 4 biomarker combinations up to 7 and 14 days, respectively. Of note, the full output of the ML model identified more than 2000 biomarker panels/table with decreasing Fl scores. Many biomarker combinations achieved very good and excellent performance metrics that can used in validation and verification studies in pre-clinical animal and human models for review and feedback by the FDA.

Exemplary Radiotherapy patients exposed to fractionated TBI

[0073] Figure 20 shows an exemplary response for BAX, FLT3-L and MCP-1 in cancer patients after receiving up to 3 three fractions of TBI (2.25 Gy). This exemplary data illustrates a more extreme scenario where the patient’s immune system is ablated (in preparation for bone marrow transplant), leaving very low levels of apparently less damaged leukocytes. This can be reflected in the BAX expression levels. In the same blood sample, plasma protein FLT3-L can show a dose dependent increase in expression, whereas MCP-1 can be persistently upregulated after the first fraction. These data can highlight that at high doses of irradiation, there is increased radiation-induced cell death of the blood leukocytes, thus, the plasma biomarkers (and blood counts) can potentially be used to determine absorbed dose.

Exemplary Discussion [0074] For example, a radiological/nuclear (R/N) incident or accident can result in exposure of potentially thousands of individuals to radiation, who may need medical intervention in a resource-constrained environment. This mass-casualty scenario can also result in a sense of radiophobia (see, e.g., Refs. 42 and 43) in millions of people wondering whether they are exposed or not. A similar mass anxiety around exposure was witnessed recently in the Covid- 19 pandemic. In preparation for an R/N emergency there is a critical need for the development of high throughput biodosimetry tests (see, e.g., Ref. 14) that can accurately provide information on biological absorbed dose or provide triage capability to distinguish between individuals exposed to doses above and below 2 Gy (see, e.g. Refs. 44 and 45) , thereby prioritizing victims who will benefit most from prompt medical attention and treatment.

[0075] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure provide can be used to provide specific treatment information following radiation exposure particularly to those individuals who are at medium to high risk of mortality. For example, accurate and prompt identification of individuals exposed to 0 to < 2 Gy can result in the system ordering no treatment whereas exposure to doses > 2 Gy (radiological triage) can result in the system ordering cytokine therapy using FDA- approved radiomitigators together with antibiotics to aid hematopoietic radiation syndrome. The accurate screening of individuals exposed to moderate to high doses (2-6 Gy) can benefit from a similar medical intervention, however differentiation between 2-4 Gy and 4-6 Gy can additionally result in the system ordering hospital/clinic outpatient or inpatient medical invention as well as provide guidance for treatment of gastrointestinal injuries. Exposure to high risk, very high doses (> 6 Gy) of irradiation, can result in the system ordering hospitalization and intensive care where a hematopoietic stem cell transplant could be needed.

[0076] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be used to rapidly detect 13 distinct blood protein biomarkers (lymphocyte, plasma and differential blood cell counts) as described in Figures 14 and 15 which can be used as input variables into the FAST-DOSE radiation dose prediction module 1140 (see Figure 11) and machine learning platform (see Figure 12) for radiological triage and dose categorization used to predict the severity of radiation exposure. [0077] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure provide an ELISA-based biodosimetry tool that can be used to rapidly detect and quantify intracellular protein biomarkers (BAX and DDB2 - see, e.g. Ref. 26) in human and NHP peripheral blood samples towards the accurate prediction of radiation exposure and biologically absorbed dose.

[0078] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can utilize human and NHP in vitro culture exemplary models to evaluate BAX and DDB2 concentration levels at 24 h and 48 h after exposure of 0-5 Gy X rays. Both biomarkers can show a strong correlation with cell viability (coefficiency > -0.82) and dose (coefficiency > 0.96). Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, BAX can exhibit significant sensitivity in blood samples exposed to 1 Gy, that remained persistently elevated up to 5 Gy, observed as a flattening of the curve with increasing dose, whereas DDB2 levels can show a dose-dependent response that can be fitted to a linear regression curve (R² > 0.79) on both days (See, e.g., Figs. 2 and 3). Prior work used the wildtype C57BL/6 mouse model and machine learning methods to determine that a combination of intracellular biomarkers DDB2, FDXR, ACTN1, peripheral blood B and T cell counts and percentages can successfully be used to reconstruct dose and distinguish between PBI and TBI exposures. (See, e.g., Ref. 22). The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can apply a similar machine learning approach and use a combination of intracellular biomarkers BAX and DDB2 (and FDXR, ACTN1, TSPYL2, (and p53) intracellular) protein expression levels), lymphocyte cell counts and viability measurements (see, e.g., Figure 10) for prediction of radiation exposure classification and dose reconstruction.

[0079] The exemplary prediction accuracy of the exemplary ELISA-based proteomic biomarker assay of using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure to discriminate the unirradiated from the irradiated samples post exposure can be determined using raw biomarker values and AUC ROC performance on each testing dataset. (See, e.g.. Figures 5 A and 5B). The median AUC and their confidence intervals (CI) for the human and NHP ex-vivo samples are 0.9914 (95% CI: 0.970 - 1.0) and 0.9941 (95% CI: 0.982 - 1.0), respectively. The exemplary binary classification model according to the exemplary embodiments of the present disclosure can be about 97.92% accurate in predicting the exposure status in humans, and about 96.15% accurate in NHPs.

[0080] For the exemplary dose reconstruction performed by the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, a combination of BAX and DDB2 (and/or FDXR, ACTN1, TSPYL2, (and p53) intracellular) protein expression levels) and lymphocyte cell counts/viability can be used to quantitatively predict radiation dose in both the human and NHP samples. (See, e.g., Figures 6A and 6B). The exemplary performance metrics can show an adequate correlation between predicted and actual dose in the human samples (R² = 0.79, RMSE = 0.80 Gy, and MAE = 0.63 Gy) and NHP (R² = 0.80, RMSE = 0.78 Gy, and MAE = 0.61 Gy). Although the prediction of dose may be slightly improved by the inclusion of the cell count/viability data (See, e.g., Table 2), their inclusion in the bioassay platform can be important towards the assessment of hematological injury and the development of a biodosimetry tool for early population triage.

Table 3: Exemplary performance metrices of reconstructed dose using ML procedure

[0081] The exemplary Boruta feature selection component of the exemplary machine learning module according to exemplary embodiments of the present disclosure also rejected Day and Sex (NHPs were all male) variables as predictors for both exposure classification and dose reconstruction.

[0082] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, measurements of BAX and DDB2 in vivo in four NHPs exposed to a single total body irradiation dose of 2.5 Gy that shows a persistent upregulation in blood samples collected on days 2 and 5 after exposure. (See, e.g., Figures 7A and 7B). This can be comparable with prior work in the NHP in-vivo model (see, e.g., Ref. 13), where both biomarkers showed a persistently increased expression up to day 8 after 2-5 Gy of total body exposures. It is known that both BAX and DDB2 are involved in apoptosis and DNA repair, which are part of the DNA damage response (DDR) cellular response to ionizing radiation. Previous studies have shown that these biomarkers can act as early predictors of individual radiosensitivity in patients undergoing radiotherapy to monitor risk and biomarker response see, e.g., Refs. 46 and 47), as well as predictive markers for therapeutic response (see, e.g. Refs. 48-50).

[0083] In summary, the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can transition two radiation responsive protein biomarkers in blood lymphocytes to a high throughput ELISA platform and apply exemplary machine learning procedures towards the development of an accurate and rapid biodosimetry tool for early population triage.

Exemplary Materials and Methods

Exemplary Peripheral Blood Sample Collection

[0084] Human peripheral blood samples (~ 10 ml) can be collected from 11 healthy donors (male and female), aged 22 - 66 years old, with no previous radiation exposure in the 6 months prior to the day of blood draw.

Non-human primate (NHP; Macaca mulatto) blood samples (~ 6 ml) can be collected from 32 (male and female) NHPs

Exemplary Blood Sample and Animal Irradiation’.

[0085] For the ex vivo studies using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, blood samples from both adult humans and NHPs can be aliquoted in 15 ml polypropylene tubes (Corning, Glendale, AZ; #352095) and irradiated ex vivo using an X-RAD 320 biological irradiator (Precision X-Ray Inc., North Branford, CT) up to total doses of 0 (mock irradiated), 1, 2, 3, 4, or 5 Gy X-rays under the following conditions: 1.5 mm Al, 0.25 Cu, 1.25 Sn and 320 kVp, 12.5 mA, FSD 40, 0.95 Gy/min, custom-made home filter. Before each sample is exposed to radiation, the dose rate can be verified using a RadCal 10X6-6 ion chamber (Monrovia, CA; calibrated annually by RadCal). [0086] For the in vivo irradiations, e.g., NHPs can be whole body irradiated to total -body doses up to 10 Gy with a Varian Trilogy linear accelerator (LINAC) accelerator (Varian Medical Systems, Palo Alto, CA) within a specially designed lucite container. An on-site medical physicist can perform all the dose calculations for radiation exposure. Prior to irradiation, while in animal housing, each animal can be anesthetized with a ketamine-containing anesthetic mixture and placed into a custom-designed Plexiglas box inside a high efficiency particulate air (HEP A) filtered transport cage on a wheeled cart for transport to the LINAC. This exemplary transport system can be configured and/or designed with, e.g., a viewing window to permit continuous monitoring and observation during transport and partial pressure of oxygen (PO2) can be monitored continuously. The transport cart can be escorted by veterinary staff and with two-way radio link to the animal facility in case of any emergency issues. After the irradiation, the animals can be transported back to housing, placed back into their cages, and continuously monitored until recovered from anesthesia. At ~2-4 weeks prior to the irradiation, control blood samples can be drawn from the same animals who are similarly anesthetized and transported to the LINAC, but unexposed to X rays.

Exemplary Sample Preparation: Isolation of Lymphocytes and Cell Lysate Preparation [0087] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, lymphocytes can be isolated from the whole blood using density gradient centrifugation. Ficoll Histopaque medium (Sigma Aldrich, St. Louis, MO, #10771-human and #10831-NHP) can be added first to 15 ml SepMate™ tubes (STEMCELL™ Technologies; Vancouver, BC, #85415), and 1 ml of blood sample can be gently poured down the side of the tube. Samples can be centrifuged at 1200 x g for 10 minutes and the top layer containing peripheral blood mononuclear cells (PBMC) can be transferred to a fresh 15 ml polypropylene tube and washed with 1XPBS (Gibco, Grand Island, NY). The washed PBMCs can be aliquoted (~1 x 10⁶ / ml) into two Matrix™ 1.0 mL microtubes (Thermo Fisher Scientific™, Waltham, MA, #3740TS) per sample with complete RPMI (15% FBS, 1% Pen- Strep) and cultured at 37 U, 5% CO2 for 1 and 2 days. After the culture time, for BAX ELISA cell lysate preparation the cells can be spun down and washed with 1XPBS, and then chilled IX Cell Extraction Buffer PTR (Abeam, Waltham, MA, #ab 193970) can be added to the cell pellet and incubated on ice for 20 minutes. After the incubation, cells can be centrifuged at 18,000 x g for 20 minutes at 4°C and for DDB2 the cells can be suspended in IxPBS with IX protease inhibitor cocktail HALT (Thermo Fisher Scientific™, Waltham, MA, #87785) and repeat freeze thawed for 3 times and the sample can be centrifuged 14,000 x g for 10 minutes, supernatant can be stored at -80°C until use.

Exemplary Cell Count and Viability

[0088] Cell count and viability staining utilizing the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be performed on the 24 and 48 hrs PBMC cultures. Cells can be stained with Acridine Orange/Propidium Iodide (AO/PI) viability dye (e.g., Logos Biosystems, Annandale, VA, #F23001) and loaded into PhotonSlide™ (Logos Bio-systems, #L 12005). A LUNA-FL™ Dual Fluorescence Cell Counter (Logos Biosystems, #L20001) can be used to automatically count and determine the viability percentage of the cells, as per manufacturer’s instructions.

Exemplary Protein Analyses of Cell Lysates by Enzyme-Linked Immunosorbent Assays (ELISA)

[0089] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be used to quantify total protein in the cell lysates using Pierce™ BCA protein assay kits (Thermo Fisher Scientific, Rockford, IL, #23225) as per the manufacturer’s instruction to develop and interpolate concentrations from a standard curve. The human and NHP immunoassays can be performed in duplicate with a conventional ELISA sandwich format for two different protein targets, BAX and DDB2, using commercially available kits from Abeam (Waltham, MA, #ab!99080) and AFG bioscience (Northbrook, IL, #EK712088), respectively. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can utilize the absorbance readings at 450 nm, with reference to the standard curve and used average difference data between control and test samples as readout. For example, Figures 8A and 8B show exemplary standard curves for BAX and DDB2 respectively. The plates can be read using BioTek Synergy Hl Multimode Microplate Reader (e g., Agilent Technologies, Santa Clara, CA) and can be analyzed using the built-in Gen5 software. Optical density readings can then be interpreted using GainData®, Arigo Biolaboratories’ online calculator to plot standard curves and interpolate unknown concentrations. Exemplary Statistical Analysis

[0090] Exemplary statistical analyses can be performed using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, and graphs can be generated using, e.g., GraphPad Prism (version 10; GraphPad Software, Inc., La Jolla, CA). Human and NHP lymphocyte viability can be analyzed according to exemplary embodiments as functions of dose separately for each day via Pearson’s correlation and linear least squares regression. Separately, both BAX and DDB2 concentrations can be correlated with lymphocyte viability using Pearson’s (DDB2) and Spearman (BAX) correlations.

[0091] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, BAX and DDB2 (and FDXR, ACTN1, TSPYL2, (and p53) intracellular) protein expression levels) in the human and NHP samples can be compared to each other across different doses using 2-, 3- way, and repeated measures ANOVA tests and by calculating or otherwise determining Pearson’s correlations between the average concentration of the biomarker and dose. For humans, a further ANOVA test can be executed using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure to determine if there is a significant difference in biomarker expression in male and female subjects. For the in vivo NHP study, data points from in-vivo irradiated NHPs can be compared to confirm differences in biomarker levels across timepoints using a repeated-measures ANOVA.

Exemplary Dose Reconstruction

[0092] According to the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, dose reconstruction calculations for the ex-vivo studies can be performed using, e.g., Python 3.10, Jupyter notebooks. Those samples (e.g., only), which can have both BAX and DDB2 (and FDXR, ACTN1, TSPYL2, (and p53) intracellular) protein expression levels) measurements in the sample, can be used for either NHPs or humans. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can perform, e.g., 50:50 splitting of the data set into training and testing parts. The Boruta procedure can be used as an initial screening step (on training data) to discard the least important variables for distinguishing between unirradiated and irradiated samples (labeled by the Exposure index variable, where 1 = irradiated, 0 = unirradiated). Separately, an exemplary regression analysis can be performed using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure to reconstruct dose quantitatively for humans or NHPs. Boruta screening can also be used.

[0093] Boruta can create “shadow features” (randomized copies of original features) and compares their importance using Random Forest regressor or classifier models. If an original feature's importance is significantly higher than the maximum importance of the shadow features using z scores, it is kept; otherwise, it is dropped. This exemplary iterative process can continue until all features are either confirmed important or unimportant using a pre-defined significance threshold.

[0094] With the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, using the retained predictors, several machine learning (ML) procedures (e.g., linear regression, random forest, XGBoost, LightGBM, CatBoost, elastic net and support vector machines for regression tasks, and logistic regression, CatBoost, XGBoost, random forest, K-nearest neighbors, and naive Bayes for classification tasks) can be fitted to the exemplary training data with repeated cross-validation and can be evaluated on testing data. For the regression task, root mean squared error (RMSE) can be used by exemplary embodiments of the present disclosure as the main metric to assess performance and mean absolute error (MAE) and coefficient of determination (R²) can also be calculated. For the classification task, balanced accuracy can be used.

[0095] Using the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure, the stacking approach can be used to integrate the outputs of these different exemplary ML models to generate an ensemble. It can be performed separately for each task. In stacking, several ML methods (e.g., levelO models) can be applied to the training data with repeated k-fold cross validation. Exemplary predictions of each levelO model on out of sample data instances (e.g., those withheld during cross validation) can be recorded. These exemplary predictions can serve as inputs to train a meta-model (level 1) which can learn how to best combine the predictions of the levelO models to predict the outcome variable. Then, the whole ensemble (e.g., levelO and level 1) can make predictions on testing data, according to exemplary embodiments of the present disclosure. Often (but not always) this approach performs better than a single best levelO model. For example, certain samples may be difficult to predict for some models, but easier for other models, so information from several models can be complementary and improve overall predictions. Achieving an improvement in performance depends on the complexity of the problem and whether it is sufficiently well represented by the training data and complex enough that there is more to learn by combining predictions. Data for days 1 and 2 can be combined, since the Boruta procedure considers the Day variable to be unimportant for both humans and NHPs for the purposes of exposure classification and dose reconstruction.

Exemplary Public health significance

[0096] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure are relevant to advancing the development of a high throughput biodosimetry device that can be used to accurately determine absorbed dose in exposed individuals (and unexposed) across clinically relevant doses (0 to 8 Gy), up to a week after a mass-casualty R/N emergency. The exemplary blood-based biodosimetry configuration of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure is intended for use across the general population, which highlights the important need to identify special populations (children, elderly, immune status, diseases, stressors, inflammation) that could potentially confound the accuracy of the dose estimations. Additionally, the effects of partial body exposures can be critical for investigation. Thus, it is essential that biodosimetry tools according to the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure perform independent/irrespective of these biological variables and exposure conditions.

Exemplary Embodiments

[0097] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can provide a simple and rapid blood test to support early medical treatment decisions after a radiological/nuclear emergency: exemplary high throughput the FAST DOSE device of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can use an integrative proteomic approach and state-of-the-art, customized ML workflow to determine absorbed radiation dose for radiological triage and dose categorization used to score the severity of radiation exposure with improved accuracy.

[0098] Integration of targeted intracellular and blood plasma protein biomarkers to increase the accuracy of individual absorbed dose estimates: The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure represent a streamlined multiparametric bioassay approach that can utilize different strengths of radio-responsive protein biomarkers in blood leukocytes and plasma for in-vivo dose reconstruction. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure illustrate the possibility of performing longitudinal proteomic measurements in tandem using the highly translational pre-clinical and human models after total and partial body exposures.

Detailed exemplary application of a large range of potentially confounding biological variables

[0099] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure span ex vivo and in vivo radiation models to evaluate whether potential confounders including sex, age, immune status, and chronic health conditions (diabetes, inflammation; mild kidney disease) are confounding and can affect the dose prediction accuracy of exemplary FAST-DOSE biodosimetry tool.

[00100] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure provide a customized ML platform for performing multiparametric dose predictions: The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure show that it is possible to utilize an ML platform to first identify the best predictors of dose (using synthetic noise variables as benchmarks of predictor performance) and then integrate the proteomic biomarker signals to build unique algorithms for accurate dose prediction across a large dose range. ML learning and regression-based modeling efforts can account for potential confounding biological variables, non-linear dose response shapes, and interactions between variables. The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can provide the use of model stacking in the field of biodosimetry- this powerful ML assembling technique is not yet commonly used in the field.

[00101] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure provide a novel blood protein biomarker signature for use in radiological emergencies up to 14 days after radiation exposure: the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure show that it is possible to develop a novel blood protein biomarker signature (a comprehensive profile of individual biomarker levels) that can be used with custom-trained ML algorithms to accurately determine clinical absorbed dose categories (includes 0-<2 Gy, 2-4 Gy, 4-6 Gy and > 6 Gy) that can be used to predict the severity of injury from a single blood sample collected in-the-field following a mass-casualty R/N incident.

[00102] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be highly translatable between human, NHP and rodent radiation models: The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can be compliant with FDA biomarker qualification guidelines to test the translation of the blood protein biomarkers from pre-clinical animal models (NHP, rodent) to human (radiotherapy).

[00103] The exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can also provide a negative exposure test: the negative exposure test of exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure can include collecting fresh- unirradiated samples from preclinical and human (radiotherapy) studies and evaluating baseline protein levels in leukocytes and plasma at 0 Gy. These samples can play a critical role in training exemplary custom machine learning (ML) procedures of the exemplary systems, methods, and computer accessible medium according to the exemplary embodiments of the present disclosure. Importantly, a negative test can provide valuable information by identifying individuals who have not been exposed and relieving large stress for the medically concerned citizens.

[00104] Figure 9 shows a block diagram of an exemplary embodiment of a system according to the present disclosure. For example, exemplary procedures in accordance with the present disclosure described herein can be performed by a processing arrangement and/or a computing arrangement (e.g., computer hardware arrangement) 905. Such processing/computing arrangement 905 can be, for example entirely or a part of, or include, but not limited to, a computer/processor 910 that can include, for example one or more microprocessors, and use instructions stored on a computer-accessible medium (e.g., RAM, ROM, hard drive, or other storage device).

[00105] As illustrated in Figure 9, for example a computer-accessible medium 915 (e.g., as described herein above, a storage device such as a hard disk, floppy disk, memory stick, CD- ROM, RAM, ROM, etc., or a collection thereof) can be provided (e g., in communication with the processing arrangement 905). The computer-accessible medium 915 can contain executable instructions 920 thereon. In addition or alternatively, a storage arrangement 925 can be provided separately from the computer-accessible medium 915, which can provide the instructions to the processing arrangement 905 so as to configure the processing arrangement to execute certain exemplary procedures, processes, and methods, as described herein above, for example. Further, the exemplary processing arrangement 905 can be provided with or include an input/output ports 935, which can include, for example a wired network, a wireless network, the internet, an intranet, a data collection probe, a sensor, etc. As shown in Figure 9, the exemplary processing arrangement 905 can be in communication with an exemplary display arrangement 930, which, according to certain exemplary embodiments of the present disclosure, can be a touch-screen configured for inputting information to the processing arrangement in addition to outputting information from the processing arrangement, for example. Further, the exemplary display arrangement 930 and/or a storage arrangement 925 can be used to display and/or store data in a user-accessible format and/or user-readable format.

[00106] According to exemplary embodiments of the present disclosure, numerous specific details have been set forth. It is to be understood, however, that implementations of the disclosed technology can be practiced without these specific details. In other instances, well- known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description. References to “some examples,” “other examples,” “one example,” “an example,” “various examples,” “one embodiment,” “an embodiment,” “some embodiments,” “example embodiment,” “various embodiments,” “one implementation,” “an implementation,” “example implementation,” “various implementations,” “some implementations,” etc., indicate that the implementation(s) of the disclosed technology so described may include a particular feature, structure, or characteristic, but not every implementation necessarily includes the particular feature, structure, or characteristic. Further, repeated use of the phrases “in one example,” “in one exemplary embodiment,” or “in one implementation” does not necessarily refer to the same example, exemplary embodiment, or implementation, although it may.

[00107] As used herein, unless otherwise specified the use of the ordinal adjectives “first,” “second,” “third,” etc., to describe a common object, merely indicate that different instances of like objects are being referred to, and are not intended to imply that the objects so described must be in a given sequence, either temporally, spatially, in ranking, or in any other manner.

[00108] While certain implementations of the disclosed technology have been described in connection with what is presently considered to be the most practical and various implementations, it is to be understood that the disclosed technology is not to be limited to the disclosed implementations, but on the contrary, is intended to cover various modifications and equivalent arrangements included within the scope of the appended claims. Although specific terms are employed herein, they are used in a generic and descriptive sense only and not for purposes of limitation.

[00109] The foregoing merely illustrates the principles of the disclosure. Various modifications and alterations to the described embodiments will be apparent to those skilled in the art in view of the teachings herein. It will thus be appreciated that those skilled in the art will be able to devise numerous systems, arrangements, and procedures which, although not explicitly shown or described herein, embody the principles of the disclosure and can be thus within the spirit and scope of the disclosure. Various different exemplary embodiments can be used together with one another, as well as interchangeably therewith, as should be understood by those having ordinary skill in the art. In addition, certain terms used in the present disclosure, including the specification and drawings, can be used synonymously in certain instances, including, but not limited to, for example, data and information. It should be understood that, while these words, and/or other words that can be synonymous to one another, can be used synonymously herein, that there can be instances when such words can be intended to not be used synonymously. Further, to the extent that the prior art knowledge has not been explicitly incorporated by reference herein above, it is explicitly incorporated herein in its entirety. All publications referenced are incorporated herein by reference in their entireties. [00110] Throughout the disclosure, the following terms take at least the meanings explicitly associated herein, unless the context clearly dictates otherwise. The term “or” is intended to mean an inclusive “or.” Further, the terms “a,” “an,” and “the” are intended to mean one or more unless specified otherwise or clear from the context to be directed to a singular form.

[00111] This written description uses examples to disclose certain implementations of the disclosed technology, including the best mode, and also to enable any person skilled in the art to practice certain implementations of the disclosed technology, including making and using any devices or systems and performing any incorporated methods.

Exemplary References:

1 Buddemeier, B. & Dillon, M. Key response planning factors for the aftermath of nuclear terrorism. (Lawrence Livermore National Lab.(LLNL), Livermore, CA (United States), 2009).

2 Buddemeier, B. Reducing the Consequences of a Nuclear Detonation; Recent Research and Guidance. Reducing the Consequences of a Nuclear Detonation; Recent Research and Guidance, na, no. 2, June 21, 2010, pp. 28-38 (2010).

3 Alai, M. & Neuscamman, S. Task 4 Improvised Nuclear Device Response Curves. (Lawrence Livermore National Lab.(LLNL), Livermore, CA (United States), (2016).

4 Knebel, A. R. et al. Allocation of scarce resources after a nuclear detonation: setting the context. Disaster medicine and public health preparedness 5, S20-S31 (2011).

5 Casagrande, R. et al. Using the model of resource and time-based triage (MORTT) to guide scarce resource allocation in the aftermath of a nuclear detonation. Disaster medicine and public health preparedness 5, S98-S110 (2011).

6 DiCarlo, A. L. et al. Radiation injury after a nuclear detonation: medical consequences and the need for scarce resources allocation. Disaster medicine and public health preparedness 5, S32-S44 (2011).

7 Singh, V. K. et al. Acute radiation syndrome: An update on biomarkers for radiation injury. Journal of Radiation and Cancer Research 9, 132-146 (2018).

8 Satyamitra, M. M., DiCarlo, A. L., Hollingsworth, B. A., Winters, T. A. & Taliaferro, L. P. (Radiation Research Society, (2022). 9 Hick, J. L et al. Proposed “exposure and symptom triage”(EAST) tool to assess radiation exposure after a nuclear detonation. Disaster medicine and public health preparedness 12, 386-395 (2018).

10 Sandgren, D. J. et al. Biodosimetry Assessment Tool (BAT) software — dose prediction algorithms. Health Physics 99, S171-S183 (2010).

11 Hu, S., Blakely, W. F. & Cucinotta, F. A. HEMODOSE: a biodosimetry tool based on multi -type blood cell counts. Health physics 109, 54 (2015).

12 Capaccio, C. et al. CytoRADx: A high-throughput, standardized biodosimetry diagnostic system based on the cytokinesis-block micronucleus assay. Radiation Research 196, 523- 534 (2021).

13 Wang, Q. et al. Development of the FAST-DOSE assay system for high-throughput biodosimetry and radiation triage. Scientific reports 10, 12716 (2020).

14 Sproull, M. T., Camphausen, K. A. & Koblentz, G. D. Biodosimetry: A future tool for medical management of radiological emergencies. Health security 15, 599-610 (2017).

15 Jacobs, A. R. et al. Role of a high throughput biodosimetry test in treatment prioritization after a nuclear incident. International journal of radiation biology 96, 57-66 (2020).

16 Balog, R. P. et al. Development of a biodosimeter for radiation triage using novel blood protein biomarker panels in humans and non-human primates. International Journal of Radiation Biology 96, 22-34 (2020).

17 Hoffmeyer, M. R., Gillis, K., Park, J. G., Murugan, V. & LaBaer, J. Making the Case for Absorbed Radiation Response Biodosimetry-Utility of a High-Throughput Biodosimetry System. Radiation Research 196, 535-546 (2021).

18 Lee, Y. et al. Candidate protein markers for radiation biodosimetry in the hematopoietically humanized mouse model. Scientific reports 8, 13557 (2018).

19 Nemzow, L., Boehringer, T., Bacon, B. & Turner, H. C. Development of a human peripheral blood ex vivo model for rapid protein biomarker detection and applications to radiation biodosimetry. Pios one 18, e0289634 (2023). 0 Wang, Q. et al. Biomarkers for Radiation Biodosimetry and Correlation with Hematopoietic Injury in a Humanized Mouse Model. Radiation Research 202:541-551 (2024). 21 Nemzow, L. et al. Validation of a blood biomarker panel for machine learning-based radiation biodosimetry in juvenile and adult C57BL/6 mice. Scientific Reports 15, 4969 (2025).

22 Shuryak, I. et al. Machine learning approach for quantitative biodosimetry of partial-body or total-body radiation exposures by combining radiation-responsive biomarkers. Scientific Reports 13, 949 (2023).

23 Kitada, S., Krajewski, S., Miyashita, T., Krajewska, M. & Reed, J. C. Gamma-radiation induces upregulation of Bax protein and apoptosis in radiosensitive cells in vivo. Oncogene 12, 187-192 (1996).

24 Pena-Blanco, A. & Garcia-Saez, A. J. Bax, Bak and beyond — mitochondrial performance in apoptosis. The FEBS journal 285, 416-431 (2018).

25 Cai, L.-H. et al. DDB2 and MDM2 genes are promising markers for radiation diagnosis and estimation of radiation dose independent of trauma and burns. Functional & Integrative Genomics 23, 294, (2023).

26 Kanagaraj, K. et al. BAX and DDB2 as biomarkers for acute radiation exposure in the human blood ex vivo and non-human primate models. Scientific Reports 14, 19345 (2024).

27 Blakely, W. F. et al. Multiple parameter radiation injury assessment using a nonhuman primate radiation model — biodosimetry applications. Health physics 98, 153-159 (2010).

28 Ossetrova, N. I., Sandgren, D. J. & Blakely, W. F. Protein biomarkers for enhancement of radiation dose and injury assessment in nonhuman primate total-body irradiation model. Radiation protection dosimetry 159, 61-76 (2014).

29 Balog, R.P. et al. Development of a point-of-care radiation biodosimeter: studies using novel protein biomarker panels in non-human primates. International journal of radiation biology, 96, 35-46 (2020).

30 Balog, R.P. et al. Development of a point-of-care radiation biodosimeter: studies using novel protein biomarker panels in non-human primates. International journal of radiation biology, 96, 35-46 (2020).

31 Herodin, F. et al. Assessment of total- and partial -body irradiation in a baboon model: preliminary results of a kinetic study including clinical, physical, and biological parameters. Health physics, 103, 143-9 (2012). 32 Ossetrova, N.I. etal. Early-response biomarkers for assessment of radiation exposure in a mouse total-body irradiation model. Health physics, 106, 772-86 (2014).

33 Ossetrova, N.I. etal. Non-human Primate Total-body Irradiation Model with Limited and Full Medical Supportive Care Including Filgrastim for Biodosimetry and Injury Assessment. Radiation Protection Dosimetry, 172, 174-191 (2016).

34 Crowther JR. The ELISA guidebook: Springer Science & Business Media; 2008.

35. Ludwig, G.V, et al. Concepts for the Development of Immunodiagnostic Assays for

Detection and Diagnosis of Biothreat Agents. In: Lindler LE, Lebeda FJ, Korch GW, editors. Biological Weapons Defense: Infectious Diseases and Counterbioterrorism. Totowa, NJ: Humana Press p. 551-79 (2005).

36 Lundberg SM, Lee SI. A Unified Approach to Interpreting Model Predictions. arXiv. 2017;1705:07874.

37 Breiman L. Random forests. Machine Learning, 45:5-32 (2001).

38 Chen T, Guestrin C. XGBoost: A scalable tree boosting system. 2016 8. Report No.: 9781450342322.

39 Zhang C, Lei X, Liu L. Predicting Metabolite-Disease Associations Based on LightGBM Model. Front Genet, 12:660275 (2021).

40 Hancock JT, Khoshgoftaar TM. CatBoost for big data: an interdisciplinary review. J Big Data, 7(1):94 (2020).

41 Awad M, Khanna R. Support Vector Regression. Berkeley, CA., USA: Apress (2015).

42 Lindberg, J. C. H. 'J'accuse. !': the continuous failure to address radiophobia and placing radiation in perspective. J Radiol Prot 41, (2021).

43 Ropeik, D. The dangers of radiophobia. Bulletin of the Atomic Scientists 72, 311-317, (2016).

44 Homer, M. J. et al. United States Department of Health and Human Services Biodosimetry and radiological/nuclear medical countermeasure programs. Radiation Protection Dosimetry 171, 85-98 (2016).

45 Singh, V. K. & Seed, T. M. A review of radiation countermeasures focusing on injuryspecific medicinals and regulatory approval status: part I. Radiation sub-syndromes, animal models and FDA-approved countermeasures. Int J Radiat Biol 93, 851-869, (2017).

46 Hoffmeyer et al. Making the Case for Absorbed Radiation Response Biodosimetry - Utility of a High-Throughput Biodosimetry System. Radiation Research, 196:535-546 (2021).

47 Palumbo, E. et al. Individual Radiosensitivity in Oncological Patients: Linking Adverse Normal Tissue Reactions and Genetic Features. Front Oncol 9, 987, (2019). Kitada, S., Krajewski, S., Miyashita, T., Krajewska, M. & Reed, J. C. Gamma-radiation induces upregulation of Bax protein and apoptosis in radiosensitive cells in vivo. Oncogene 12, 187-192 (1996). Mackey, T. J., Borkowski, A., Amin, P., Jacobs, S. C. & Kyprianou, N. bcl-2/bax ratio as a predictive marker for therapeutic response to radiotherapy in patients with prostate cancer. Urology 52, 1085-1090, (1998). Pisani, C. et al. Apoptotic and predictive factors by Bax, Caspases 3/9, Bcl-2, p53 and Ki- 67 in prostate cancer after 12 Gy single-dose. Scientific Reports 10, 7050, (2020). Wang, P.-G. et al. Lower expression of Bax predicts poor clinical outcome in patients with glioma after curative resection and radiotherapy /chemotherapy. Journal of Neuro¬

Oncology 141, 71-81, (2019).

Claims

WHAT IS CLAIMED IS;

1. A method for detecting a radiation exposure, comprising: providing at least one blood lymphocyte or plasma protein digital information into a machine learning model; and generating, by the machine learning model, exposure data associated with the radiation exposure based on the provided at least one blood lymphocyte or plasma protein digital information.

2. The method of claim 1, wherein the at least one lymphocyte or plasma protein digital information is from a blood sample.

3. The method of claim 1, wherein the exposure data comprises a radiation exposure prediction data.

4. The method of claim 1, wherein the exposure data comprises a radiation dose prediction data.

5. The method of claim 1, wherein the machine learning model generates the exposure data based on a single blood lymphocyte or plasma protein digital information.

6. The method of claim 5, wherein the exposure data comprises a radiation exposure prediction data that is based on the single lymphocyte or plasma protein digital information.

7. The method of claim 1, wherein the machine learning model generates the exposure data based on a combination of at least two blood biomarkers from the at least one lymphocyte or plasma protein digital information, a differential blood cell count or a biomarker ratio information.

8. The method of claim 7, wherein the exposure data comprises a radiation dose prediction data based on the at least two blood biomarkers from the at least one lymphocyte or plasma protein digital information, the differential blood cell counts or the biomarker ratio information.

9. The method of claim 1, wherein an accuracy of the generated exposure data is directly correlated with a number of lymphocyte or blood plasma proteins, differential blood cell counts or biomarker ratio protein information that is input into the machine learning model.

10. The method of claim 1, wherein an accuracy of the generated exposure data is directly correlated with a type of lymphocyte or blood plasma protein, differential blood cell counts or biomarker ratio information that is input into the machine learning model.

11. The method of claim 1, wherein the machine learning model generates the exposure data up to a week for a triage or up to 14 days for a dose categorization and a radiation severity scoring after an exposure to an ionizing radiation.

12. The method of claim 1, further comprising, generating, by the machine learning model, a specific treatment based on the generated exposure data.

13. A system for detecting a radiation exposure, the method comprising: at least one processor configured to: provide at least one blood lymphocyte or plasma protein digital information into a machine learning model; and generate, by the machine learning model, exposure data associated with the radiation exposure based on the provided at least one blood lymphocyte or plasma protein digital information.

14. The system of claim 13, wherein the at least one lymphocyte or plasma protein digital information is from a blood sample.

15. The system of claim 13, wherein the exposure data comprises a radiation exposure prediction data.

16. The system of claim 13, wherein the exposure data comprises a radiation dose prediction data.

17. The system of claim 13, wherein the machine learning model generates the exposure data based on a single blood lymphocyte or plasma protein digital information.

18. The system of claim 17, wherein the exposure data comprises a radiation exposure prediction data that is based on the single lymphocyte or plasma protein digital information.

19. The system of claim 13, wherein the machine learning model generates the exposure data based on a combination of at least two blood biomarkers from the at least one lymphocyte or plasma protein digital information, a differential blood cell counts or a biomarker ratio information.

20. The system of claim 19, wherein the exposure data comprises a radiation dose prediction data based on the at least two blood biomarkers from the at least one lymphocyte or plasma protein digital information, the differential blood cell counts or the biomarker ratio information.

21. The system of claim 13, wherein an accuracy of the generated exposure data is directly correlated with a number of lymphocyte or blood plasma proteins, differential blood cell counts or biomarker ratio protein information that is input into the machine learning model.

22. The system of claim 13, wherein an accuracy of the generated exposure data is directly correlated with a type of lymphocyte or blood plasma protein, differential blood cell counts or biomarker ratio information that is input into the machine learning model.

23. The system of claim 13, wherein the machine learning model generates the exposure data up to a week for a triage or up to 14 days for a dose categorization and a radiation severity scoring after an exposure to an ionizing radiation.

24. The system of claim 13, further comprising, generating, by the machine learning model, a specific treatment based on the generated exposure data.

25. A non-transitory computer accessible medium which includes software thereon for facilitating detection of a radiation exposure wherein, when at least one computer processor executes the software, the computer processor is configured to perform the procedures, comprising: providing at least one blood lymphocyte or plasma protein digital information into a machine learning model; and generating, by the machine learning model, exposure data associated with the radiation exposure based on the provided at least one blood lymphocyte or plasma protein digital information.

26. The non-transitory computer accessible medium of claim 25, wherein the at least one lymphocyte or plasma protein digital information is from a blood sample.

27. The non-transitory computer accessible medium of claim 25, wherein the exposure data comprises a radiation exposure prediction data.

28. The non-transitory computer accessible medium of claim 25, wherein the exposure data comprises a radiation dose prediction data.

29. The non-transitory computer accessible medium of claim 25, wherein the machine learning model generates the exposure data based on a single blood lymphocyte or plasma protein digital information.

30. The non-transitory computer accessible medium of claim 29, wherein the exposure data comprises a radiation exposure prediction data that is based on the single lymphocyte or plasma protein digital information.

31. The non-transitory computer accessible medium of claim 25, wherein the machine learning model generates the exposure data based on a combination of at least two blood biomarkers from the at least one lymphocyte or plasma protein digital information, a differential blood cell counts or a biomarker ratio information.

32. The non-transitory computer accessible medium of claim 31, wherein the exposure data comprises a radiation dose prediction data based on the at least two blood biomarkers from the at least one lymphocyte or plasma protein digital information, the differential blood cell counts or the biomarker ratio information.

33. The non-transitory computer accessible medium of claim 25, wherein an accuracy of the generated exposure data is directly correlated with a number of lymphocyte or blood plasma proteins, differential blood cell counts or biomarker ratio protein information that is input into the machine learning model.

34. The non-transitory computer accessible medium of claim 25, wherein an accuracy of the generated exposure data is directly correlated with a type of lymphocyte or blood plasma protein, differential blood cell counts or biomarker ratio information that is input into the machine learning model.

35. The non-transitory computer accessible medium of claim 25, wherein the machine learning model generates the exposure data up to a week for a triage or up to 14 days for a dose categorization and a radiation severity scoring after an exposure to an ionizing radiation.

36. The non-transitory computer accessible medium of claim 25, further comprising, generating, by the machine learning model, a specific treatment based on the generated exposure data.

37. A method for detecting and treating a radiation exposure, comprising: providing at least one blood protein biomarker information into a machine learning model; generating, by the machine learning model, radiation dose information based on the provided at least one protein information; and directing a specific medical treatment based on the radiation dose information.

38. A system for detecting and treating a radiation exposure, comprising: at least one computer processor configured to: provide at least one blood protein biomarker information into a machine learning model; generate, by the machine learning model, radiation dose information based on the provided at least one protein information; and direct a specific medical treatment based on the radiation dose information.

39. A non-transitory computer accessible medium which includes software thereon for facilitating detection and treatment of a radiation exposure wherein, when at least one computer processor executes the software, the computer processor is configured to perform the procedures, comprising: providing at least one blood protein biomarker information into a machine learning model; generating, by the machine learning model, radiation dose information based on the provided at least one protein information; and directing a specific medical treatment based on the radiation dose information.