[go: up one dir, main page]

WO2024202466A1 - Programme informatique, procédé de traitement d'informations et dispositif de traitement d'informations - Google Patents

Programme informatique, procédé de traitement d'informations et dispositif de traitement d'informations Download PDF

Info

Publication number
WO2024202466A1
WO2024202466A1 PCT/JP2024/002133 JP2024002133W WO2024202466A1 WO 2024202466 A1 WO2024202466 A1 WO 2024202466A1 JP 2024002133 W JP2024002133 W JP 2024002133W WO 2024202466 A1 WO2024202466 A1 WO 2024202466A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
image data
patient
endovascular treatment
vascular
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
PCT/JP2024/002133
Other languages
English (en)
Japanese (ja)
Inventor
耕太郎 楠
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Terumo Corp
Original Assignee
Terumo Corp
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Terumo Corp filed Critical Terumo Corp
Publication of WO2024202466A1 publication Critical patent/WO2024202466A1/fr
Anticipated expiration legal-status Critical
Pending legal-status Critical Current

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
    • A61B1/04—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor combined with photographic or television appliances
    • A61B1/045—Control thereof
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
    • A61B1/313—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor for introducing through surgical openings, e.g. laparoscopes
    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B8/00—Diagnosis using ultrasonic, sonic or infrasonic waves
    • A61B8/12—Diagnosis using ultrasonic, sonic or infrasonic waves in body cavities or body tracts, e.g. by using catheters

Definitions

  • the present invention relates to a computer program, an information processing method, and an information processing device.
  • Intravascular treatments such as percutaneous coronary intervention (PCI) are used as minimally invasive treatments for ischemic heart diseases, such as angina pectoris or myocardial infarction.
  • Intravascular imaging diagnostic devices such as ultrasound imaging diagnostic devices using ultrasound (IVUS: IntraVascular Ultra Sound) and optical coherence imaging diagnostic devices using near-infrared rays (OCT: Optical Coherence Tomography), are used for preoperative diagnosis in endovascular treatments or to confirm the results after surgery (see, for example, Patent Document 1).
  • the prognosis of patients who have undergone endovascular treatment is thought to be influenced by the experience and skills of the doctor performing the treatment, as well as the patient's medical history, but it is difficult to predict the risk of a patient being readmitted to hospital after endovascular treatment.
  • the objective of this disclosure is to provide a computer program, an information processing method, and an information processing device that can calculate the risk of readmission for patients who have undergone endovascular treatment.
  • the present invention provides a computer program that (1) causes a computer to execute a process of acquiring tomographic image data obtained by scanning a patient's blood vessels after endovascular treatment and the patient's medical record information, calculating vascular information indicating the characteristics of the blood vessels after endovascular treatment based on the acquired tomographic image data, and inputting the calculated vascular information and the acquired medical record information into a learning model that outputs the readmission risk of the patient who has undergone endovascular treatment when vascular information based on the tomographic image data of the blood vessels and the patient's medical record information are input.
  • the vascular information includes information regarding the presence or absence of vascular dissection, poor stent compression, poor stent expansion, protrusion, or minimum stent area (MSA), which are calculated based on tomographic image data after endovascular treatment.
  • MSA minimum stent area
  • the computer program of (1) or (2) above acquires tomographic image data obtained by scanning the patient's blood vessels before endovascular treatment, calculates vascular information indicating characteristics of the blood vessels before endovascular treatment based on the acquired tomographic image data of the blood vessels before endovascular treatment, and calculates the risk of readmission of the patient who underwent endovascular treatment by inputting the calculated vascular information based on the tomographic image data of the blood vessels before endovascular treatment, the vascular information based on the tomographic image data of the blood vessels after endovascular treatment, and the acquired medical record information into the learning model.
  • the vascular information includes information relating to plaque burden, lipid plaque, attenuating plaque, calcified plaque, or unstable plaque calculated based on tomographic image data before endovascular treatment.
  • the tomographic image data includes ultrasound tomographic image data obtained by scanning a blood vessel using ultrasound, and optical coherence tomographic image data obtained by scanning a blood vessel using light.
  • the medical record information includes information relating to renal damage.
  • the information processing method of the present invention is (7) an information processing method for acquiring tomographic image data and the patient's medical record information obtained by scanning the patient's blood vessels after endovascular treatment, calculating vascular information indicating the characteristics of the blood vessels after endovascular treatment based on the acquired tomographic image data, and inputting the calculated vascular information and the acquired medical record information into a learning model that outputs the readmission risk of the patient who underwent endovascular treatment when vascular information based on the tomographic image data of the blood vessels and the patient's medical record information are input.
  • the information processing device of the present invention (8) is an information processing device that includes a learning model that outputs the risk of readmission of a patient who has undergone endovascular treatment when vascular information based on vascular tomographic image data and the patient's medical record information are input, an acquisition unit that acquires the tomographic image data and the patient's medical record information obtained by scanning the patient's blood vessels after endovascular treatment, and a processing unit that calculates vascular information indicating the characteristics of the blood vessels after endovascular treatment based on the acquired tomographic image data, and calculates the risk of readmission of a patient who has undergone endovascular treatment by inputting the calculated vascular information and the acquired medical record information.
  • FIG. 1 is a block diagram showing an example of the configuration of an information processing device according to a first embodiment
  • FIG. 2 is an explanatory diagram showing an information processing method according to the first embodiment
  • 4 is a flowchart showing an information processing procedure according to the first embodiment
  • FIG. 13 is a schematic diagram showing an example of a readmission risk display screen.
  • FIG. 11 is an explanatory diagram showing an information processing method according to a second embodiment.
  • (Embodiment 1) 1 is a block diagram showing an example of the configuration of an information processing device 1 according to embodiment 1.
  • the information processing device 1 is a computer, and includes a processing unit 11, a storage unit 12, a display unit 13, an operation unit 14, and a communication unit 15. Each unit is connected via a bus.
  • the information processing device 1 may be a multi-computer consisting of multiple computers, or may be a virtual machine virtually constructed by software.
  • the processing unit 11 is an arithmetic processing device or processor having one or more CPUs (Central Processing Units), MPUs (Micro-Processing Units), GPUs (Graphics Processing Units), GPGPUs (General-purpose computing on graphics processing units), TPUs (Tensor Processing Units), etc.
  • the processing unit 11 reads out and executes a computer program P (program product) stored in the memory unit 12, thereby implementing the information processing method of embodiment 1 and calculating the risk of readmission of a patient who has undergone endovascular treatment.
  • a computer program P program product
  • the communication unit 15 includes a communication circuit for performing communication processing with the diagnostic imaging device 3.
  • the communication unit 15 transmits and receives various information with the diagnostic imaging device 3.
  • the imaging diagnostic device 3 constitutes an imaging diagnostic system using, for example, a dual-type catheter equipped with both intravascular ultrasound (IVUS) and optical coherence tomography (OCT) functions.
  • the imaging diagnostic device 3 includes an imaging diagnostic catheter, an MDU (Motor Drive Unit), and an image processing device.
  • the imaging diagnostic catheter has a sensor unit that emits ultrasound or light at the tip.
  • the MDU can perform a pull-back operation to rotate the sensor unit and shaft inserted into the probe of the imaging diagnostic catheter in a circumferential direction while pulling them toward the MDU side at a constant speed.
  • the sensor unit rotates while moving from the tip side to the base end by the pull-back operation, and continuously scans the inside of the blood vessel at a predetermined time interval, and the imaging diagnostic device 3 can continuously generate multiple tomographic images approximately perpendicular to the probe based on the scanning results.
  • the image processing unit acquires signal data based on reflected waves of ultrasound or light emitted from the sensor unit in the radial direction of the diagnostic imaging catheter toward the tubular organ, and generates IVUS image data (ultrasound tomographic image data) and OCT image data (optical coherence tomographic image data) that visualize the reflection intensity versus the distance from the sensor unit based on the acquired signal data.
  • the diagnostic imaging device 3 transmits the generated IVUS image data and OCT image data to the information processing device 1.
  • the information processing device 1 receives the IVUS image data and OCT image data transmitted from the diagnostic imaging device 3 at the communication unit 15.
  • the IVUS image data and OCT image data will be collectively referred to as tomographic image data.
  • the memory unit 12 has, for example, a main memory unit and an auxiliary memory unit.
  • the main memory unit is a temporary storage area such as SRAM (Static Random Access Memory), DRAM (Dynamic Random Access Memory), or flash memory, and temporarily stores data necessary for the processing unit 11 to execute arithmetic processing.
  • the auxiliary memory unit is a storage device such as a hard disk or EEPROM (Electrically Erasable Programmable ROM).
  • the auxiliary memory unit stores the computer program P executed by the processing unit 11, the learning model 2, the electronic medical record DB 12a, the tomographic image DB 12b, and various other data necessary for processing. Details of the learning model 2 will be described later.
  • the electronic medical record DB 12a stores the medical record information of patients undergoing endovascular treatment.
  • the patient's name, date of birth, sex, height, weight, and medical history are stored in a patient ID that identifies the patient. It is preferable that the medical record information includes information indicating the presence or absence of renal impairment.
  • the tomographic image DB 12b stores the IVUS image data and OCT image data of the patient that the information processing device 1 receives via the communication unit 15. Specifically, the tomographic image DB 12b stores the IVUS image data and OCT image data before the endovascular treatment and the IVUS image data and OCT image data after the endovascular treatment in association with the patient ID.
  • the auxiliary storage unit may be an external storage device connected to the information processing device 1.
  • the computer program P may be written to the auxiliary storage unit during the manufacturing stage of the information processing device 1, or the computer program P may be distributed by an external information processing device 1 and acquired by the information processing device 1 via communication and stored in the auxiliary storage unit.
  • the computer program P may be recorded in a readable manner on a recording medium 1a such as a magnetic disk, optical disk, or semiconductor memory, and may be read from the recording medium 1a by a reading device and stored in the auxiliary storage unit.
  • the display unit 13 is, for example, a display device such as a liquid crystal panel or an organic EL (Electro Luminescence) display.
  • a display device such as a liquid crystal panel or an organic EL (Electro Luminescence) display.
  • the operation unit 14 is, for example, an input device such as a touch panel, a hardware keyboard, or a mouse built into the display unit 13.
  • a user can operate the information processing device 1 using the operation unit 14 to input information required for the readmission risk of patients who have undergone endovascular treatment and to display the readmission risk.
  • FIG. 2 is an explanatory diagram showing an information processing method according to the first embodiment.
  • the information processing device 1 includes a first feature calculation unit 11a and a second feature calculation unit 11b as functional units of the processing unit 11.
  • the first feature calculation unit 11a calculates vascular information representing the characteristics of vascular lesions from the IVUS image data and OCT image data of the patient before endovascular treatment.
  • the vascular information calculated from the IVUS image data and OCT image data before endovascular treatment includes information related to plaque burden, lipid plaque, attenuating plaque, calcified plaque, and unstable plaque.
  • the vascular information calculated in this manner is input to the learning model 2.
  • the information related to plaque burden includes the ratio of the plaque cross-sectional area to the blood vessel cross-sectional area. It is preferable that the plaque burden is calculated using both IVUS image data and OCT image data.
  • the first feature calculation unit 11a calculates the blood vessel diameter based on the IVUS image data and calculates the lumen diameter based on the OCT image data.
  • the blood vessel diameter is a distance obtained by detecting a closed curve showing the boundary of the external elastic lamina (blood vessel) on the image, setting the center of gravity based on the detected closed curve, and then measuring the linear distance between the center of gravity and the closed curve on the image.
  • the blood vessel diameter may be a statistical value such as the average diameter, maximum diameter, and minimum diameter of the distance measured at multiple circumferential positions.
  • the method of calculating the blood vessel diameter is not particularly limited, and it is sufficient that the square value roughly corresponds to the cross-sectional area of the inner space of the blood vessel when a state in which no plaque is attached is assumed.
  • the lumen diameter is a distance obtained by detecting a closed curve showing the boundary of the lumen on an image, setting a center of gravity based on the detected closed curve, and then measuring the linear distance between the center of gravity and the closed curve on the image.
  • the lumen diameter may be a statistical value such as an average diameter, a maximum diameter, or a minimum diameter of the distance measured at a plurality of circumferential positions.
  • the method of calculating the lumen diameter is not particularly limited, and the square value may be any value that roughly corresponds to the cross-sectional area of the inner space of the blood vessel to which plaque is attached.
  • the blood vessel diameter may be a numerical value that represents the size of the inner space of the blood vessel before plaque is attached
  • the lumen diameter may be a numerical value that represents the size of the inner space of the blood vessel to which plaque is attached.
  • the first feature amount calculation unit 11a calculates the ratio (plaque burden) of the cross-sectional area of the plaque to the blood vessel media area (blood vessel cross-sectional area) based on the calculated blood vessel diameter and lumen diameter.
  • the plaque burden is expressed as (blood vessel media area - blood vessel lumen area) / blood vessel media area.
  • the first feature amount calculator 11a calculates the maximum or average value of plaque burden (plaque burden per unit length) as vascular information.
  • the average plaque burden is expressed as the sum of the plaque cross-sectional area over the entire lesion/lesion length.
  • the information related to lipid plaque includes the presence or absence of lipid plaque, the number of lipid plaques, the maximum angle indicating the circumferential spread of lipid plaque in the cross section of the blood vessel, the area of lipid plaque, etc.
  • the area of lipid plaque includes, for example, the maximum or average value of the lipid plaque area in the lesion (lipid plaque area per unit length). The average value is expressed as the sum of the lipid plaque area over the entire lesion/lesion length. It is preferable that the information related to lipid plaque is calculated based on OCT image data and IVUS image data.
  • Information related to attenuating plaque includes the presence or absence of attenuating plaque, the number of attenuating plaques, the maximum angle indicating the circumferential spread of attenuating plaque in the cross section of the blood vessel, the area of attenuating plaque, etc.
  • the area of attenuating plaque includes, for example, the maximum or average value of the attenuating plaque area in the lesion (attenuating plaque area per unit length). The average value is expressed as the sum of the attenuating plaque areas over the entire lesion/lesion length.
  • Attenuating plaque is, for example, lipid plaque.
  • Attenuating plaque is plaque defined based on IVUS image data, and is therefore calculated based on IVUS image data.
  • Information related to calcified plaque includes the presence or absence of calcified plaque, the number of calcified plaques, the maximum angle indicating the circumferential spread of calcified plaque in the cross section of the blood vessel, the area of calcified plaque, etc.
  • the area of calcified plaque includes, for example, the maximum or average value of the area of calcified plaque in the lesion (calcified plaque area per unit length). The average value is expressed as the total value of the area of calcified plaque over the entire lesion/lesion length. It is preferable that the presence or absence and the number of calcified plaques are calculated based on IVUS image data. It is preferable that the area of calcified plaques is calculated based on OCT image data.
  • Information related to unstable plaque includes the presence or absence of unstable plaque, the number of unstable plaques, the area of unstable plaques, etc. It is preferable that information related to unstable plaques is calculated using both IVUS image data and OCT image data.
  • the second feature calculation unit 11b calculates vascular information representing the characteristics of the vascular lesion from the IVUS image data and OCT image data of the patient after endovascular treatment.
  • the vascular information calculated from the IVUS image data and OCT image data after endovascular treatment includes information related to the presence or absence of vascular dissection, poor stent crimping, poor stent expansion, tissue protrusion (vascular tissue protruding into the lumen from the stent mesh), and minimum stent area.
  • the vascular information calculated in this way is input to learning model 2.
  • Information related to vascular dissection includes, for example, the presence or absence of a vascular dissection, the maximum angle indicating the circumferential extent of the vascular dissection in the vascular cross section, and the length of the vascular dissection in the longitudinal direction of the vascular dissection.
  • Information related to stent poor crimping includes, for example, the presence or absence of stent poor crimping, the maximum distance between the stent and the blood vessel wall in the area where the stent poor crimping occurs, and the length of the blood vessel in the longitudinal direction in the area where the stent poor crimping occurs.
  • Information related to poor stent expansion includes the ratio of the blood vessel diameter in the normal area to the minimum stent area. It is preferable that the minimum stent area is calculated using OCT image data, and the blood vessel diameter in the normal area is calculated using IVUS image data.
  • the learning model 2 is a trained model that has been trained to output a patient's risk of readmission from vascular information based on IVUS and OCT images and the patient's medical record information through machine learning using teacher data.
  • the learning model 2 performs a predetermined calculation on the input values and outputs the calculation results, and the memory unit 12 stores data such as coefficients and thresholds of functions that define this calculation as the learning model 2.
  • the processing unit 11 By reading the data stored as the learning model 2, the processing unit 11 becomes able to execute calculation processing of the risk of readmission from vascular information and medical record information extracted from IVUS and OCT images.
  • the learning process of the learning model 2 is performed by a learning computer.
  • Data related to the learned learning model 2 may be provided in the form of distribution via a communication network, similar to the computer program P, or may be provided in the form of being recorded on a recording medium 1a.
  • the learning model 2 is a neural network having an input layer 2a to which vascular information calculated based on IVUS images and OCT images of the patient before endovascular treatment, vascular information calculated based on IVUS images and OCT images of the patient after endovascular treatment, and medical record information of the patient are input, an intermediate layer 2b that extracts features of the vascular information and medical record information, and an output layer 2c that outputs the risk of readmission calculated based on the extracted features.
  • the input layer 2a of the neural network has multiple neurons to which vascular information and medical record information are input, and passes each input data to the intermediate layer 2b.
  • the intermediate layer 2b has multiple layers made up of multiple neurons. Each layer extracts features related to the risk of readmission from the input data and passes them on to the next layer in turn, with the last layer passing them on to the output layer 2c.
  • the output layer 2c includes neurons that output the calculation results, and these neurons output the readmission risk of a patient who has undergone endovascular treatment.
  • the readmission risk is, for example, a real value between 0 and 1. "0" indicates the lowest readmission risk, and "1" indicates the highest readmission risk.
  • the learning model 2 is a general neural network, but it may also be a model having a configuration such as other neural networks such as a transformer, an SVM (Support Vector Machine), a Bayesian network, or a regression tree.
  • a transformer an example has been described in which the learning model 2 is a general neural network, but it may also be a model having a configuration such as other neural networks such as a transformer, an SVM (Support Vector Machine), a Bayesian network, or a regression tree.
  • SVM Small Vector Machine
  • Bayesian network a Bayesian network
  • a learning method for the learning model 2 will be described.
  • the computer collects a plurality of IVUS images, OCT images, and medical record information that are the source of the training data.
  • the computer generates training data by adding training data indicating the risk of re-admission of the patient to the vascular information and medical record information calculated based on the IVUS images and OCT images of the patient.
  • the computer generates the learning model 2 by performing machine learning or deep learning on the pre-training neural network model using the generated training data.
  • the computer inputs the vascular information and medical record information contained in the learning data into the pre-learning neural network model, and obtains the readmission risk output from the output layer 2c through arithmetic processing in the intermediate layer 2b.
  • the computer compares the readmission risk output from the output layer 2c with the readmission risk indicated by the teacher data, and optimizes the parameters used in the arithmetic processing in the intermediate layer 2b so that the readmission risk output from the output layer 2c approaches the correct value.
  • the parameters are, for example, weights (coupling coefficients) between neurons.
  • the computer optimizes various parameters using the steepest descent method or the like.
  • the information processing device 1 obtains a trained learning model 2 by repeatedly performing the above process based on the training data of a large number of patients contained in the training data.
  • FIG. 3 is a flowchart showing the information processing procedure according to the first embodiment.
  • the processing unit 11 of the information processing device 1 reads out the medical record information of a patient who is to be evaluated for risk of readmission from the electronic medical record DB 12a (step S11).
  • the processing unit 11 acquires IVUS image data and OCT image data before and after endovascular treatment of the patient who is to be evaluated for risk of readmission (step S12).
  • the processing unit 11 may be configured to acquire the IVUS image data and OCT image data from the imaging diagnostic device 3 via the communication unit 15, or may be configured to read out these data from the tomographic image DB 12b.
  • the processing unit 11 which functions as the first feature calculation unit 11a, calculates vascular information such as information related to plaque burden, lipid plaque, attenuating plaque, calcified plaque, and unstable plaque based on the IVUS image data and OCT image data before the endovascular treatment (step S13).
  • the processing unit 11 which functions as the second feature calculation unit 11b, calculates vascular information such as information related to vascular dissection, poor stent compression, and poor stent expansion based on the IVUS image data and OCT image data after endovascular treatment (step S14).
  • the processing unit 11 inputs the vascular information before the endovascular treatment calculated in step S13, the vascular information calculated in step S14, and the medical record information acquired in step S11 into the learning model 2 to calculate the patient's risk of readmission (step S15).
  • the processing unit 11 displays the patient's risk of readmission on the display unit 13 (step S16), and ends the process.
  • FIG. 4 is a schematic diagram showing an example of the readmission risk display screen 4.
  • the processing unit 11 may be configured to display the readmission risk display screen 4 as shown in FIG. 4 in step S16.
  • the readmission risk display screen 4 includes, for example, a medical record information display section 41, a pre-treatment tomographic image display section 42, a post-treatment tomographic image display section 43, a pre-treatment vascular information display section 44, a post-treatment vascular information display section 45, a readmission risk display gauge 46, and an output result explanation section 47.
  • the medical record information display unit 41 is a frame that displays the electronic medical record information of the patient who is the subject of an evaluation of the risk of readmission.
  • the processing unit 11 displays the medical record information acquired in step S11 on the medical record information display unit 41.
  • the pre-treatment tomographic image display section 42 and the post-treatment tomographic image display section 43 are frames that display tomographic image data before and after endovascular treatment.
  • the pre-treatment vascular information display section 44 is a frame that displays vascular information before endovascular treatment
  • the post-treatment vascular information display section 45 is a frame that displays vascular information after endovascular treatment.
  • the readmission risk display gauge 46 includes a bar to indicate the degree of risk, ranging from low risk to high risk, and a line image indicating the degree of risk, and indicates the patient's risk of readmission depending on the display position of the line image.
  • the output result explanation unit 47 displays the importance of each of the multiple vascular information and medical record information with respect to the readmission risk output by the learning model 2.
  • the processing unit 11 can calculate the importance of each of the multiple vascular information and medical record information with respect to the readmission risk output by the learning model 2, for example, by using XAI (XAI: Explainable AI) technology.
  • XAI is a technology that outputs the factor contribution rate for each piece of information input to the machine learning model and presents to humans in an explainable manner which input information led to the output result (readmission risk).
  • XRAI EXplanation with Ranked Area Integrals
  • the output result explanation unit 47 displays, for example, the importance of each of the various vascular information before and after endovascular treatment, and the electronic medical record information, i.e., the magnitude of the influence on the output result of the learning model 2, in numerical or graphical form.
  • the output result explanation unit 47 displays characters representing the various vascular information before and after endovascular treatment, and the electronic medical record information, in association with a meter bar indicating the importance of each piece of information.
  • the magnitude of the importance of each piece of information is represented by the length of the meter bar.
  • the information processing device 1 configured in this manner can calculate the risk of readmission of a patient who has undergone endovascular treatment based on the tomographic image data of the patient's blood vessels and the patient's medical record information.
  • the information processing device 1 according to the second embodiment differs from the first embodiment in that IVUS image data and OCT image data are also input to the learning model. Since the other configurations of the information processing device 1 are similar to those of the information processing device 1 according to the first embodiment, the same reference numerals are used for similar parts and detailed description will be omitted.
  • the learning model 202 includes, for example, a convolutional neural network (CNN) that has been trained by deep learning.
  • the learning model 202 according to the second embodiment is a neural network having an input layer 202a to which, for example, vascular information calculated based on IVUS images and OCT images of a patient before endovascular treatment, vascular information calculated based on IVUS images and OCT images of a patient after endovascular treatment, IVUS image data before and after endovascular treatment, OCT image data before and after endovascular treatment, and medical record information of the patient are input, an intermediate layer 202b that extracts features of the vascular information, tomographic image data, and medical record information, and an output layer 202c that outputs a readmission risk calculated based on the extracted features.
  • CNN convolutional neural network
  • step S15 described in embodiment 1 the processing unit 11 inputs vascular information before and after endovascular treatment, IVUS image data before and after endovascular treatment, OCT image data before and after endovascular treatment, and medical record information into the learning model 202 to calculate the patient's risk of readmission.
  • vascular information that cannot be extracted by rule-based processing can be extracted from tomographic image data, and the risk of readmission to the patient can be calculated more accurately.
  • Information processing device 1a Recording medium 2: Learning model 3: Image diagnostic device 4: Readmission risk display screen 11: Processing unit 12: Memory unit 13: Display unit 14: Operation unit 15: Communication unit 41: Medical record information display unit 42: Pre-treatment tomographic image display unit 43: Post-treatment tomographic image display unit 44: Pre-treatment vascular information display unit 45: Post-treatment vascular information display unit 46: Readmission risk display gauge 47: Output result explanation unit 202: Learning model P: Computer program

Landscapes

  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Surgery (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Biomedical Technology (AREA)
  • Veterinary Medicine (AREA)
  • Pathology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Biophysics (AREA)
  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Heart & Thoracic Surgery (AREA)
  • Medical Informatics (AREA)
  • Molecular Biology (AREA)
  • Animal Behavior & Ethology (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Optics & Photonics (AREA)
  • Image Analysis (AREA)

Abstract

La présente invention comprend les étapes consistant à : acquérir des données d'image tomographique obtenues par balayage d'un vaisseau sanguin d'un patient après un traitement endovasculaire et des informations de dossier médical du patient ; calculer, sur la base des données d'image tomographique acquises, des informations de vaisseau sanguin indiquant un élément caractéristique du vaisseau sanguin après le traitement endovasculaire ; et entrer les informations de vaisseau sanguin calculées et les informations de dossier médical acquises dans un modèle entraîné pour calculer le risque de nouvelle hospitalisation du patient qui a reçu le traitement endovasculaire, le modèle entraîné étant configuré pour sortir le risque de nouvelle hospitalisation du patient qui a reçu le traitement endovasculaire lors de l'entrée des informations de vaisseau sanguin sur la base des données d'image tomographique du vaisseau sanguin et des informations de dossier médical du patient.
PCT/JP2024/002133 2023-03-30 2024-01-25 Programme informatique, procédé de traitement d'informations et dispositif de traitement d'informations Pending WO2024202466A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2023055715 2023-03-30
JP2023-055715 2023-03-30

Publications (1)

Publication Number Publication Date
WO2024202466A1 true WO2024202466A1 (fr) 2024-10-03

Family

ID=92904911

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2024/002133 Pending WO2024202466A1 (fr) 2023-03-30 2024-01-25 Programme informatique, procédé de traitement d'informations et dispositif de traitement d'informations

Country Status (1)

Country Link
WO (1) WO2024202466A1 (fr)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110313788A1 (en) * 2010-06-17 2011-12-22 Cerner Innovation, Inc. Readmission risk assesment
JP2017527418A (ja) * 2014-07-24 2017-09-21 ライトラボ・イメージング・インコーポレーテッド ステント及び血管の可視化及び診断のシステム、装置、及び方法
JP2017536917A (ja) * 2014-12-10 2017-12-14 コーニンクレッカ フィリップス エヌ ヴェＫｏｎｉｎｋｌｉｊｋｅ Ｐｈｉｌｉｐｓ Ｎ．Ｖ． ステント内再狭窄予測のためのデバイス、システム、及び方法
CN109637663A (zh) * 2018-11-14 2019-04-16 浙江大学山东工业技术研究院 一种基于机器学习的经皮冠状动脉介入治疗心血管不良事件的预测装置
US20210251577A1 (en) * 2020-02-17 2021-08-19 Siemens Healthcare Gmbh Machine-Based Risk Prediction for Peri-Procedural Myocardial Infarction or Complication from Medical Data
WO2021199960A1 (fr) * 2020-03-30 2021-10-07 テルモ株式会社 Programme, procédé de traitement d'informations, et système de traitement d'informations
JP2023032648A (ja) * 2021-08-27 2023-03-09 キヤノンメディカルシステムズ株式会社 医用情報処理装置、医用情報処理方法、およびプログラム

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20110313788A1 (en) * 2010-06-17 2011-12-22 Cerner Innovation, Inc. Readmission risk assesment
JP2017527418A (ja) * 2014-07-24 2017-09-21 ライトラボ・イメージング・インコーポレーテッド ステント及び血管の可視化及び診断のシステム、装置、及び方法
JP2017536917A (ja) * 2014-12-10 2017-12-14 コーニンクレッカ フィリップス エヌ ヴェＫｏｎｉｎｋｌｉｊｋｅ Ｐｈｉｌｉｐｓ Ｎ．Ｖ． ステント内再狭窄予測のためのデバイス、システム、及び方法
CN109637663A (zh) * 2018-11-14 2019-04-16 浙江大学山东工业技术研究院 一种基于机器学习的经皮冠状动脉介入治疗心血管不良事件的预测装置
US20210251577A1 (en) * 2020-02-17 2021-08-19 Siemens Healthcare Gmbh Machine-Based Risk Prediction for Peri-Procedural Myocardial Infarction or Complication from Medical Data
WO2021199960A1 (fr) * 2020-03-30 2021-10-07 テルモ株式会社 Programme, procédé de traitement d'informations, et système de traitement d'informations
JP2023032648A (ja) * 2021-08-27 2023-03-09 キヤノンメディカルシステムズ株式会社 医用情報処理装置、医用情報処理方法、およびプログラム

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
BODEWES, THOMAS ET AL.: "Risk factors for 30-day unplanned readmission following infrainguinal endovascular interventions", JOURNAL OF VASCULAR SURGERY, vol. 65, no. 2, 23 January 2017 (2017-01-23), pages 484 - 494, XP029897600, DOI: https://doi.org/10.1016/j.jvs. 2016.08.09 3 *

Similar Documents

Publication Publication Date Title
CN113544737B (zh) 用于动脉图像区域及其特征的分类的系统和方法
US12336763B2 (en) Systems and methods for treatment planning based on plaque progression and regression curves
Tesche et al. Machine learning and deep neural networks applications in coronary flow assessment: the case of computed tomography fractional flow reserve
US20110173027A1 (en) Health-risk metric determination and/or presentation
JP7426824B2 (ja) 非侵襲的イメージングベースのｆｆｒのインタラクションモニタリング
US20200359911A1 (en) Arterial wall characterization in optical coherence tomography imaging
JP2025120362A (ja) 情報処理装置、情報処理方法及びプログラム
WO2021193015A1 (fr) Programme, procédé de traitement d'informations, dispositif de traitement d'informations et procédé de génération de modèle
JP7536861B2 (ja) プログラム、情報処理方法、情報処理装置及びモデル生成方法
JP7561833B2 (ja) コンピュータプログラム、情報処理方法及び情報処理装置
JP7536862B2 (ja) プログラム、情報処理方法、情報処理装置及びモデル生成方法
WO2025002905A1 (fr) Caractérisation de tissu athéroscléreux basée sur une dynamique de traitement endovasculaire
US20230238148A1 (en) Information processing device, information processing method, and computer program
WO2024202466A1 (fr) Programme informatique, procédé de traitement d'informations et dispositif de traitement d'informations
JP7727586B2 (ja) 情報処理方法、プログラムおよび情報処理装置
EP4254428A1 (fr) Prédiction d'étape de procédé intravasculaire
WO2023186610A1 (fr) Prédiction d'étape de procédure intravasculaire
JP7667134B2 (ja) プログラム、情報処理方法、情報処理装置及びモデル生成方法
JP2021114139A (ja) 医用情報処理装置及び医用情報処理プログラム
EP4510962A1 (fr) Prédiction de la conformité d?un vaisseau en réponse à de multiples traitements potentiels
Shah et al. Adjunctive intra-coronary imaging for the assessment of coronary artery disease
CA3206860A1 (fr) Traitement d'une forme d'onde ultrasonore doppler arterielle
JP7753240B2 (ja) プログラム、情報処理装置及び情報処理方法
RU2834685C1 (ru) Способ оценки корректности установки потоконаправляющего стента
JP2025052747A (ja) 情報処理方法、プログラム及び情報処理装置

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 24778612

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE