WO2025054443A1

WO2025054443A1 - Deterministic seismocardiogram analysis for monitoring and diagnosis of valvular heart disease

Info

Publication number: WO2025054443A1
Application number: PCT/US2024/045572
Authority: WO
Inventors: Farrokh Ayazi; Brian SANG
Original assignee: Georgia Tech Research Institute; Georgia Tech Research Corp
Current assignee: Georgia Tech Research Institute; Georgia Tech Research Corp
Priority date: 2023-09-06
Filing date: 2024-09-06
Publication date: 2025-03-13
Anticipated expiration: 2026-03-06

Abstract

An exemplary embodiment of the present disclosure provides a system for monitoring a status of a heart disease. The system can comprise a wearable sensor and a controller. The wearable sensor can be configured generate or capture a signal indicative of vibrations from a heart of a user. The controller can be configured to extract S2 sound time splits of the heart based, at least in part, on the signal.

Description

DETERMINISTIC SEISMOCARDIOGRAM ANALYSIS FOR MONITORING AND DIAGNOSIS OF VALVULAR HEART DISEASE

CROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims the benefit of U.S. Provisional Application Serial No. 63/536,802, filed on 6 September 2023, which is incorporated herein by reference in its entirety as if fully set forth below.

FIELD OF THE DISCLOSURE

[0002] The various embodiments of the present disclosure relate generally to wearable devices for monitoring heart conditions.

BACKGROUND

[0003] Cardiovascular disease is a significant public health issue that affects more than 37 million individuals globally. Heart failure can result from various heart diseases such as aortic stenosis (AS). AS is a valvular heart disease caused by the calcification of the aortic valve. Over time, this calcification reduces the excursion of the aortic valve and thus a reduction of extracardiac perfusion. This calcification is usually due to age, but other contributing factors include high blood pressure, diabetes, obesity, male sex, and anatomic variants. AS is considered one of the most common valvular heart diseases as it affects 3% of population above 65 in developed countries. In addition, mortality rate can be as high as 50% within a year of the onset of symptom, but it can be mitigated with either surgical aortic valve replacement or transcatheter aortic valve replacement (TAVR). Screening and detection of the progression of AS would allow for earlier diagnosis, especially for people at risk. For asymptomatic AS patients, it is recommended to undergo an echocardiogram scan, the gold standard for heart assessment, every six-12 months; however, echocardiograms are costly, require office visits, and can have long wait times. Therefore, widespread availability of wearable devices for detection and monitoring progression of AS that are cost effective and can be used for at-home use are important.

[0004] AS can be detected through various means, in particular, paradoxical splitting of the second heart sound (S2). The S2 heart sound comprises of two main peaks, the aortic valve closure sound peak (A2) and the pulmonic valve closure sound peak (P2), as shown from a phonocardiogram (PCG) signal in FIG. 1. These two main peaks represent acoustic vibrations generated by the sharp incisura of aortic and pulmonic blood pressure due to the rapid deceleration of blood following the closure of the aortic and pulmonic valves. In healthy individuals, A2 and P2 have been empirically determined to occur sequentially. The time difference between A2 and P2 is referred to as S2 splitting. Paradoxical splitting occurs when P2 precedes A2 as A2 sound is delayed due to the increase of the impedance of the aorta and the obstruction to left ventricular outflow, both occurring due to AS.

[0005] Previously, paradoxical splitting has only been computationally analyzed using PCG signals from nonwearable digital stethoscopes. While digital stethoscopes are a popular means of capturing and extracting information from heart sounds, they are not suitable for long-term monitoring as a wearable device because of its bulky frame, its ability to pick up noise from other airborne sounds, and its requirement of active listening from trained professionals.

[0006] Seismocardiogram (SCG) is a noninvasive measurement of the acoustic vibrations of the heart through wearable accelerometers. SCG signals are traditionally captured by accelerometers placed on the exterior of the chest walls, and they have been popularly used in clinical studies. Studies have found that SCG signals from off-the-shelf accelerometers strongly correlate with echocardiogram readings. Specifically, the aortic peak of S2 heart sound from SCG occurs almost simultaneously with the echocardiogram reading’s aortic valve closure time frame. Not only that, but studies have also found that SCG signals to have slightly higher accuracy rate than PCG signals for estimating total systolic time when compared with echocardiogram readings. In addition, noninvasive wearable sensors are considered one of the most promising devices for at-home use especially for high-risk patients due to their ease of use, lightweight, and overall accessibility to a larger group of patients. However, off-the-shelf accelerometers lack the sensitivity to capture S2 heart splitting accurately as off-the-shelf accelerometers cannot capture the S2 pulmonic sound. To this date, there has not been any report of capturing and identifying S2 splitting from a wearable device, or from SCG signals. Accordingly, there is a need for such devices.

[0007] Additionally, the respiration phase may cause the S2 splitting to be modulated. It is known that during inspiration, the S2 splitting is larger than during expiration. As such, accurate measurement of S2 splitting can enable extraction of respiration rate of an individual, among other things. BRIEF SUMMARY

[0008] An exemplary embodiment of the present disclosure provides a system for monitoring a status of a heart and/or a heart disease. The system can comprise a wearable sensor and a controller. The wearable sensor can be configured to generate or capture a signal indicative of vibrations from a heart of a user. The controller can be configured to extract S2 sound time splits of the heart based, at least in part, on the signal.

[0009] In any of the embodiments disclosed herein, the sensor can comprise an accelerometer.

[0010] In any of the embodiments disclosed herein, the signal can be a seismocardiogram signal.

[0011] In any of the embodiments disclosed herein, the controller can be configured extract S2 sound time splits of the heart, at least in part, by determining S2 time windows in the signal. [0012] In any of the embodiments disclosed herein, the controller can be configured to identify S2 time windows in the signal, at least in part, by computing a Shannon energy of the signal.

[0013] In any of the embodiments disclosed herein, the controller can be further configured to identify A2 and P2 sounds in the S2 time windows.

[0014] In any of the embodiments disclosed herein, the controller can be further configured to identify A2 and P2 sounds in the S2 time windows, at least in part, by generating a timefrequency representation of the signal.

[0015] In any of the embodiments disclosed herein, the time-frequency representation of the signal can be a smoothed pseudo Wigner-Ville distribution.

[0016] In any of the embodiments disclosed herein, the controller can be further configured to predict a status of the heart disease based, at least in part, on the S2 sound time splits.

[0017] In any of the embodiments disclosed herein, the heart disease can be aortic stenosis.

[0018] In any of the embodiments disclosed herein, the heart disease can be valvular heart disease.

[0019] Another embodiment of the present disclosure provides a method of monitoring a status of a heart disease, the method comprising: receiving a signal from a wearable sensor, the signal indicative of vibrations from a heart of a user; extracting S2 sound time splits of the heart based, at least in part, on the signal.

[0020] In any of the embodiments disclosed herein, extracting S2 sound time splits can comprise determining S2 time windows in the signal. [0021] In any of the embodiments disclosed herein, determining S2 time windows in the signal can comprise computing a Shannon energy of the signal.

[0022] In any of the embodiments disclosed herein, extracting S2 sound time splits in the signal can further comprise identifying A2 and P2 sounds in the S2 time windows.

[0023] In any of the embodiments disclosed herein, identifying A2 and P2 sounds in the S2 sounds can comprise generating a time- frequency representation of the signal.

[0024] In any of the embodiments disclosed herein, the time-frequency representation of the signal can be a smoothed pseudo Wigner-Ville distribution.

[0025] In any of the embodiments disclosed herein, the method can further comprise predicting a status of the heart disease based, at least in part, on the S2 time splits.

[0026] These and other aspects of the present disclosure are described in the Detailed Description below and the accompanying drawings. Other aspects and features of embodiments will become apparent to those of ordinary skill in the art upon reviewing the following description of specific, exemplary embodiments in concert with the drawings. While features of the present disclosure may be discussed relative to certain embodiments and figures, all embodiments of the present disclosure can include one or more of the features discussed herein. Further, while one or more embodiments may be discussed as having certain advantageous features, one or more of such features may also be used with the various embodiments discussed herein. In similar fashion, while exemplary embodiments may be discussed below as device, system, or method embodiments, it is to be understood that such exemplary embodiments can be implemented in various devices, systems, and methods of the present disclosure.

BRIEF DESCRIPTION OF THE DRAWINGS

[0027] The following detailed description of specific embodiments of the disclosure will be better understood when read in conjunction with the appended drawings. For the purpose of illustrating the disclosure, specific embodiments are shown in the drawings. It should be understood, however, that the disclosure is not limited to the precise arrangements and instrumentalities of the embodiments shown in the drawings.

[0028] FIG. 1 provides a phonocardiogram (PCG) of a single heartbeat, in which the S2 sound encompasses both A2 and P2 peaks and occurs after S 1 sound.

[0029] FIG. 2 provides a block diagram of a system for monitoring a status of a heart disease, in accordance with some embodiments of the present disclosure. [0030] FIGs. 3A-B provide plots of SCG data from subject 2 with a time frame of 2.5-12.5 s represented with the normalized average Shannon energy to show peaks of SI by “X” (FIG.

3 A), and ECG signals S 1 peaks shown by abrupt peak in the same time-domain matching with the captured ACM SI peaks (FIG. 3B).

[0031] FIG. 4 provides a plot of SCG data from subject 2 with the S2 peaks landmarked by the circle within the same time domain of 2.5-12.5 s.

[0032] FIGs. 5A-B provide plots of cardiac cycles of a healthy subject with distinct S2 peaks shown with circle (FIG. 5A) and of an unhealthy subject with murmur shown in the left circle and distinct S2 peak shown in right circle (FIG. 5B).

[0033] FIG. 6 provides a plot of instantaneous frequency measurements from PCG signals of S2 sound with the highest peaks representing the start of each respective sound and the time difference of each peak is S2 time split.

[0034] FIGs. 7A-B provide plots showing a high BMI subject with normal S2 splitting from subject 7 with a time split of 39.28 ms (FIG. 7A) and a sixty-seven-year-old subject with normal S2 splitting from subject 9 with a time split of 24.36 ms (FIG. 7B).

[0035] FIGs. 8A-C provide plots showing mild AS paradoxical S2 splitting in subject 11 with a time split of -14.63 ms (FIG. 8A), moderate AS paradoxical S2 splitting in subject 14 with a time split of -24.27 ms (FIG. 8B), and severe AS paradoxical S2 splitting in subject 16 with a time split of-19.63 ms (FIG. 8C).

[0036] FIGs. 9A-B provides plots showing moderate aortic regurgitation paradoxical S2 splitting in subject 17 with a time split of -31.63 ms (FIG. 9A), and severe mitral regurgitation paradoxical S2 splitting in subject 18 with a time split of-14.27 ms (FIG. 9B).

[0037] FIGs. 10A-C provide correlation plots of both ACM and PCG signals of: (FIG. 10A) systolic; (FIG. 10B) diastolic; and (FIG. 10C) interbeat time interval.

[0038] FIGs. 11A-B provide (FIG. 11A) a comparison of 102 SCG S2 signals superimposed and (FIG. 1 IB) 101 PCG S2 signals superimposed, in which each plots’ signals were averaged and the average of each was represented by a black thicker line.

[0039] FIGs. 12A-B provide plots showing A2 and P2 intensity of S2 splitting at primary pulmonic location (FIG. 12 A), and A2 and slight P2 intensity of S2 splitting at secondary location (FIG. 12B). DETAILED DESCRIPTION

[0040] Although preferred exemplary embodiments of the disclosure are explained in detail, it is to be understood that other exemplary embodiments are contemplated. Accordingly, it is not intended that the disclosure is limited in its scope to the details of construction and arrangement of components set forth in the following description or illustrated in the drawings. The disclosure is capable of other exemplary embodiments and of being practiced or carried out in various ways. Also, in describing the preferred exemplary embodiments, specific terminology will be resorted to for the sake of clarity.

[0041] To facilitate an understanding of the principles and features of the present disclosure, various illustrative embodiments are explained below. The components, steps, and materials described hereinafter as making up various elements of the embodiments disclosed herein are intended to be illustrative and not restrictive. Many suitable components, steps, and materials that would perform the same or similar functions as the components, steps, and materials described herein are intended to be embraced within the scope of the disclosure. Such other components, steps, and materials not described herein can include, but are not limited to, similar components or steps that are developed after development of the embodiments disclosed herein.

[0042] As used in the specification and the appended claims, the singular forms “a,” “an” and “the” include plural referents unless the context clearly dictates otherwise.

[0043] Also, in describing the preferred exemplary embodiments, terminology will be resorted to for the sake of clarity. It is intended that each term contemplates its broadest meaning as understood by those skilled in the art and includes all technical equivalents which operate in a similar manner to accomplish a similar purpose.

[0044] Ranges can be expressed herein as from “about” or “approximately” one particular value and/or to “about” or “approximately” another particular value. When such a range is expressed, another exemplary embodiment includes from the one particular value and/or to the other particular value.

[0045] Similarly, as used herein, “substantially free” of something, or “substantially pure”, and like characterizations, can include both being “at least substantially free” of something, or “at least substantially pure”, and being “completely free” of something, or “completely pure”. [0046] By ‘ ‘comprising” or “containing” or “including” is meant that at least the named compound, member, particle, or method step is present in the composition or article or method, but does not exclude the presence of other compounds, materials, particles, method steps, even if the other such compounds, material, particles, method steps have the same function as what is named.

[0047] Mention of one or more method steps does not preclude the presence of additional method steps or intervening method steps between those steps expressly identified. Similarly, it is also to be understood that the mention of one or more components in a device or system does not preclude the presence of additional components or intervening components between those components expressly identified.

[0048] The materials described as making up the various members of the invention are intended to be illustrative and not restrictive. Many suitable materials that would perform the same or a similar function as the materials described herein are intended to be embraced within the scope of the invention. Such other materials not described herein can include, but are not limited to, for example, materials that are developed after the time of the development of the invention.

[0049] Reference will now be made in detail to exemplary embodiments of the disclosed technology, examples of which are illustrated in the accompanying drawings and disclosed herein. Wherever convenient, the same references numbers will be used throughout the drawings to refer to the same or like parts.

[0050] Disclosed herein are systems and methods that use a highly sensitive accelerometer to computationally extract S2 time split accurately and robustly to identify normal and paradoxical S2 splitting in healthy subjects and subjects with valvular heart disease.

[0051] As shown in FIG. 2, an exemplary embodiment of the present disclosure provides a system for monitoring a status of a heart disease. The system can comprise a wearable sensor 210 and a controller 220 (e.g., computing device). The sensor can be many sensors capable of detecting vibrations. The wearable sensor can be worn by a user and can be configured generate a signal indicative of vibrations from a heart of a user. For example, in some embodiments, the sensor can comprise an accelerometer positioned on the pulmonic region of the user proximate the user’s heart. The accelerometer can detect vibrations from the user’s heart and generate a signal indicative of those vibrations, such as a seismocardiogram (SCG) signal. The controller can receive the signal and can extract S2 sound time splits of the heart based, at least in a part, on the signal. A status of a heart disease, such as AS, can then be determined based on the S2 time splits. [0052] The S2 heart sound time splits can be determined a number of ways in accordance with various embodiments of the present disclosure. In some embodiments, the SI and S2 time windows can first be determined. This can be performed by computing a Shannon energy of the signal. Once the S2 time windows are identified, the A2 and P2 sounds can be identified in the S2 windows. This can be performed, at least in part, by generating a time-frequency representation of the signal, such as a smoothed pseudo Wigner-Ville distribution. Various details of algorithms for determining S2 time splits in a user are further explained in the example section below.

[0053] The controller/computing device 220 shown in FIG. 2 illustrates an exemplary computing device that can be used to implement the methods/algorithms (or one or more steps of the methods/algorithms) disclosed herein. For example, the computing device shown in FIG. 2 can receive data from the sensor 210 (e.g., accelerometer) and process the data in accordance with the various processes disclosed herein. As will be appreciated by one of skill in the art, the computing device 220 can be configured to implement all or some of the features described in relation to the methods disclosed herein. As shown, the computing device 220 may include a processor 222, an input/output ("I/O") device 224, a memory 230 containing an operating system ("OS") 232 and a program 236. In certain example implementations, the computing device 220 may be a single server or may be configured as a distributed computer system including multiple servers or computers that interoperate to perform one or more of the processes and functionalities associated with the disclosed embodiments. In some embodiments, computing device 220 may be one or more servers from a serverless or scaling server system. In some embodiments, the computing device 220 may further include a peripheral interface, a transceiver, a mobile network interface in communication with the processor 222, a bus configured to facilitate communication between the various components of the computing device 220, and a power source configured to power one or more components of the computing device 220.

[0054] A peripheral interface, for example, may include the hardware, firmware and/or software that enable(s) communication with various peripheral devices, such as sensors, media drives (e.g., magnetic disk, solid state, or optical disk drives), other processing devices, or any other input source used in connection with the disclosed technology. In some embodiments, a peripheral interface may include a serial port, a parallel port, a general-purpose input and output (GPIO) port, a game port, a universal serial bus (USB), a micro-USB port, a high definition multimedia interface (HD MI) port, a video port, an audio port, a Bluetooth™ port, a near- field communication (NFC) port, another like communication interface, or any combination thereof. [0055] In some embodiments, a transceiver may be configured to communicate with compatible devices and sensors when they are within a predetermined range. A transceiver may be compatible with one or more of: radio-frequency identification (RFID), near-field communication (NFC), Bluetooth™, low-energy Bluetooth™ (BLE), WiFi™, ZigBee™, ambient backscatter communications (ABC) protocols or similar technologies.

[0056] A mobile network interface may provide access to a cellular network, the Internet, or another wide-area or local area network. In some embodiments, a mobile network interface may include hardware, firmware, and/or software that allow(s) the processor(s) 222 to communicate with other devices via wired or wireless networks, whether local or wide area, private or public, as known in the art. A power source may be configured to provide an appropriate alternating current (AC) or direct current (DC) to power components.

[0057] The processor 222 may include one or more of a microprocessor, microcontroller, digital signal processor, co-processor or the like or combinations thereof capable of executing stored instructions and operating upon stored data. The memory 230 may include, in some implementations, one or more suitable types of memory (e.g. such as volatile or non-volatile memory, random access memory (RAM), read only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), magnetic disks, optical disks, floppy disks, hard disks, removable cartridges, flash memory, a redundant array of independent disks (RAID), and the like), for storing files including an operating system, application programs (including, for example, a web browser application, a widget or gadget engine, and or other applications, as necessary), executable instructions and data. In one embodiment, the processing techniques described herein may be implemented as a combination of executable instructions and data stored within the memory 230.

[0058] The processor 222 may be one or more known processing devices, such as, but not limited to, a microprocessor from the Pentium™ family manufactured by Intel™ or the Turion™ family manufactured by AMD™. The processor 222 may constitute a single core or multiple core processor that executes parallel processes simultaneously. For example, the processor 222 may be a single core processor that is configured with virtual processing technologies. In certain embodiments, the processor 222 may use logical processors to simultaneously execute and control multiple processes. The processor 222 may implement virtual machine technologies, or other similar known technologies to provide the ability to execute, control, run, manipulate, store, etc. multiple software processes, applications, programs, etc. The processor 222 may also comprise multiple processors, each of which is configured to implement one or more features/steps of the disclosed technology. One of ordinary skill in the art would understand that other types of processor arrangements could be implemented that provide for the capabilities disclosed herein.

[0059] In accordance with certain example implementations of the disclosed technology, the computing device 220 may include one or more storage devices configured to store information used by the processor 222 (or other components) to perform certain functions related to the disclosed embodiments. In one example, the computing device 220 may include the memory 230 that includes instructions to enable the processor 222 to execute one or more applications, such as server applications, network communication processes, and any other type of application or software known to be available on computer systems. Alternatively, the instructions, application programs, etc. may be stored in an external storage or available from a memory over a network. The one or more storage devices may be a volatile or non-volatile, magnetic, semiconductor, tape, optical, removable, non-removable, or other type of storage device or tangible computer-readable medium.

[0060] In one embodiment, the computing device 220 may include a memory 230 that includes instructions that, when executed by the processor 222, perform one or more processes consistent with the functionalities disclosed herein. Methods, systems, and articles of manufacture consistent with disclosed embodiments are not limited to separate programs or computers configured to perform dedicated tasks. For example, the computing device 220 may include the memory 230 that may include one or more programs 236 to perform one or more functions of the disclosed embodiments.

[0061] The processor 222 may execute one or more programs located remotely from the computing device 220. For example, the computing device 220 may access one or more remote programs that, when executed, perform functions related to disclosed embodiments.

[0062] The memory 230 may include one or more memory devices that store data and instructions used to perform one or more features of the disclosed embodiments. The memory 230 may also include any combination of one or more databases controlled by memory controller devices (e.g., server(s), etc.) or software, such as document management systems, Microsoft™ SQL databases, SharePoint™ databases, Oracle™ databases, Sybase™ databases, or other relational or non-relational databases. The memory 230 may include software components that, when executed by the processor 222, perform one or more processes consistent with the disclosed embodiments. In some examples, the memory 230 may include a database 234 configured to store various data described herein. For example, the database 234 can be configured to store the software repository 102 or data generated by the repository intent model 104 such as synopses of the computer instructions stored in the software repository 102, inputs received from a user (e.g., responses to questions or edits made to synopses), or other data that can be used to train the repository intent model 104.

[0063] The computing device 220 may also be communicatively connected to one or more memory devices (e.g., databases) locally or through a network. The remote memory devices may be configured to store information and may be accessed and/or managed by the computing device 220. By way of example, the remote memory devices may be document management systems, Microsoft™ SQL database, SharePoint™ databases, Oracle™ databases, Sybase™ databases, or other relational or non-relational databases. Systems and methods consistent with disclosed embodiments, however, are not limited to separate databases or even to the use of a database.

[0064] The computing device 220 may also include one or more I/O devices 224 that may comprise one or more user interfaces 226 for receiving signals or input from devices (e.g., sensors) and providing signals or output to one or more devices that allow data to be received and/or transmitted by the computing device 220. For example, the computing device 220 may include interface components, which may provide interfaces to one or more input devices, such as one or more keyboards, mouse devices, touch screens, track pads, trackballs, scroll wheels, digital cameras, microphones, sensors, and the like, that enable the computing device 220 to receive data from a user.

[0065] In example embodiments of the disclosed technology, the computing device 220 may include any number of hardware and/or software applications that are executed to facilitate any of the operations. The one or more I/O interfaces may be utilized to receive or collect data and/or user instructions from a wide variety of input devices. Received data may be processed by one or more computer processors as desired in various implementations of the disclosed technology and/or stored in one or more memory devices.

[0066] While the computing device 220 has been described as one form for implementing the techniques described herein, other, functionally equivalent, techniques may be employed. For example, some or all of the functionality implemented via executable instructions may also be implemented using firmware and/or hardware devices such as application specific integrated circuits (ASICs), programmable logic arrays, state machines, etc. Furthermore, other implementations of the computing device 220 may include a greater or lesser number of components than those illustrated.

EXAMPLES

[0067] Below, an exemplary system is described. The system is provided for illustrative purposes only and should be construed as limiting the scope of the claims submitted herewith.

[0068] Data Collection and Data Preparation

[0069] A sensitive, wearable accelerometer contact microphone (ACM) from StethX Microsystems, Atlanta, GA, USA, was utilized. The ACM uses a hermetically sealed microelectromechanical system (MEMS) device operating in vacuum with nanogap capacitive transducers that can capture wideband cardiopulmonary acoustic vibrations from de to 10 kHz with micro-g resolution without sensitivity to airborne sounds. The ACM sensor board can have a small form factor of 27 x 15 x 2.5 mm containing the 2.8 x 2.8 x 0.9 mm hermetically sealed MEMS chip and a CMOS application-specific integrated circuit (ASIC) for converting the MEMS chip capacitive output to robust digital signal.

[0070] SCG signals were collected from the ACM of ten individuals with no known heart condition and eight individuals with underlying cardiovascular condition of AS and other related heart diseases as diagnosed by a cardiologist with disease severity determined by echocardiograms readings. The healthy individuals were between the age of 20-69 years old with Body mass index (BMI) ranging from 21.5 to 37.9, while the unhealthy individuals were from 64 to 97 years old with BMI ranging from 23.2 to 32.8. The individuals were asked to sit upright and to breathe normally through their nose as the ACM was placed with medical tape on their pulmonic region, situated near the pulmonic valve, which is between the second and third intercostal spaces at the left sternal border and about an inch left from the sternum.

[0071] The ACM was connected via wire to a data acquisition hub unit (i.e., controller). In some embodiments, however, the sensor can be connected to the controller wirelessly. The 60 s of continual data collection was done on each subject. The SCG data from the ACM were then captured and processed in MATLAB. The SCG signal was filtered with a Butterworth bandpass filter of 5-200 Hz, corresponding to the heart sound frequency range typically used for SCG signals of 5-100 Hz to ensure the frequency contents of SCG signals along with matching the frequency range from past S2 splitting measurements. Additional filtering processes were not utilized to minimize data distortion. The filtered data were then parsed into smaller time windows to accurately measure S2 time splitting.

[0072] Time Window: Second Heartbeat Segmentation

[0073] To create S2 time windows for healthy subjects, heartbeat segmentation was completed based solely on the SCG signals collected from the ACM. S2 time window segmentation is extracted by the normalized average Shannon energy along with peak detection as SCG signals of healthy individuals usually comprise of SI and S2, with S2 having a smaller peak than S 1 ; past literatures have done similar extraction, but not with an accelerometer placed on the pulmonic region.

[0074] The method can comprise peak detection above a threshold to automatically reject extra peaks and augment weaker peaks that are possibly S 1 or S2, determined based on the use of the diastolic and systolic time along with the relative amplitude difference between the two. To accurately determine if the SI peaks identified by the ACM are indeed SI, one subject, specifically subject 2, was also attached with the BIOP AC Systems MP 160 Bionomadix Wireless electrocardiogram (ECG) sensors to compare if the ECG signals’ heartbeat interval matches with the S 1 landmarked major peaks. ECG signals and ACM data were downsampled accordingly to have their time-domain match.

[0075] The segmentation was extracted with the ACM placed on the pulmonic region. As shown in FIG. 3A, “X” was used as a landmark to represent which peaks were automatically analyzed to be considered SI. ECG measurements were also used to compare with the same time frames and shown similar heartbeat intervals as shown in FIG. 3B even when the ACM was placed on the pulmonic region.

[0076] The following method has shown to be 89% accurate in determining if the landmarked X is SI when compared with ECG signals. One reason as to why it is not more accurate may be due to some S 1 from the SCG signal having low normalized average Shannon energy which can cause missed peaks. SI heartbeat interval captured via the ACM placed on the sternum of the chest has also been shown to have similar heartbeat interval as the ECG heartbeat interval.

[0077] S2 peaks, within the same dataset, were also calculated and picked through normalized average Shannon energy as shown in FIG. 4 landmarked with circles. The method was able to accurately determine which peaks were S2 as they were the next biggest peaks, in terms of amplitude, after S 1 and had a relatively short period due to the known time difference of systole and diastole.

[0078] The following method was able to accurately determine which peaks were S2 86% of the time. Inaccurate S2 peaks were determined to be missed due to the variability of high normalized average Shannon energy of some S2 peaks due to the ACM location on the pulmonic region. Visual inspection of S2 s landmarked peaks, segmented through this method, was also conducted on all healthy subjects to affirm that these S2 time windows contained S2 for accurate time split measurement.

[0079] However, for subjects with AS and other valvular heart diseases, majority of the subjects likely have murmurs in their heart signals and their S2 sounds might have a relatively small normalized average Shannon energy due to a decrease in the aorta sound, which makes peak detection harder to identify, especially for S2 sound. Therefore, for unhealthy subjects, S2 was manually parsed by looking at the sharpest peaks that occur after S 1 and after an assumed systolic time. For example, SCG data of both healthy and unhealthy heart diagnosis subjects have been captured and are shown in FIGs. 5A-B. In FIG. 5A, SCG data of subject 4 show distinct S2 peaks with no other major peaks other than SI. However, subject 16’s SCG data, shown in FIG. 5B, have the presence of murmurs during the end of systole but still show prominent S2 peaks as shown with the dashed circle. Therefore, even with the presence of murmurs, S2 time split can still be calculated as S2 peaks tend to be louder than murmurs.

[0080] Using the identified S2 peaks for healthy and unhealthy subjects, time windows were created and set with a range of 90 ms before and after each S2 peak for S2 time split computation.

[0081] S2 Time Split Extraction

[0082] To extract the timing of S2 splitting and the occurrences of both A2 and P2 via SCG signals from the ACM data, time-frequency representation method can be utilized as an effective mean to determine and show the A2 and P2 peaks because of the importance of instantaneous frequency and amplitude; A2 and P2 sounds have been documented to start at their instantaneous frequency respective peaks, which can be defined as the location of signals’ spectral peaks as shown in FIG. 6. Exemplary time-frequency representation methods for measuring heart sounds and S2 time split from digital stethoscopes can include short-time Fourier transform (STFT), continuous wavelet transform (CWT), Wigner-Ville distribution (WVD), and smoothed pseudo Wigner-Ville distribution (SPWVD) to measure the instantaneous frequency and instantaneous amplitude. SPWVD and WVD are known to have superior resolution in comparison to STFT and CWT as SPWVD and WVD encompass both STFT and auto-correlation. However, utilization of WVD with varying frequencies such as heart sounds can cause cross terms to appear in graphs. These cross terms can be significantly reduced through smoothing in the time and frequency window domain via SPWVD. The SPWVD equation uses independent time windows, represented by g, and frequency windows, represented by H, to smoothen the cross terms as shown in the following equation, where t represents the time, represents the frequency, and r represents the autocorrelation time lag:

Equation 1 :

[0083] Utilizing the S2 time windows previously calculated, SPWVD was calculated in MATLAB using Equation 1 which outputted a 3-D plot with x-axis as timc. v-axis as frequency, and z-axis as intensity. A grading plot is then created with x-axis as time and the y-axis being the product of frequency and intensity. Afterward, A2 and P2 peaks are shown as the two largest peaks due to the peak’s respective highest frequency and intensity. Peak detection is then utilized in the grading plot to identify the time location of the two largest peaks as A2 is considered the largest peak and P2 is considered the second largest peak and measure S2 time split difference between A2 and P2. S2 time splitting can be calculated when there is a presence of two distinct peaks, which can be determined when the normalized amplitude of second largest peak, P2, is at least 20% the size of the largest peak, A2. Negative time splitting is shown if P2 occurs before A2 from both the output of the SPWVD and grading plot and is then considered paradoxical splitting instance.

[0084] Validation of S2 Time Split: SCG Versus PCG

[0085] To validate that the ACM’s SCG signals have similar fidelity for capturing S2 time split to that of PCG’s signals, the ACM SCG signals were compared to Eko CORE 3.0 digital stethoscope’s PCG signals. An initial study to compare and determine the variances of systolic, diastolic, and interbeat intervals from both the SCG and PCG signals from a subject, specifically subject 5, was done by placing both devices simultaneously near the pulmonic region. The ACM sensor was attached on the chest walls with medical tape while the digital stethoscope was attached below the ACM with a chest band Velcro strap. Both recordings were done simultaneously for 60 s. Afterward, the systolic, diastolic, and interbeat intervals of SCG and PCG are compared by utilizing the SI- S2 segmentation method and confirmed via manual inspection.

[0086] In addition, the S2 signal waveform of both SCG and PCG signals was compared on four healthy subjects, specifically subjects 1, 2, 3, and 5. Because of the relatively small pulmonic region and the damping effect, the ACM and digital stethoscope recordings were carried out consecutively. The ACM was first placed on the pulmonic region via medical tape and recorded for 90 s. Consecutively, the digital stethoscope recordings were carried out. The ACM was first placed on the pulmonic region via medical tape and recorded for 90 s. The signals were then parsed to locate and collect S2 signals. Afterward, all the SCG’s S2 signals were then averaged and compared to the average of all the PCG’s S2 sound signals. Finally, the S2 time split range was extracted and compared by the S2 time split computation from all four subjects for both their SCG and PCG signals. The signals were then parsed to locate and collect S2 signals. Afterward, all the SCG and PCG S2 signals were then averaged and compared. Finally, the SCG and PCG’s S2 time split range was extracted and compared.

[0087] Placement Sensitivity

[0088] Sensor placement regarding measuring S2 time split has also been considered as the pulmonic sound might not be fully captured by the ACM. For subject 2, the ACM was placed once on the pulmonic region. To compare the placement sensitivity, the ACM was also placed an inch to the left of the original placement and recorded for 60 s known as the “inching” method. The graded intensity curve of P2 amplitude, extracted via SPWVD, of both placements was compared visually. The intensity curves were compared at the peak of inspiration determined by the chest movement measured through a low-pass filter of 5 Hz of the SCG data.

[0089] RESULTS AND DISCUSSION

[0090] S2 Time Splits: Healthy Versus Unhealthy Subjects

[0091] Subjects with no known heart diseases have almost all their S2 splits showing normal splitting. As shown in Table I, normal S2 time split from subject 1-10 ranges from 9.97 to 64.23 ms, which is considered within the range of normal S2 time split reported by others. S2 splits have been captured on individuals with high BMI of above 30, who tend to have higher fat content in the pulmonic region as shown in subjects 5-8, showing that ACM can capture S2 splits on a wide range of subjects. In addition, we tested the ACM on subjects

NUMBER OF PARADOXICAL S2 SPLI WITH HEALTHY AND UNHEALTHY HEART DIAGNOSIS W DATA COLLECTION

Time Split

Subject Gende Age BMI Paradoxical Range Heart Diagnosis

# r Splits (ms)

1 F 20 24.8 0 12.80 to 59.62 Normal

2 M 23 21.7 0 18.12 to 59.74 Normal

3 M 22 26.8 0 15.21 to 56.32 Normal

4 M 21 27.8 1 -9.75 to 51.14 Normal

5 M 24 36.2 0 9.97 to 61.42 Normal

6 M 37 32.8 0 25.91 to 64.23 Normal

7 F 48 37.9 0 21.29 to 63.42 Normal

8 F 55 32.0 3 -10.45 to 35.32 Normal

9 F 67 26.4 0 16.86 to 28.82 Normal

10 F 69 29.4 0 11.00 to 31.54 Normal

11 M 74 24.7 10 -25.23 to 34.32 Mild Aortic Stenosis

12 F 88 Missing 10 -37.15 to 27.77 Moderate Aortic Stenosi

13 M 67 27.4 10 -21.43 to 38.98 Moderate Aortic Stenosi

14 M 86 23.2 14 -21.45 to 32.48 Moderate Aortic Stenosi

15 M 83 24.4 19 -38.23 to 21.18 Moderate to Severe Aort

Stenosis

16 F 76 32.8 23 -30.00 to 31.84 Severe Aortic Stenosis

17 F 97 23.2 11 -36.43 to 28.29 Moderate Aortic

Regurgitation

18 M 64 25.7 13 -24.87 to 37.83 Severe Mitral

Regurgitation who share the same age range as subjects who typically have heart disease as shown with subjects 8- 10. We also placed the ACM on patients with known heart conditions, specifically AS, that can cause paradoxical splitting in S2. AS is known as the narrowing of the aorta, thus increasing the impedance and hangout interval of the aorta to then cause a delay in the incisura of the blood pressure curve as heart sounds which is also a result of an increase in the pressure of left ventricle and causes the S2 paradoxical splitting to occur. As shown in Table I, multiple instances of paradoxical splitting have been detected by the ACM within one minute of data collection from varying patients with diagnosis of AS and other relating heart diseases.

[0092] FIGs. 7A-B show the plots used for determining S2 splitting. The top graph illustrates the filtered SCG data, the middle graph shows the instantaneous frequency and intensity of instantaneous amplitude, and the bottom graph is the product of instantaneous frequency and instantaneous amplitude which shows the distinct A2 and P2 peaks. As shown in FIG. 7A, normal S2 splitting from subject 7 with a BMI of 37.9 has been detected with splitting of 39.28 ms with distinct differences between A2 and P2 along with a pronounced P2 peak of above 20% of the amplitude of A2. The peaks in the graph are known to be A2 and P2 as the P2 peak has a smaller amplitude than A2 in the graded plot and the S2 time splits are in the range of S2 time split extracted from. Subjects of older age have also been found to have majority normal splitting as shown in FIG. 7B with subject 9 with a time split of 24.36 ms along with subjects 8 and 10. In the subjects with no known abnormal heart conditions, there were a total of four instances of paradoxical splitting found from the SCG data of the ACM. Three instances of paradoxical splitting occurred in subject 8, mainly due to the location of that subject’s murmurs at the start of diastole, thus increasing the amplitude of P2. Like other SCG measurements, most of the frequency contents for A2 and P2 are between 5 and 80 Hz.

[0093] One example of paradoxical splitting is shown in FIG. 8A from subject 11 who has mild AS. There have been ten occurrences of paradoxical splitting within 1 min of data collection. The graded intensity curve in FIG. 8A shows a clear distinction of P2 and A2 peaks with a time split of-14.63 ms. Moderate AS subjects have also been tested to have paradoxical splitting with occurrences ranging from 10 to 19 times in 1 min of data captured as shown from subjects 12- 15. One example of a moderate AS subject, subject 14, has a total of 14 paradoxical S2 splits in 1 min with one of the splits being -24.97 ms as shown in FIG. 8B. Subject 16, a subject with severe AS, had their SCG signal captured and determined to have a total of 23 paradoxical splitting with one example of an S2 time split being - 19.63 ms as shown in FIG. 8C. Even though, the subject with severe AS was shown to have strong murmurs near the end of systole that could potentially affect S2, as shown in FIG. 5B, the A2 and P2 peaks still have higher instantaneous amplitude and show the presence of two major peaks of instantaneous amplitude, S2 time split can still be captured as shown with the clear distinct A2 and P2 peaks above murmurs, as shown in FIG. 8C.

[0094] As reflected from Table I, there is an increase in the number of S2 paradoxical splitting in a set time with an increase in severity of the AS. The valvular heart disease severity has been determined through the subjects’ echocardiograms parameters determined by cardiologists through factors such as aortic valve area, mean pressure gradient, and peak velocity as shown in Table II. In contrast to the past literature, not only subjects diagnosed with severe AS had paradoxical splitting, but also subjects with mild to moderate AS had paradoxical splitting detected and measured via the ACM, as shown in FIG. 8A and 8B. More subject data can be tested to determine if severity is a direct correlation with the number of paradoxical splitting in a set time frame. In addition, the overall S2 time split range has not followed a correlation with heart disease severity as there are many other factors that can determine S2 time split such as other heart diseases.

TABLE II

RECOMMENDATIONS FOR GRADING AORTIC STENOSIS

SEVERITY VIA ECHOCARDIOGRAMS _

Mild Moderate Severe

AS AS AS

Peak velocity (m/s) 2.6-2.9 3.0-4.0 >4.0

Mean gradient <20 20-40 >40

(mmHg)

Aortic valve area >1.5 1.0-1.5 <1.0

(cm²)

Indexed aortic >0.85 0.60-0.85 <0. valve area (cm² /m²)

Velocity ratio >.50 0.25-0.50 <0.25

[0095] Paradoxical splitting has also been found in subjects with heart disease that relate with AS. Subject 17 was diagnosed with moderate aortic regurgitation as their main heart diagnoses and has 11 counts of paradoxical splitting with one S2 time split shown in FIG. 9A at a time split of -31.62 ms. Subjects with aortic regurgitation may also have paradoxical splitting because AS can cause aortic regurgitation. Mitral regurgitation can also be present in the setting of AS. Subject 18, who had a diagnosis of severe mitral regurgitation, has also been found with paradoxical splitting where in one instance the S2 time split was -14.27 ms, as shown in FIG. 9B. Further studies can be conducted to determine if these subjects’ valvular heart disease diagnosis can cause paradoxical splitting or if they have a related cardiac abnormality that is known to cause paradoxical splitting.

[0096] Validation of S2 Time Split: SCG Versus PCG

[0097] The systolic, diastolic, and interbeat interval of the simultaneous recordings of the ACM’s SCG signal and digital stethoscope’s PCG signal were calculated from 60 s of data collection, thus resulting to a total of 68 heartbeats, and showed strong similarity as shown in the correlation plots of FIGs. 10A-C. [0098] The systolic interval difference is calculated by the time difference between S 1 and S2 peaks. The systolic time interval correlation between SCG and PCG signals was shown to be significant with an R² value of 0.9848 (FIG. 10A). Similar correlations were shown with diastolic time difference, as measured by the time difference between S2 and SI peaks with R2 of 0.9927 (FIG. 10B). The overall interbeat interval of SCG and PCG also had strong correlation with R² value 0.9935 (FIG. 10C). These strong correlations show the similarity of the highly sensitive ACM’s SCG signal compared to the digital stethoscope’s PCG signal, and similar correlations have been reported in the past literature from off-the-shelf accelerometers placed on the sternum. To provide qualitative analysis of the SCG S2 waveform, SCG S2 signals and PCG S2 signals were averaged and visually compared. From the 90 s of data captured from the ACM and digital stethoscope, S2 plots from subject 5 from each sensor were superimposed on each other and averaged with the black line representing the averaged S2 signal, as shown in FIGs. 11A-B. There are slight differences with the overall waveform of the S2 signals from SCG and PCG. However, as previously mentioned, the sound peaks are based on their instantaneous frequency and instantaneous amplitude; therefore, the robustness of the SPWVD method shows the capturing of pulmonic sound from the ACM SCG signals. Similar results were shown with the comparison of SCG and PCG S2 signals from other three healthy subjects. In addition, these sounds and peaks are caused from the same source of sound as past studies have found that SCG signals to strongly correlate to echocardiogram readings, especially with S2 sound and total systolic time.

[0099] In addition to showing the correlation and similarity of SCG and PCG S2 time split, S2 time split ranges were calculated via the SPWVD method. S2 time split range of subject 5 from the ACM SCG ranged from 9.97 to 61.42 ms while the S2 time split range calculated from the digital stethoscope PCG ranged from 9.57 to 61.25 ms. The other three healthy subjects tested had their S2 time split range be a maximum difference of ±1.84 ms between SCG and PCG. Because of the SCG and PCG similarity of S2 waveforms, similarity of S2 time split, and foundations of S2 sound with the robustness of SPWVD extraction of A2 and P2, SCG signals from ACM are shown to have similar fidelity to that of PCG signals from digital stethoscope for capturing S2 time split.

[00100] Placement Sensitivity

[00101] To extract S2 time split, A2 and P2 can be captured and identified. The P2 sound might not always be present, or it may be below the 20% amplitude of A2. This might be due to the placement of the ACM as the pulmonic region is relatively small and people’s heart location varies. Therefore, a variable to consider when testing S2 time split on a subject is the sensor placement. To compare placement sensitivity, calculations through SPWVD was calculated and shown.

[00102] As shown in FIGs. 12A-B, a comparison of two different P2 values intensity was recorded at the start of inspiration and when paradoxical splitting occurred for subject 2. The graded intensity curve of the primary location is shown in FIG. 12A, and S2 split captured one inch to the left of the pulmonic region, noted as the secondary location, has the graded intensity curve shown in FIG. 12B. Primary location’s P2 value has a much stronger intensity than the pulmonic sound peak of the secondary location. Not only that, but the hue of the SPWVD of the primary is also more intense. Therefore, the placement of ACM is important to capture the pulmonic sound and to be able to capture and measure S2 splits accurately as one inch difference in placement location can cause for the pulmonic sound to be captured more effectively due to patient variability and due to some patient’s relatively small pulmonic region and variability. Past literatures have shown similar findings of the importance and signal variations caused by the placement of accelerometers for SCG signals.

[00103] Using a sensitive, wearable ACM, the S2 time splits from SCG signals, along with the paradoxical S2 splits, were accurately identified by computation extraction via the SPWVD method. Normal S2 splitting in healthy subjects, also from healthy subjects of higher BMI and of similar age of subjects with valvular heart disease, has been extracted. Paradoxical S2 splitting in unhealthy subjects have been extracted and identified, specifically in subjects with AS from mild to severe severity. ACM’s SCG signals were compared and determined to have similar fidelity of digital stethoscope’s PCG signals. In addition, both SCG and PCG signals from the same subject have extracted and captured similar S2 time split range. ACM can be used for longitudinal and at-home use to detect paradoxical splitting of S2, which can be used to assess severity and monitor the progression of AS and can be used as a posttreatment monitoring device to determine if paradoxical S2 splitting is present during a subject’s recovery phase.

[00104] It is to be understood that the embodiments and claims disclosed herein are not limited in their application to the details of construction and arrangement of the components set forth in the description and illustrated in the drawings. Rather, the description and the drawings provide examples of the embodiments envisioned. The embodiments and claims disclosed herein are further capable of other embodiments and of being practiced and carried out in various ways. Also, it is to be understood that the phraseology and terminology employed herein are for the purposes of description and should not be regarded as limiting the claims.

[00105] Accordingly, those skilled in the art will appreciate that the conception upon which the application and claims are based may be readily utilized as a basis for the design of other structures, methods, and systems for carrying out the several purposes of the embodiments and claims presented in this application. It is important, therefore, that the claims be regarded as including such equivalent constructions.

[00106] Furthermore, the purpose of the foregoing Abstract is to enable the United States Patent and Trademark Office and the public generally, and especially including the practitioners in the art who are not familiar with patent and legal terms or phraseology, to determine quickly from a cursory inspection the nature and essence of the technical disclosure of the application. The Abstract is neither intended to define the claims of the application, nor is it intended to be limiting to the scope of the claims in any way.

Claims

CLAIMS What is claimed is:

1. A system for monitoring a status of a heart, comprising: a wearable sensor configured to generate a signal indicative of vibrations from a heart of a user; a controller configured to extract S2 sound time splits of the heart based, at least in part, on the signal.

2. The system of claim 1, wherein the sensor comprises an accelerometer.

3. The system of claim 1, wherein the signal is a seismocardiogram signal.

4. The system of claim 1, wherein the controller is configured to extract S2 sound time splits of the heart, at least in part, by determining S2 time windows in the signal.

5. The system of claim 1, wherein the controller is configured to extract paradoxical splitting in S2 heart sound.

6. The system of claim 1, wherein the controller is configured to identify S2 time windows in the signal, at least in part, by computing a Shannon energy of the signal.

7. The system of claim 6, wherein the controller is further configured to identify A2 and P2 sounds in the S2 time windows.

8. The system of claim 7, wherein the controller is further configured to identify A2 and P2 sounds in the S2 time windows, at least in part, by generating a time-frequency representation of the signal.

9. The system of claim 8, wherein the time-frequency representation of the signal is a smoothed pseudo Wigner-Ville distribution.

10. The system of claim 1, wherein the controller is further configured to predict a status of a heart disease based, at least in part, on the S2 sound time splits.

11. The system of claim 10, wherein the heart disease is aortic stenosis.

12. The system of claim 10, wherein the heart disease is a valvular heart disease.

13. A method of monitoring a status of a heart, the method comprising: receiving a signal from a wearable sensor, the signal indicative of vibrations from a heart of a user; extracting S2 sound time splits of the heart based, at least in part, on the signal.

14. The method of claim 13, wherein the sensor comprises an accelerometer.

15. The method of claim 13, wherein the signal is a seismocardiogram signal.

16. The method of claim 13, wherein extracting S2 sound time splits comprises determining S2 time windows in the signal.

17. The method of claim 13, wherein extracting S2 sound time splits enables the detection of paradoxical splitting in S2 heart sound.

18. The method of claim 16, wherein determining S2 time windows in the signal comprises computing a Shannon energy of the signal.

19. The method of claim 16, wherein extracting S2 sound time splits in the signal further comprises identifying A2 and P2 sounds in the S2 time windows.

20. The method of claim 16, wherein identifying A2 and P2 sounds in the S2 sounds comprises generating a time-frequency representation of the signal.

21. The method of claim 20, wherein the time-frequency representation of the signal is a smoothed pseudo Wigner-Ville distribution.

22. The method of claim 13, further comprising predicting a status of a heart disease based, at least in part, on the S2 time splits.

23. The method of claim 22, wherein the heart disease is aortic stenosis.