SYSTEM, APPARATUS AND METHOD FOR UTILIZING BIO-POTENTIAL DATA OBTAINED BY MULTIPLE BIOPOTENTIAL SENSORS FOR PRENATAL TRACKING - Patent application

Abstract

The method includes receiving an input dataset for a pregnant woman within a time window, the input dataset including fetal heart rate data; completing missing data samples in the input dataset to provide a completed dataset; removing segments from the completed dataset based on at least one criterion to provide a corrected dataset; detecting at least one event occurring within the time window based on at least one of the corrected dataset or a baseline fetal heart rate; classifying the event as either an acceleration, deceleration, or baseline change based on at least one of the corrected dataset or the baseline fetal heart rate; and generating an output indicative of the event to enable action to be taken with respect to the pregnant woman.

Description

[Technical field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application is an international (PCT) patent application related to and claiming the benefit of commonly owned, co-pending U.S. Provisional Patent Application No. 63 / 304,120, filed on January 28, 2022, and entitled “SYSTEM, DEVICE, AND METHOD FOR PRENATAL CLINICAL DECISION SUPPORT.”

[0002] Technical Field The present disclosure relates to systems and methods for providing clinical decision support to clinicians providing care to pregnant women. In particular, the present disclosure relates to systems and methods for providing clinical decision support based on the results of remote non-stress tests performed through the use of sensor-equipped garments (e.g., garments such as belts having sensors such as biopotential and acoustic sensors). [Background technology]

[0003] A non-stress test involves the review of fetal heart rate, maternal heart rate and uterine activity data for 20-40 minutes. This data is reviewed manually by a physician or experienced nurse to determine if the test is problematic or reassuring. The clinician looking at the data manually scans the data and looks for relevant clinical phenomena, which are usually changes in fetal heart rate. For example, these phenomena include fetal heart rate acceleration, deceleration, or a certain level of fetal heart rate variability. The clinical relevance of some of these phenomena may vary based on their temporal relationship to the appearance of uterine activity contractions. Once these clinical phenomena are found, they are marked and a clinical decision is made as to whether the NST is reactive or non-responsive according to a set of rules defined by the American College of Obstetricians and Gynecologists (ACOG). Based on this decision, the next steps in the clinical workflow are devised. Summary of the Invention

[0004] In some embodiments, a method includes receiving, by a computing device, an input dataset related to a pregnant woman within a time window, the input dataset including at least fetal heart rate data, the fetal heart rate data calculated based at least on raw biopotential data obtained by a plurality of biopotential sensors in a garment worn by the pregnant woman, the method includes completing, by the computing device, at least one missing data sample in the input dataset to provide a completed dataset; removing, by the computing device, at least one segment from the completed dataset based on at least one criterion to provide a corrected dataset; calculating, by the computing device, a baseline fetal heart rate based at least on the corrected dataset; detecting, by the computing device, at least one event occurring within the time window based on at least one of the corrected dataset or the baseline fetal heart rate; classifying, by the computing device, the at least one event as either an acceleration, a deceleration, or a baseline change based on at least one of the corrected dataset or the baseline fetal heart rate; and generating, by the computing device, an output indicative of the at least one event, the output configured to enable at least one action to be taken with respect to the pregnant woman.

[0005] In some embodiments, the at least one criterion for removing at least one segment from the completed data set includes removing the at least one segment based on at least one segment including a periodic change in fetal heart rate, in some embodiments, the periodic change in fetal heart rate includes at least one of (a) fetal heart rate changes of more than 25 beats per minute in consecutive samples, or (b) applying a threshold based on a median and a standard deviation.

[0006] In some embodiments, the method further includes determining, by the computing device, whether the input data set is valid or not valid, and if the input data set is determined to be not valid, receiving, by the computing device, a further input data set before performing the step of complementing the input data set. The step of complementing the input data set by the computing device is performed on the further input data set. In some embodiments, the step of determining whether the input data set is valid or not is performed by a process including: identifying, by the computing device, a plurality of stable segments of the corrected data set; determining, by the computing device, a total duration of the plurality of stable segments; and determining, by the computing device, whether the input data set is valid or not based at least on a comparison of the total duration of the plurality of stable segments to a predetermined threshold duration. Each stable segment includes a sequence of a predetermined amount of consecutive samples having a fetal heart rate variability less than a predetermined threshold variability.

[0007] In some embodiments, determining, by the computing device, the baseline fetal heart rate is performed by a process including: calculating, by the computing device, a virtual baseline fetal heart rate; calculating, by the computing device, a virtual high fetal heart rate limit; calculating, by the computing device, a virtual low fetal heart rate limit; removing, by the computing device, an outlier data point from the corrected data set; and calculating, by the computing device, a baseline fetal heart rate based at least on the application of rounding to the corrected data set after removing the outlier data point. An outlier data point is a data point having a fetal heart rate higher than the virtual high fetal heart rate limit or a fetal heart rate lower than the virtual low fetal heart rate limit. In some embodiments, the virtual baseline fetal heart rate is an average value of the fetal heart rates in the corrected data set. In some embodiments, the steps of calculating the virtual high fetal heart rate limit and calculating the virtual low fetal heart rate limit are performed by a process that includes: calculating, by a computing device, a standard deviation of a virtual baseline fetal heart rate; calculating, by a computing device, a virtual high fetal heart rate limit as the virtual baseline fetal heart rate plus two times the standard deviation; and calculating, by a computing device, a virtual low fetal heart rate limit as the virtual baseline fetal heart rate minus two times the standard deviation.

[0008] In some embodiments, at least one of the at least one event is classified as an acceleration, and at least one of the at least one event is classified as an acceleration based on a process including: identifying, by a computing device, a first intersection between a trace of the corrected data set and a baseline fetal heart rate; identifying, by a computing device, a second intersection between the trace of the corrected data set and the baseline fetal heart rate; identifying, by a computing device, a maximum fetal heart rate within a time interval between the first intersection and the second intersection; and identifying, by a computing device, the time interval as a value including an acceleration based on at least a duration of the time interval and the maximum fetal heart rate within the time interval. In some embodiments, identifying the time interval as a value including an acceleration is further based on a gestational age of the pregnant woman. In some embodiments, the method further includes determining, by the computing device, a fetal heart rate baseline variability. In some embodiments, determining the fetal heart rate baseline variability by the computing device is performed by a process that includes determining, by the computing device, a highest fetal heart rate within a variability time interval following the second intersection, determining, by the computing device, a lowest fetal heart rate within the variability time interval, and subtracting, by the computing device, the lowest fetal heart rate within the variability time interval from the highest fetal heart rate within the variability time interval to provide the fetal heart rate baseline variability. The variability time interval extends from the second intersection until the earlier of (1) a predetermined duration following the second intersection, or (2) the onset of a subsequent acceleration or deceleration.

[0009] In some embodiments, the at least one event includes a deceleration, and the at least one event is classified as a deceleration based on a process including: identifying, by a computing device, a first intersection between a trace of the corrected data set and a baseline fetal heart rate; identifying, by a computing device, a second intersection between the trace of the corrected data set and the baseline fetal heart rate; identifying, by a computing device, a minimum fetal heart rate within a time interval between the first intersection and the second intersection; and identifying, by a computing device, the time interval as a value including a deceleration based on at least a difference between the baseline fetal heart rate and the minimum fetal heart rate within the time interval. In some embodiments, the method further includes classifying, by the computing device, the deceleration as either a fast deceleration, a slow deceleration, a variable deceleration, or a long deceleration. In some embodiments, the input data set further includes a uterine contraction signal. The deceleration is classified as one of a fast deceleration, a slow deceleration, a variable deceleration, or a long deceleration based on a process that includes: determining, by a computing device, a delay, which is an amount of time elapsed between the first intersection and a time of a minimum fetal heart rate within the time interval; determining, by a computing device, a time of a peak of a uterine contraction within the time interval; determining, by a computing device, a time difference between the time of the minimum fetal heart rate within the time interval and a time of a peak of the uterine contraction signal within the time interval; and classifying, by the computing device, the deceleration as one of a fast deceleration, a slow deceleration, a variable deceleration, or a long deceleration based on at least the delay, the time difference, and the duration of the uterine contraction.

[0010] In some embodiments, two iterations of the following steps are performed for the time window: (1) removing, by the computing device, at least one segment from the completed dataset based on at least one criterion to provide a corrected dataset; (2) calculating, by the computing device, a baseline fetal heart rate based at least on the corrected dataset; and (3) detecting, by the computing device, at least one event that occurred within the time window based on at least one of the corrected dataset or the baseline fetal heart rate.

[0011] In some embodiments, the method includes repeating, for a subsequent time window, the steps of receiving, by the computing device, an input dataset of the pregnant woman, completing, by the computing device, at least one missing data sample in the input dataset to provide a completed dataset, removing, by the computing device, at least one segment from the completed dataset based on at least one criterion to provide a corrected dataset, calculating, by the computing device, a baseline fetal heart rate based at least on the corrected dataset, detecting, by the computing device, at least one event that occurred within the time window based on at least one of the corrected dataset or the baseline fetal heart rate, and classifying, by the computing device, the at least one event as either an acceleration, a deceleration, or a baseline change based on at least one of the corrected dataset or the baseline fetal heart rate. In some embodiments, the subsequent time window does not overlap with the time window. In some embodiments, the subsequent time window overlaps with the time window by an overlap duration. In some embodiments, the overlap duration is between 5% and 95% of the duration of the time window. [Brief description of the drawings]

[0012] Certain embodiments of the invention are herein described, by way of example only, with reference to the accompanying drawings. Referring now in detail and specifically to the drawings, it is emphasized that the particulars shown are by way of example and for illustrative discussion of embodiments of the invention. In this regard, the description given in conjunction with the drawings will make apparent to those skilled in the art how embodiments of the invention may be practiced.

[0013] [Figure 1] 1 shows a flow chart of an exemplary clinical decision support process. [Diagram 2] 1 illustrates a flowchart of an exemplary process for identifying the next valid fetal heart rate (FHR) data segment. [Diagram 3] 1 shows a flowchart of an exemplary process for estimating an FHR baseline. [Figure 4] 1 shows a flowchart of an exemplary process for identifying FHR acceleration. [Diagram 5] 1 shows a flowchart of an exemplary process for extracting FHR baseline variability. [Figure 6A] 1 shows a flowchart of an exemplary process for identifying FHR decelerations. [Figure 6B] 1 shows an exemplary chart of fetal heart rate and uterine contractions. [Figure 7] 1 shows a flowchart of an exemplary process for classifying FHR decelerations. [Figure 8] 1 shows a flow chart of exemplary criteria for classifying monitoring sessions based on FHR. [Figure 9A] 1 illustrates an example clinical decision support system output for a first patient. [Figure 9B] 13 illustrates an example clinical decision support system output for a second patient. [Figure 9C] 1 shows an example clinical decision support system output for a first patient, with clinical phenomena annotated. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] Although various detailed embodiments of the present disclosure are disclosed herein in conjunction with the accompanying figures, it should be understood that the disclosed embodiments are merely exemplary. Moreover, the examples given in connection with the various embodiments of the present disclosure are intended to be illustrative and not limiting.

[0015] Throughout this specification, the following terms take on the meanings expressly associated therewith, unless the context clearly indicates otherwise. The phrases "in one embodiment" and "in some embodiments" as used herein do not necessarily refer to the same embodiment. Furthermore, the phrases "in another embodiment" and "in some other embodiments" as used herein do not necessarily refer to different embodiments, although they may. Thus, as described below, various embodiments can be readily combined without departing from the scope or spirit of the present disclosure.

[0016] Moreover, the term "based on" is not exclusive and allows for based on additional unrecited factors unless the context clearly dictates otherwise. Also, throughout this specification, the meanings of a, an, and the include plural references. The meaning of "in" includes "in" and "on."

[0017] It is understood that at least one aspect / function of various embodiments described herein may be performed in real-time and / or dynamically. As used herein, the term "real-time" refers to an event / action that may occur instantaneously or nearly instantaneously in time when another event / action occurs. For example, "real-time processing," "real-time computation," and "real-time execution" all relate to the performance of a computation during the actual time that an associated physical process (e.g., a user interacting with an application on a mobile device) occurs, such that the results of the computation can be used in directing the physical process.

[0018] As used herein, the terms "dynamically" and "automatically," as well as logical and / or linguistic related and / or derivatives thereof, mean that certain events and / or actions can be triggered and / or occur without human intervention. In some embodiments, events and / or actions according to the present disclosure can occur in real-time and / or based on at least one predetermined periodicity of nanoseconds, nanoseconds, milliseconds, milliseconds, seconds, seconds, minutes, minutes, hourly, hours, daily, daily, weekly, monthly, etc.

[0019] Exemplary embodiments relate to systems and methods configured to receive data describing a condition of a pregnant woman, such as maternal heart rate data, fetal heart rate data, and uterine activity (e.g., contraction) data, perform remote non-stress testing on such data, and provide clinical decision support to a clinician based on such non-stress testing. Some embodiments of the present disclosure may be understood in part by reference to the following description and accompanying drawings. In some embodiments, the fetal heart rate is determined based on data obtained through the use of sensors (e.g., biopotential sensors, acoustic sensors) integrated into at least one garment (e.g., belt) worn by the pregnant woman, according to at least one of the techniques described in U.S. Patent Application Publication No. 2021 / 0267535. In some embodiments, the uterine activity is determined based on data obtained through the use of sensors (e.g., biopotential sensors, acoustic sensors) integrated into at least one garment (e.g., belt) worn by the pregnant woman, according to at least one of the techniques described in U.S. Patent Application Publication No. 2021 / 0267535.

[0020] In some embodiments, the analysis is performed in two stages. In some embodiments, in a first stage, basic clinical phenomena (e.g., heart rate acceleration or deceleration) are calculated with a certain degree of confidence and marked on top of the fetal heart rate ("FHR") tracing. In some embodiments, in a second stage, these basic phenomena are used to determine meaningful clinical decisions that are presented to the physician.

[0021] In some embodiments, the first stage includes capturing basic clinical phenomena on the FHR trace using analytical techniques including digital signal processing ("DSP") and machine learning ("ML"), and presenting such clinical phenomena to a clinician. In some embodiments, the first stage includes a first sub-stage of capturing events and outputting features of such events, and a second sub-stage of filtering and presenting such events.

[0022] In some embodiments, events are captured utilizing an iterative process to estimate FHR baseline, identify FHR baseline variability, capture FHR accelerations and decelerations, and classify such events into subtypes. FIG. 1 shows a flow chart of an exemplary iterative method 100 (e.g., iterative process). In step 110, a time window of relevant data (e.g., FHR data, MHR data, and contraction data) is received. In some embodiments, the data received in step 110 is referred to as the "input data set." In some embodiments, the window is 10 minutes long. In some embodiments, when the method begins, the time window is a first time window, and thereafter, the windows are subsequent time windows. In some embodiments, each time window does not overlap with adjacent time windows (e.g., the immediately preceding and following time windows). In some embodiments, each time window overlaps with adjacent time windows.In some embodiments, adjacent time windows overlap each other by 5% of the duration of each time window (e.g., each time window has a duration of 10 minutes and the amount of overlap is half a minute), or 10% of the duration of each time window, or 15% of the duration of each time window, or 20% of the duration of each time window, or 25% of the duration of each time window, or 30% of the duration of each time window, or 35% of the duration of each time window, or 40% of the duration of each time window, or 45% of the duration of each time window, or 50% of the duration of each time window, or or 55% of the duration of each time window, or 60% of the duration of each time window, or 65% of the duration of each time window, or 70% of the duration of each time window, or 75% of the duration of each time window, or 80% of the duration of each time window, or 85% of the duration of each time window, or 90% of the duration of each time window, or 95% of the duration of each time window, or a portion of each time window between 5% and 95% of the duration of each time window, or any other amount between an overlap of one data point in duration to an overlap of the entire time window but for one non-overlapping data point.

[0023] In step 120, it is determined whether the time window contains a valid data segment. In some embodiments, the validity determination is performed according to method 200, described below. If a valid data segment is not found, the method returns to step 110 and the next window of data is received. If a valid data segment is found, the method proceeds to step 130.

[0024] In step 130, missing data samples are input into the input data set (e.g., the input data set is completed). In some embodiments, the missing samples are filled by interpolation. In some embodiments, the interpolation is performed using a bidirectional autoregressive interpolation model. In some embodiments, the result of the completion of the input data set in step 130 is referred to as the "completed data set." After step 130, the method proceeds to step 140. In step 140, certain types of data are removed from the completed data set based on at least one criterion. In some embodiments, the data removal criterion in step 140 includes removing data that represents cyclic or transient changes in FHR. In some embodiments, the data removal criterion in step 140 includes removing data that represents periods of significant FHR variability. In some embodiments, the data removal criterion in step 140 includes removing segments of baseline FHR that differ by more than 25 beats per minute. In some embodiments, the data removal criterion in step 140 includes removing cyclic or transient changes by performing a moving average smoothing process and then applying a median ±2 standard deviation threshold. In some embodiments, significant variability and segments of baseline FHR that differ by more than 25 beats per minute are removed by identifying and removing segments of baseline FHR that have consecutive samples that differ by more than 25 beats per minute. In some embodiments, the result of data removal in step 140 is referred to as a "corrected data set."

[0025] After step 140, in step 150, a baseline FHR is determined (e.g., calculated). In some embodiments, following the removal of step 140, a baseline is extracted from the original signal that does not include the removed segments (e.g., from the corrected data set resulting from step 140). In some embodiments, the baseline FHR is determined according to method 300 described below. In some embodiments, the baseline FHR is determined as a computational baseline. In some embodiments, the baseline FHR is determined using the ACOG guidelines as the average of valid samples within a predefined time window. In step 160, accelerations and decelerations are identified. In some embodiments, accelerations are identified according to method 400 described below. In some embodiments, decelerations are identified according to method 600 described below.

[0026] In step 170, it is determined whether steps 140, 150, and 160 have been performed for the first or second iteration of the current data. If the first iteration is complete, the method returns to step 140 and steps 140, 150, 160 are repeated as described above. If the second iteration is complete, the method proceeds to step 180. In step 180, the events identified in step 160 are classified as described herein (e.g., as a particular type of acceleration, a particular type of deceleration, or a change in baseline). After step 180, in step 190, FHR variability is calculated. In some embodiments, the variability is calculated according to method 500, described below. After step 190, the method returns to step 110 and the next window of data is received.

[0027] In some embodiments, the iterative process illustrated in FIG. 1 includes identifying the next valid segment of the FHR trace. FIG. 2 illustrates a flow chart of an exemplary method 200 for identifying the next valid segment. In step 210, an FHR session is received as input. In step 215, a first stable segment of the FHR session is detected. In some embodiments, a stable segment is a sequence of five adjacent samples (e.g., a predetermined amount of consecutive samples) with an FHR difference of less than 25 bpm. In step 220, the method advances the FHR session to the next segment. In step 225, the method determines whether the next segment includes five adjacent samples with an FHR difference greater than 25 bpm. If the segment does not include five adjacent samples with an FHR difference greater than 25 bpm, the method returns to step 220. If the segment includes five adjacent samples with an FHR difference greater than 25 bpm, the method proceeds to step 230.

[0028] In step 230, the method advances the session to the next stable segment. In some embodiments, similar to step 215, a stable segment is a sequence of five adjacent samples with an FHR difference of less than 25 bpm. In step 235, the length of the unstable segment (e.g., the length of time between the stable segment identified in step 230 and the previous stable segment) is evaluated. If the length is less than a threshold length, an interpolation is performed to generate substitute FHR data for the unstable segment. In some embodiments, the threshold length is 2.5 seconds. In some embodiments, the interpolation is a spline interpolation. In some embodiments, the interpolation is a second order polynomial interpolation. Additionally, the number of samples of the unstable segment is counted. If the length of the unstable segment is greater than the threshold length, no interpolation is performed and the number of samples of the unstable segment is counted. In some embodiments, if the interpolation is performed, the method proceeds to step 240. In step 240, a smoothing process is performed at the edges of the interpolated segment.

[0029] After step 235, if the entire session has not yet been evaluated, the method returns to step 215 to continue moving through the samples within the session to evaluate further segments of the session. If the entire session has been evaluated, the method proceeds to step 245.

[0030] In parallel with step 245, where the total number of interpolation points is counted as a trace quality criterion, in step 250, the total number of samples in the session with a BPM below a minimum valid BPM threshold M is counted. In some embodiments, the minimum valid BPM threshold M is 50 bpm. In some embodiments, samples with a BPM below the minimum valid BPM threshold M are converted to invalid samples. In step 255, the samples are evaluated on a pass / fail basis based on certain thresholds for certain parameters. In some embodiments, the parameters include a percentage of signal loss. In some embodiments, the threshold for the percentage of signal loss is 40%. In some embodiments, the parameters include a confidence level of the interpolation points. In some embodiments, the confidence level of the interpolation points is ±1 bpm. In some embodiments, the parameters include a minimum of stable time per sample. In some embodiments, the minimum number of stable minutes is 2 minutes per 10 minute sample. In some embodiments, a single stable segment of minimum length is sufficient.

[0031] In some embodiments, as a result of the above process and in accordance with the European College of Obstetrics and Gynaecology (ECOG) guidelines, there must be two stable periods within a 10 minute sample to define an FHR baseline. In some embodiments, if there are no indeterminate samples, the baseline is indeterminate for the current sample. In some embodiments, if there are indeterminate samples, the previous frame is referenced to determine the baseline FHR.

[0032] In some embodiments, the iterative process illustrated in FIG. 1 includes estimating a baseline FHR for a given segment of the FHR tracing. FIG. 3 shows a flowchart of an exemplary method 300 for estimating a baseline FHR. In step 310, a session of FHR data is received. In step 320, a virtual baseline R(n) is calculated. In some embodiments, the virtual baseline R(n) is calculated (e.g., as the arithmetic mean of all FHR data points) according to the following formula: TIFF2025504014000002.tif1361

[0033] In some embodiments, the virtual baseline R(n) is calculated over a 10 minute time window containing 2400 FHR samples. In some embodiments, the virtual baseline R(n) is calculated using a moving median window of 2 minute length with 1 minute overlap.

[0034] In step 330, a virtual upper FHR limit and a virtual lower FHR limit are calculated. In some embodiments, the virtual upper FHR limit and the virtual lower FHR limit are calculated using the parameter α. In some embodiments, α is equal to two times the standard deviation of the virtual baseline R(n). In some embodiments, the virtual upper FHR limit is calculated as R(n)+α and the virtual lower FHR limit is calculated as R(n)-α. In step 340, outlier data points having an FHR higher than the virtual upper FHR limit or lower than the virtual lower FHR limit are removed from the data set.

[0035] In step 350, an ecog baseline is calculated. In some embodiments, the ecog baseline is a 5 bpm rounded baseline of the calculated baseline. In some embodiments, the baseline is the average FHR remaining in the sample after the outlier data points are removed in step 340. In step 360, the first and last segments are extrapolated to upsample the virtual baseline and remove it from the FHR trace. The output of step 360 is used as input for acceleration / deceleration extraction. In step 370, the method proceeds to the next segment. In some embodiments, each subsequent segment overlaps with the previous segment by 2 minutes.

[0036] In some embodiments, the iterative process depicted in FIG. 1 includes extracting (e.g., identifying) FHR accelerations and identifying long-term accelerations and baseline changes based on the identified FHR accelerations. In some embodiments, an FHR acceleration is a visually apparent sudden increase in FHR. In some embodiments, an increase is a sudden increase if less than 30 seconds elapse between the onset of the increase and the peak FHR. In some embodiments, from 32 weeks gestation onwards, the acceleration has a peak of 15 bpm or more above the baseline FHR and a duration from onset to return to baseline between 15 seconds and 2 minutes. In some embodiments, before 32 weeks gestation, the acceleration has a peak of 10 bpm or more above the baseline FHR and a duration from onset to return to baseline between 10 seconds and 2 minutes. In some embodiments, the long-term acceleration has a duration between 2 and 10 minutes. In some embodiments, if the acceleration lasts for 10 minutes or more, there is considered to be a change in the baseline FHR.

[0037] 4 shows a flow chart of an exemplary method 400 for extracting FHR acceleration. In step 410, the method receives an FHR trace as input. In step 420, a first intersection point X1 between the baseline FHR and the FHR trace is identified. In step 430, a second intersection point X2 between the baseline FHR and the FHR trace is identified. In step 440, a peak FHR Ymax of the FHR trace between X1 and X2 is determined.

[0038] In step 450, it is determined whether the period between X1 and X2 is considered to be an acceleration. In some embodiments, for pregnant women less than 32 weeks pregnant, the period is considered to be an acceleration if (a) the time elapsed between X1 and X2 is 10 seconds or more, (b) the increase in FHR from baseline FHR to peak FHR Ymax is 10 bpm or more, and (c) the time elapsed between the time of potential onset X1 and peak FHR Ymax is 30 seconds or less. In some embodiments, for pregnant women 32 weeks pregnant or more, the period is considered to be an acceleration if (a) the time elapsed between X1 and X2 is 15 seconds or more, (b) the increase in FHR from baseline FHR to peak FHR Ymax is 15 bpm or more, and (c) the time elapsed between potential onset X1 and peak FHR Ymax is 30 seconds or less. After step 450, if the period between X1 and X2 is not considered to be an acceleration, the method returns to step 410 and the next period is considered. If the period is considered to be an acceleration, the method proceeds to step 460.

[0039] In step 460, an acceleration characteristic is calculated. In some embodiments, an acceleration is considered to be a change in baseline FHR if the elapsed time between X1 and X2 is 10 minutes or more. In some embodiments, an acceleration is considered to be a long-term acceleration if the elapsed time between X1 and X2 is 2 minutes or more but less than 10 minutes. Following step 460, method 400 is completed for the given acceleration and is run again for subsequent potential accelerations.

[0040] In some embodiments, the iterative process shown in FIG. 1 includes extracting FHR baseline variability. FIG. 5 shows a flow chart of an exemplary method 500 for extracting FHR baseline variability. In step 510, an FHR trace is received. In step 520, a maximum FHR value Ymax is calculated for a time period within 2 minutes after point X2 and before any subsequent acceleration or deceleration. In step 530, a minimum FHR value Ymin is calculated for the same time period within 2 minutes after point X2 and before any subsequent acceleration or deceleration. In step 540, the variability is calculated by subtracting Ymax-Ymin. In some embodiments, the variability is calculated on the FHR trace for each 2 minute period on the segment between the identified acceleration and deceleration.

[0041] In some embodiments, the iterative process illustrated in FIG. 1 includes extracting FHR decelerations. FIG. 6A shows a flow chart of an exemplary method 600 for extracting FHR decelerations, and FIG. 6B shows a graph 680 of a fetal heart rate signal 682, a fetal heart rate baseline 684, a uterine activity signal 686, and other points as described below. In step 610, an FHR trace is received as an input. In step 620, a first intersection point X3 between the baseline FHR and the FHR trace is identified. In step 630, a subsequent second intersection point X4 between the baseline FHR and the FHR trace is identified. In step 640, a minimum value Ymin of the FHR trace during the period between X3 and X4 is calculated. In step 650, it is determined whether the period between X3 and X4 is a candidate deceleration. In some embodiments, if the minimum FHR between X3 and X4 is at least 15 bpm lower than the baseline FHR, the period between X3 and X4 is identified as a candidate deceleration. In step 660, deceleration characteristics are calculated. In some embodiments, the features calculated for the deceleration include fall time (e.g., duration from start to nadir), slope from start to nadir, rise time (e.g., duration from nadir to offset), and slope from nadir to end. In step 670, the deceleration is output as a candidate deceleration.

[0042] In some embodiments, the iterative process depicted in FIG. 1 includes classifying the extracted FHR decelerations. For example, in some embodiments, the FHR deceleration may be classified as an early deceleration. In some embodiments, an early deceleration is a visually evident and usually symmetrical decline and return of FHR associated with a uterine contraction. In some embodiments, a decline is defined as having an elapsed time of 30 seconds or more between the start of the deceleration and the nadir of the deceleration (e.g., the minimum FHR during the deceleration). In some embodiments, the magnitude of the decline is measured from the start to the nadir. During an early deceleration, the nadir coincides with the peak of the associated uterine contraction. In most cases, the start, nadir, and return of the deceleration coincide with the start, peak, and end, respectively, of the associated uterine contraction.

[0043] In some embodiments, a deceleration of the FHR may be classified as a slow deceleration. In some embodiments, a slow deceleration is a visually evident and usually symmetrical decline and return of the FHR associated with a uterine contraction. In some embodiments, a decline is defined as having an elapsed time of 30 seconds or more between the onset of the deceleration and the nadir of the deceleration (e.g., the minimum FHR during the deceleration). In some embodiments, the magnitude of the decline is measured from onset to nadir. During a slow deceleration, the nadir occurs after the peak of the associated uterine contraction. In most cases, the onset, nadir, and return of the deceleration occur after the onset, peak, and end, respectively, of the associated uterine contraction.

[0044] In some embodiments, a deceleration in FHR may be classified as a variable deceleration. In some embodiments, a variable deceleration is a visually evident sudden decrease in FHR. In some embodiments, a sudden decrease in FHR is defined as less than 30 seconds elapsed from the onset of deceleration to the nadir of deceleration (e.g., minimum FHR during deceleration). In some embodiments, the magnitude of the decrease is measured from onset to nadir. In some embodiments, a variable deceleration is defined as a decrease in FHR of 15 beats per minute or more and having a duration of 15 seconds or more and 2 minutes or less. When a variable deceleration is associated with a uterine contraction, its onset, depth, and duration typically vary between successive uterine contractions.

[0045] In some embodiments, the decay of the FHR may be classified as a long-term deceleration. In some embodiments, a long-term deceleration is a visually evident decrease in FHR below the baseline FHR. In some embodiments, a long-term deceleration is defined as a decrease in FHR of 15 beats per minute or more and having a duration of 2 minutes or more and 10 minutes or less. In some embodiments, a decrease in FHR lasting 10 minutes or more is considered a change in the baseline FHR.

[0046] FIG. 7 shows a flow chart of an exemplary method 700 for classifying FHR decelerations (e.g., as early, late, variable, or long decelerations, as defined above). In step 710, candidate decelerations (e.g., the output of method 600) are received as input. In step 720, for each candidate deceleration, the minimum FHR value Ymin is identified and a delay Zd is calculated as the elapsed time between the start X3 and the time of the first occurrence of Ymin. In step 730, for each candidate deceleration, a peak in the uterine contraction signal UCmax is identified and correlated with the candidate deceleration. In step 740, for each candidate deceleration, a time difference τ (see FIG. 6B) is calculated as the difference between the time of the nadir of the candidate deceleration and the time of the peak of the correlated uterine contraction.

[0047] In step 750, each candidate deceleration is classified based on the delay Zd and based on a comparison of the time difference τ to the correlating uterine contraction duration. In some embodiments, a candidate deceleration is classified as a slow deceleration if (a) τ is less than or equal to half a uterine contraction plus or minus 5 seconds, (b) τ is greater than zero, and (c) the delay Zd is greater than 30 seconds. In some embodiments, a candidate deceleration is classified as a fast deceleration if (a) τ is less than or equal to half a uterine contraction plus or minus 5 seconds, (b) τ is less than zero, and (c) the delay Zd is greater than 30 seconds. In some embodiments, a candidate deceleration is classified as a long deceleration if (a) τ is greater than half a uterine contraction plus or minus 5 seconds, (b) the total duration of the candidate deceleration is less than or equal to 600 seconds, and (c) the delay Zd is between 15 and 30 seconds. In some embodiments, a candidate deceleration is classified as a variable deceleration if (a) the total duration of the candidate deceleration is less than or equal to 120 seconds, and (b) the delay Zd is between 15 and 30 seconds.

[0048] In some embodiments, if a deceleration having a total duration of more than one minute is identified based on execution of step 750, an extended deceleration epoch alert is generated. In some embodiments, if three adjacent variable decelerations within a 20 minute duration are identified based on execution of step 750, a variable deceleration epoch alert is generated. After step 750, method 700 is complete.

[0049] In some embodiments, after the iterative process shown in FIG. 1 is performed for a given session, the marking service outputs an aggregated list of events. In some embodiments, each event in such aggregated list includes morphological and temporal parameters. In some embodiments, each event includes a description of the event and timing information such as start time, end time, peak time, rise time, and fall time. In some embodiments, each event (e.g., each acceleration and deceleration) is classified with a level of "clinical confidence" from high to low. In some embodiments, the clinical confidence includes an amplitude confidence and a duration confidence. In some embodiments, the amplitude confidence represents the confidence of the event amplitude, defined as the relative distance between the maximum peak of the event and the ACOG criterion for the event (e.g., the peak of the event minus the ACOG criterion divided by the ACOG criterion). In some embodiments, the duration confidence represents the confidence of the duration of the event, defined as the percentage shift between the maximum duration of the event and the corresponding ACOG parameter relative to the ACOG parameter standard (e.g., event duration minus ACOG parameter divided by ACOG parameter). Providing an example with numbers selected for clarity of explanation, for a given event type, the ACOG guideline threshold is 10 bpm and the detected candidate event has a value of 13 bpm. In such an example, the clinical amplitude confidence is equal to ((13-10) / 10)=0.3. As another example, the ACOG guideline may define a threshold of 10 bpm and the detected candidate event may have a value of 16 bpm. In such an example, the clinical amplitude confidence is equal to ((16-10) / 10)=0.6. With this type of confidence, a higher confidence value indicates a greater certainty of the event (e.g., if the ACOG criteria define deceleration as having a threshold of 10 bpm and a detected candidate event is 16 bpm, a confidence of 0.6 indicates a greater certainty that a deceleration occurred than a confidence of 0.3).

[0050] In some embodiments, the user interface allows the clinician to select a threshold confidence level and display only those events that meet the selected threshold. In some embodiments, each event (e.g., each acceleration and each deceleration) is classified with a level of "algorithmic confidence" that represents the confidence of the process described above that the event is present based on internal factors. In some embodiments, the algorithmic confidence of each event is defined as the average of the square root of the distance between the upper and lower confidence intervals (e.g., 95% confidence intervals) of any imputed (e.g., interpolated) samples within a contiguous segment of the event. Providing an example with numbers selected for clarity of explanation, a given interpolated data point may have a predicted value of 144 bpm, with an upper 95% confidence interval of 146 bpm and a lower 95% confidence interval of 142 bpm. In this case, the confidence interval is 4 bpm (e.g., 146 bpm-142 bpm). If a data point is not interpolated, its confidence interval is zero. To generate the algorithmic confidence of an event, the square roots of the confidence intervals of each data point within the event are averaged. For example, if an event has eight data points with confidence intervals of 0,0,0,1,4,1,0,0, the square roots of the eight confidence intervals 0,0,0,1,2,1,0,0 are averaged to (0+0+0+1+2+1+0+0) / 8 to produce an algorithmic confidence of 0.5 for that event. Thus, if there are no imputed samples between the start and end of the event, the algorithmic confidence will be zero. With this type of confidence, a higher confidence indicates more uncertainty about the event, as it includes more interpolated data points and / or interpolated data points with a larger 95% confidence interval.

[0051] In some embodiments, the events are filtered for presentation. In some embodiments, the events are filtered based on a set of predefined parameters for each event type. In some embodiments, the predefined parameters include a value and a confidence level. In some embodiments, events that do not meet the predefined parameters are filtered. In some embodiments, the unfiltered events are provided to the clinician via a user interface, such as marking the events on an FHR trace. In some embodiments, providing the results to the clinician completes the first stage of the analysis.

[0052] In some embodiments, in a second stage of analysis, the clinical phenomenon captured in the first stage is analyzed and based on this analysis, a clinical decision is proposed to the clinician along with known clinical guidelines. In some embodiments, the second stage includes supporting a basic decision. In some embodiments, the basic decision includes classifying the session as responsive or non-responsive. In some embodiments, the session is classified as responsive if at least two accelerations are detected during the first iteration. In some embodiments, the basic decision includes classifying the session as normal, suspicious, or pathological. In some embodiments, the session is classified as normal, suspicious, or pathological based on whether each of the following features are classified as reassuring, non-reassuring, or abnormal: (a) FHR baseline during the session, (b) FHR variability during the session, (c) FHR deceleration during the session, and (d) FHR acceleration during the session. FIG. 8 is a table defining exemplary criteria for defining the above parameters as reassuring, non-reassuring, or abnormal. In some embodiments, the session is classified as normal if all four features fall into the reassuring category. In some embodiments, a session is classified as suspicious if one of the features is classified in the non-reassuring category and the remaining features are classified in the reassuring category. In some embodiments, a session is classified as pathological if two or more features are classified in the non-reassuring category or if one or more features are classified in the abnormal category.

[0053] In some embodiments, assisting in basic decisions also includes providing an alert to the clinician based on a non-comfortable condition, such as tachycardia. In some embodiments, an alert is provided based on the baseline FHR. In some embodiments, an alert is provided indicating tachycardia if the baseline FHR is greater than 160 bpm. In some embodiments, an alert is provided indicating bradycardia if the baseline FHR is less than 110 bpm for at least 10 minutes.

[0054] In some embodiments, an alert is provided based on FHR variability. In some embodiments, an alert indicating no variability is provided if there is a sustained undetectable amplitude in the FHR range of 5 bpm or less. In some embodiments, an alert indicating minimal variability is provided if there is a detectable but amplitude range of 5 bpm or less. In some embodiments, an alert indicating significant variability is provided if the FHR variation is 25 bpm or more. In some embodiments, an alert indicating sinusoidal variability is provided if the FHR baseline exhibits a sinusoidal pattern with a frequency of 3-5 cycles per minute sustained for at least 20 minutes.

[0055] In some embodiments, an alert is provided based on FHR acceleration. In some embodiments, an alert indicating long-term acceleration is provided if FHR acceleration is detected to last between 2 and 10 minutes. In some embodiments, an alert indicating a baseline shift is displayed if FHR acceleration lasting longer than 10 minutes is detected.

[0056] In some embodiments, an alert is provided based on FHR decelerations. In some embodiments, an alert indicating slow decelerations is provided when an FHR deceleration is detected in which the onset, peak, and end of an associated contraction is followed by the onset, nadir, and recovery of the FHR deceleration, as described above. In some embodiments, an alert indicating variable decelerations is provided when an FHR deceleration is detected that is a rapid decrease below a baseline FHR. In some embodiments, a variable deceleration can be classified as non-repetitive and brief if it lasts less than 20 seconds, and repetitive if at least three such decelerations occur over a 20 minute period. In some embodiments, an alert indicating long decelerations is provided when a gradual or rapid decrease in FHR of 15 bpm or more below a baseline FHR is detected that continues for at least 2 minutes from onset to return to the baseline FHR. In some embodiments, an alert is provided based on associated FHR and maternal heart rate (MHR) patterns.

[0057] In some embodiments, the second stage includes "advanced" decision support. In some embodiments, the advanced decision support includes providing a recommendation to continue the observation session. In some embodiments, the advanced decision support includes providing a recommendation to conduct a subsequent observation session. In some embodiments, the recommendation to continue the observation session is provided based on the guidelines and the current findings. For example, in some embodiments, if no more than one acceleration is found during the first 20 minute observation session, a recommendation is provided to continue observation for another 20 minutes. In some embodiments, after such additional 20 minute session, a conclusion of responsiveness or non-responsiveness is provided. In some embodiments, a recommendation is provided to conduct a subsequent observation session based on a finding that the current observation session is uninterpretable and a decision cannot be provided.

[0058] In some embodiments, one or more exemplary processes described herein are executed locally (e.g., on the client side) at the wearable device, e.g., using an on-board computing system of the wearable device. In some embodiments, one or more exemplary processes described herein are executed remotely from the wearable device, e.g., at a remote server, using cloud computing, etc. In some embodiments, the applicable computing system may include hardware components such as a processor, including local or remote processing components. In some embodiments, the processor may include any type of data processing capability, such as hardware logic circuitry, e.g., application specific integrated circuits (ASICs) and programmable logic, or computing devices, e.g., microcomputers or microcontrollers, including programmable microprocessors. In some embodiments, the processor may include data processing capability provided by a microprocessor. In some embodiments, the microprocessor may include memory, processing, interface resources, controllers, and counters. In some embodiments, the microprocessor may include one or more programs stored in memory.

[0059] In some embodiments, the associated computing system may include one or more computer devices implemented as a local or on-premise computing system, a cloud computing system, or a hybrid cloud computing system having one or more hardware and / or software components located in a cloud computing system and one or more hardware and / or software components located in a local computing system. In some embodiments, the terms "cloud", "internet cloud", "cloud computing", "cloud architecture", and similar terms may include at least one of the following: (1) a large number of computers connected via a real-time communication network (e.g., the Internet); (2) providing the ability to run programs or applications simultaneously on a large number of connected computers (e.g., physical machines, virtual machines (VMs)); (3) network-based services that appear to be provided by actual server hardware, but are actually provided by virtual hardware (e.g., virtual servers) simulated by software running on one or more actual machines (e.g., capable of being moved in place and scaled up (or down) without impacting end users). Of course, the foregoing examples are illustrative and not limiting.

[0060] Similarly, a computing system may include storage, such as, for example, one or more local and / or remote data storage solutions, such as a local hard drive, solid state drive, flash drive, database or other local data storage solution or any combination thereof, and / or remote data storage solutions, such as a server, mainframe, database or cloud service, distributed database or other suitable data storage solution or any combination thereof. In some embodiments, storage may include a suitable non-transitory computer-readable medium, such as, for example, one or more buffers and / or caches of random access memory (RAM), read-only memory (ROM), other memory devices, or any combination thereof.

[0061] In some embodiments, an exemplary computing system may implement a computer engine for performing any of the processes described herein. In some embodiments, the terms "computer engine" and "engine" identify at least one software component and / or a combination of at least one software component and at least one hardware component designed / programmed / configured to manage / control other software and / or hardware components (libraries, software development kits (SDKs), objects, etc.).

[0062] Examples of hardware elements may include a processor, a microprocessor, a circuit, a circuit element (e.g., a transistor, a resistor, a capacitor, an inductor, etc.), an integrated circuit, an application specific integrated circuit (ASIC), a programmable logic device (PLD), a digital signal processor (DSP), a field programmable gate array (FPGA), a logic gate, a register, a semiconductor device, a chip, a microchip, a chipset, etc. In some embodiments, the one or more processors may be implemented as a complex instruction set computer (CISC) or reduced instruction set computer (RISC) processor, an x86 instruction set compatible processor, a multi-core, or any other microprocessor or central processing unit (CPU). In various implementations, the one or more processors may be a dual core processor, a dual core mobile processor, etc.

[0063] Examples of software may include a software component, a program, an application, a computer program, an application program, a system program, a machine program, an operating system software, middleware, firmware, a software module, a routine, a subroutine, a function, a method, a procedure, a software interface, an application program interface (API), an instruction set, computing code, computer code, a code segment, a computer code segment, a word, a value, a symbol, or any combination thereof. The decision whether an embodiment is implemented using hardware and / or software elements may vary according to any number of factors, such as desired computation speed, power levels, thermal tolerances, processing cycle budgets, input data rates, output data rates, memory resources, data bus speeds, and other design or performance constraints.

[0064] In some embodiments, the system includes a non-transitory computer memory storing software instructions and at least one computer processor that, upon execution of the software instructions, is configured to: receive an input dataset related to a pregnant woman within a time window; the input dataset includes at least fetal heart rate data, the fetal heart rate data calculated based at least on raw biopotential data obtained by a plurality of biopotential sensors in a garment worn by the pregnant woman; imputing at least one missing data sample to the input dataset to provide a completed dataset; removing at least one segment from the completed dataset based on at least one criterion to provide a corrected dataset; calculating a baseline fetal heart rate based at least on the corrected dataset; detecting at least one event occurring within the time window based on at least one of the corrected dataset or the baseline fetal heart rate; classifying the at least one event as either an acceleration, deceleration, or baseline change based on at least one of the corrected dataset or the baseline fetal heart rate; generating an output indicative of the at least one event; the output is configured to enable at least one action to be taken with respect to the pregnant woman.

[0065] In some embodiments, the at least one criterion for removing at least one segment from the completed data set includes removing at least one segment based on at least one segment including a cyclical change in fetal heart rate, in some embodiments, the cyclical change in fetal heart rate includes at least one of: (a) a change in fetal heart rate of 25 beats per minute or more in consecutive samples, or (b) applying a threshold based on a median and a standard deviation.

[0066] In some embodiments, the computer processor is further configured to determine whether the input dataset is valid or not valid, and if determining that the input dataset is not valid, to receive a further input dataset before completing the input dataset. Completing the input dataset is performed on the further input dataset. In some embodiments, the step of determining whether the input dataset is valid or not includes identifying a plurality of stable segments of the corrected dataset, determining a total duration of the plurality of stable segments, and determining whether the input dataset is valid or not based at least on a comparison of the total duration of the plurality of stable segments to a predetermined threshold duration. Each stable segment includes a sequence of a predetermined amount of consecutive samples having a fetal heart rate variability less than a predetermined threshold variability.

[0067] In some embodiments, determining the baseline fetal heart rate is performed by a process that includes calculating a virtual baseline fetal heart rate, calculating a virtual high fetal heart rate limit, calculating a virtual low fetal heart rate limit, removing outlier data points from the corrected data set, and calculating the baseline fetal heart rate based at least on applying rounding to the corrected data set after removing the outlier data points. An outlier data point is a data point that has a fetal heart rate higher than the virtual high fetal heart rate limit or a fetal heart rate lower than the virtual low fetal heart rate limit. In some embodiments, the virtual baseline fetal heart rate is an average value of the fetal heart rates in the corrected data set. In some embodiments, the steps of calculating the virtual high fetal heart rate limit and calculating the virtual low fetal heart rate limit are performed by a process that includes calculating the standard deviation of a virtual baseline fetal heart rate, calculating the virtual high fetal heart rate limit as the virtual baseline fetal heart rate plus two times the standard deviation, and calculating the virtual low fetal heart rate limit as the virtual baseline fetal heart rate minus two times the standard deviation.

[0068] In some embodiments, at least one of the at least one event is classified as an acceleration, and at least one of the at least one event is classified as an acceleration based on a process including: identifying a first intersection between a trace of the corrected data set and a baseline fetal heart rate; identifying a second intersection between the trace of the corrected data set and the baseline fetal heart rate; identifying a maximum fetal heart rate within a time interval between the first intersection and the second intersection; and identifying the time interval as a value that includes an acceleration based on at least a duration of the time interval and the maximum fetal heart rate within the time interval. In some embodiments, identifying the time interval as a value that includes an acceleration is further based on a gestational age of the pregnant woman. In some embodiments, the method further includes determining a fetal heart rate baseline variability. In some embodiments, determining the fetal heart rate baseline variability is performed by a process that includes determining a highest fetal heart rate within a variability time interval following the second intersection, determining a lowest fetal heart rate within the variability time interval, and subtracting the lowest fetal heart rate within the variability time interval from the highest fetal heart rate within the variability time interval to provide the fetal heart rate baseline variability. The variability time interval extends from the second intersection until the earlier of (1) a predetermined duration following the second intersection, or (2) the onset of a subsequent acceleration or deceleration.

[0069] In some embodiments, the at least one event includes a deceleration, and the at least one event is classified as a deceleration based on a process including: identifying a first intersection between a trace of the corrected data set and a baseline fetal heart rate; identifying a second intersection between the trace of the corrected data set and the baseline fetal heart rate; identifying a minimum fetal heart rate within a time interval between the first intersection and the second intersection; and identifying the time interval as a value including a deceleration based on at least a difference between the baseline fetal heart rate and the minimum fetal heart rate within the time interval. In some embodiments, the method further includes classifying the deceleration as either an early deceleration, a late deceleration, a variable deceleration, or a long deceleration. In some embodiments, the input data set further includes a uterine contraction signal. The deceleration is classified as one of a fast deceleration, a slow deceleration, a variable deceleration, or a long deceleration based on a process that includes determining a delay, which is the amount of time elapsed between the first intersection and a time of a minimum fetal heart rate within the time interval, determining a time of a peak of a uterine contraction within the time interval, determining a time difference between the time of the minimum fetal heart rate within the time interval and a time of a peak of the uterine contraction signal within the time interval, and classifying the deceleration as one of a fast deceleration, a slow deceleration, a variable deceleration, or a long deceleration based on at least the delay, the time difference, and the duration of the uterine contraction.

[0070] In some embodiments, two iterations of the steps of: (1) removing at least one segment from the completed dataset based on at least one criterion to provide a corrected dataset; (2) calculating a baseline fetal heart rate based at least on the corrected dataset; and (3) detecting at least one event that occurred within the time window based on at least one of the corrected dataset or the baseline fetal heart rate are performed for the time window.

[0071] In some embodiments, the method includes repeating, for a subsequent time window, the steps of receiving an input dataset of a pregnant woman, completing at least one missing data sample in the input dataset to provide a completed dataset, removing at least one segment from the completed dataset based on at least one criterion to provide a corrected dataset, calculating a baseline fetal heart rate based at least on the corrected dataset, detecting at least one event that occurred within the time window based on at least one of the corrected dataset or the baseline fetal heart rate, and classifying the at least one event as either an acceleration, a deceleration, or a baseline change based on at least one of the corrected dataset or the baseline fetal heart rate. In some embodiments, the subsequent time window does not overlap with the time window. In some embodiments, the subsequent time window overlaps with the time window by an overlap duration. In some embodiments, the overlap duration is between 5% and 95% of the duration of the time window.

[0072] Research results

[0073] Pregnant women with high-risk singleton pregnancies ≥32 weeks gestation underwent self-administered remote nonstress testing (NST) using a sensor-equipped belt sold under the trade name INVU by Nuvo Group, Limited, Tel Aviv, Israel. Remote NSTs performed during the study period replaced in-clinic NSTs. NSTs were reviewed by clinicians via a web application. Clinicians determined clinical interpretability and responsiveness. Interpretable sessions were retrospectively analyzed for responsiveness by a clinical decision support system (CDSS), as described herein, and compared with clinician assessments. Mean time to reach responsiveness by CDSS was compared with mean time of NST session.

[0074] A total of 30 remote NSTs were performed among 9 pregnant women during the study period. Of the 30 remote NSTs, 27 (i.e., 90%) were judged interpretable by the reviewing clinician. Figure 9A shows an NST that was judged interpretable. Figure 9B shows an NST that was judged uninterpretable.

[0075] Of the sessions judged interpretable, 27 (i.e., 100%) were defined as responsive by the reviewing clinician. Twenty-five of these 27 sessions (i.e., 92.6%) were also judged as responsive by the CDSS. Figure 9C shows the CDSS results overlaid on the NST in Figure 9A, with the accelerations identified by the CDSS shown in contrasting colors. Of the two sessions judged as non-responsive by the CDSS, both were borderline cases in terms of the ACOG guidelines.

[0076] Among cases identified as responsive by the CDSS, the mean time to CDSS recognition was 22.40 ± 6.68 minutes. The mean duration of NST sessions performed as part of this study was significantly longer, 36.26 ± 16.57 minutes (t test, p value < 0.0001). Thus, this study demonstrated the feasibility of remote NST performed using the CDSS described herein. Furthermore, this study demonstrated that the CDSS described herein accurately identifies responsiveness, potentially reducing the length of NST sessions and improving clinician workflow. A poster summarizing the results of this study is attached as Appendix A.

[0077] While multiple embodiments of the present invention have been described, it is understood that these embodiments are illustrative and not limiting, and that numerous variations may be apparent to those skilled in the art. For example, all dimensions discussed herein are provided by way of example only, and are not intended to be limiting dimensions.

Claims

1. 1. A computer-implemented method comprising: The method includes receiving, by a computing device, an input data set relating to a pregnant woman within a time window; the input data set includes at least fetal heart rate data; the fetal heart rate data is calculated based at least on raw biopotential data obtained by a plurality of biopotential sensors in a garment worn by the pregnant woman; The method comprises: imputing, by the computing device, at least one missing data sample into the input data set to provide a completed data set; removing, by the computing device, at least one segment from the completed data set based on at least one criterion to provide a corrected data set; calculating, by the computing device, a baseline fetal heart rate based at least on the corrected data set; detecting, by the computing device, at least one event that occurred within the time window based on at least one of the corrected data set or the baseline fetal heart rate; classifying, by the computing device, the at least one event as either an acceleration, a deceleration, or a baseline change based on at least one of the corrected data set or the baseline fetal heart rate; generating, by the computing device, an output indicative of at least the at least one event; Including, the output is configured to enable at least one action to be taken with respect to the pregnant woman. method.

2. 2. The method of claim 1, wherein the at least one criterion for removing the at least one segment from the completed data set comprises removing the at least one segment based on the at least one segment including a cyclic change in fetal heart rate.

3. 3. The method of claim 2, wherein the periodic variation in the fetal heart rate includes at least one of: (a) the fetal heart rate changing by more than 25 beats per minute in consecutive samples; or (b) applying a threshold based on a median and a standard deviation.

4. determining, by the computing device, whether the input data set is valid or not valid; if the input data set is determined to be invalid, receiving, by the computing device, a further input data set before performing the step of complementing the input data set; Further comprising: and the step of complementing the input data set by the computing device is performed on the further input data set. The method of claim 1.

5. The step of determining whether the input data set is valid or not valid comprises: identifying, by the computing device, a plurality of stable segments of the corrected data set; determining, by the computing device, a total duration of the plurality of stable segments; determining, by the computing device, whether the input data set is valid or not valid based at least on a comparison of the total duration of the plurality of stable segments to a predetermined threshold duration; It is executed by a process that includes Each stable segment includes a sequence of a predetermined amount of consecutive samples having fetal heart rate variability less than a predetermined threshold variability. The method according to claim 4.

6. The step of determining, by the computing device, the baseline fetal heart rate includes: calculating, by the computing device, a virtual baseline fetal heart rate; calculating, by the computing device, a virtual high fetal heart rate limit; calculating, by the computing device, a virtual low fetal heart rate limit; removing, by the computing device, outlier data points from the corrected data set; and calculating, by the computing device, the baseline fetal heart rate based at least on application of rounding to the corrected data set after removing the outlier data points; It is executed by a process that includes the outlier data points are data points having a fetal heart rate higher than the virtual high fetal heart rate limit or a fetal heart rate lower than the virtual low fetal heart rate limit. The method of claim 1.

7. 7. The method of claim 6, wherein the virtual baseline fetal heart rate is the average of the fetal heart rates in the corrected data set.

8. The steps of calculating a virtual high fetal heart rate limit and calculating a virtual low fetal heart rate limit include: calculating, by the computing device, a standard deviation of the virtual baseline fetal heart rate; calculating, by the computing device, the virtual high fetal heart rate limit as the virtual baseline fetal heart rate plus two times the standard deviation; calculating, by the computing device, the virtual low fetal heart rate limit as the virtual baseline fetal heart rate minus two standard deviations; executed by a process including The method according to claim 6.

9. At least one of the at least one event is classified as an acceleration, and at least one of the at least one event is identifying, by the computing device, a first intersection point between a trace of the corrected data set and the baseline fetal heart rate; identifying, by the computing device, a second intersection point between a trace of the corrected data set and the baseline fetal heart rate; identifying, by the computing device, a maximum fetal heart rate within a time interval between the first intersection point and the second intersection point; identifying, by the computing device, the time interval as a value that includes acceleration based on at least a duration of the time interval and the maximum fetal heart rate within the time interval; are classified as accelerated based on the process, The method of claim 1.

10. The method of claim 9 , wherein identifying the time interval as a value that includes the acceleration is further based on a gestational age of the pregnant woman.

11. 10. The method of claim 9, further comprising determining, by the computing device, fetal heart rate baseline variability.

12. The step of determining the fetal heart rate baseline variability by the computing device comprises: determining, by the computing device, a highest fetal heart rate within a dispersion time interval following the second intersection; determining, by the computing device, a minimum fetal heart rate within the variability time interval; subtracting, by the computing device, the lowest fetal heart rate within the variability time interval from the highest fetal heart rate within the variability time interval to provide the fetal heart rate baseline variability; It is executed by a process that contains the variation time interval extends from the second intersection to the earlier of (1) a predetermined duration following the second intersection, or (2) the onset of a subsequent acceleration or deceleration; The method of claim 11.

13. The at least one event includes a deceleration, and the at least one event includes: identifying, by the computing device, a first intersection point between a trace of the corrected data set and the baseline fetal heart rate; identifying, by the computing device, a second intersection point between a trace of the corrected data set and the baseline fetal heart rate; identifying, by the computing device, a minimum fetal heart rate within a time interval between the first intersection point and the second intersection point; identifying, by the computing device, the time interval as a value that includes a deceleration based at least on a difference between the baseline fetal heart rate and the minimum fetal heart rate within the time interval; are classified as decelerating based on the process, The method of claim 1.

14. The method of claim 13 , further comprising classifying, by the computing device, the deceleration as either a fast deceleration, a slow deceleration, a variable deceleration, or a long deceleration.

15. The input data set further comprises a uterine contraction signal; The deceleration is determining, by the computing device, a delay that is an amount of time elapsed between the first intersection point and a time of the minimum fetal heart rate within the time interval; determining, by the computing device, a time of peak uterine contraction within the time interval; determining, by the computing device, a time difference between a time of the minimum fetal heart rate within the time interval and a time of a peak of the uterine contraction signal within the time interval; classifying, by the computing device, the deceleration as one of the fast deceleration, the slow deceleration, the variable deceleration, or the long deceleration based on at least the delay, the time difference, and the duration of the contraction; the fast deceleration, the slow deceleration, the variable deceleration, or the long deceleration based on a process including: The method of claim 14.

16. 2. The method of claim 1, wherein two iterations of the following steps are performed for the time window: (1) removing, by the computing device, at least one segment from the completed dataset based on at least one criterion to provide a corrected dataset; (2) calculating, by the computing device, a baseline fetal heart rate based at least on the corrected dataset; and (3) detecting, by the computing device, at least one event that occurred within the time window based on at least one of the corrected dataset or the baseline fetal heart rate.

17. For subsequent time windows, receiving, by the computing device, an input data set of the pregnant woman; imputing, by the computing device, at least one missing data sample into the input data set to provide a completed data set; removing, by the computing device, at least one segment from the completed data set based on at least one criterion to provide a corrected data set; calculating, by the computing device, a baseline fetal heart rate based at least on the corrected data set; detecting, by the computing device, at least one event that occurred within the time window based on at least one of the corrected data set or the baseline fetal heart rate; classifying, by the computing device, the at least one event as either an acceleration, a deceleration, or a baseline change based on at least one of the corrected data set or the baseline fetal heart rate; This includes repeating The method of claim 1.

18. The method of claim 17 , wherein the subsequent time window does not overlap with the time window.

19. The method of claim 17 , wherein the subsequent time window overlaps with the time window by an overlap duration.

20. The method of claim 19 , wherein the overlap duration is between 5% and 95% of the duration of the time window.

Images