HEART RATE COMPLEXITY

Technical Field

The present disclosure generally relates to assessing clinical event risk based on heart rate complexity. More particularly, the present disclosure describes various embodiments of a system and a method for assessing the risk of clinical events based on heart rate complexity of a subject.

Background

Heart rate variability (HRV) has been widely studied for many years and is also routinely measured in Holter monitor tests. A Holter monitor is portable device that measures and records the heart's electrocardiogram (ECG) activity continuously for a period of time. HRV is considered as a sensitive marker of autonomic activity and is most commonly measured using time or frequency domain parameters. However, due to the complex interactions in heart rate regulation, its mechanism remains a controversial topic in physiology and the clinical application of HRV analysis has remained limited.

More recently, a different approach was developed for heart rate analysis. Particularly, heart rate complexity (HRC) was analyzed through various non-linear measures, such as through entropy. Studies of HRC in healthy individuals, for example, have found a tendency towards 1/f noise patterns. 1/f noise or pink noise is a signal with a frequency spectrum such that the power spectral density is inversely proportional to the frequency of the signal. Pink noise is the most common signal in biological systems.

Additionally, perturbations of this healthy state are associated with changes in HRC and have been associated with a wide range of conditions including heart failure, senility, psychiatric disorders, traumatic brain injury, and strokes. This is postulated to be due to derangements in the balance of the autonom ic system and have been shown to correlate with severity as well as clinical outcomes in both cardiac disease and non- cardiac conditions such as trauma and stroke.

It has also been shown that fractal complexity measures of heart rate dynamics in patients with implantable cardioverter defibrillator (ICDs) may predict future events such as ICD shocks or death. These were based on specialized high-resolution ECG recordings which were only done at a single time-point. More recently, pre-enrolment Holter data from the SCD-HeFT (Sudden Cardiac Death in Heart Failure Trial) study was used to risk stratify patients in the study and the long-term fractal exponent, a2, from detrended fluctuation analysis was found to be most predictive of sudden cardiac death events. Notably, all these studies used different techniques to analyze HRC, though these are all based on open domain mathematical formulae. However, much of this early work did not find widespread application of HRC analysis.

Earlier attempts at predicting the onset of heart failure, for instance, relied on changes in parameters such as the OptiVol fluid index and thoracic impedance which increases with the accumulation of fluid in the lungs. However, these parameters such as the have failed in clinical studies to be reliably predictive. Furthermore, clinical deterioration often preceded the changes in these parameters and hence they did not improve the patients’ care in this respect.

Therefore, in order to address or alleviate at least one of the aforementioned problems and/or disadvantages, there is a need to provide an improved system and method for assessing the risk of clinical events based on heart rate complexity.

Summary

According to an aspect of the present disclosure, there is a system and method for assessing risks of clinical events based on heart rate complexity of a subject. The system comprises a physiological sensing device connectable to the subject; and an electronic device communicable with the physiological sensing device, the electronic device comprising a processor configured for performing the method. The method comprises: receiving, from the physiological sensing device, a physiological dataset measured from the subject, the physiological dataset comprising a set of heart rate data; generating a heart rate time series from the heart rate data; calculating a set of entropy data from the heart rate time series; calculating an HRC index from the entropy data; comparing the HRC index against predefined reference data; and generating a risk signal in response to a determination from said comparing that the HRC index is indicative of increased risks of clinical events to the subject.

An advantage of the present disclosure is that the HRC index can be used as an indicator of increased risks of clinical events to the subject. The HRC index serves as a diagnostic modality that helps to quantify a subject’s health state. Clinicians can rely on the HRC index with other physiological data to look for derangements in the physiology of the subject that may predict impending significant adverse cardiovascular events. This allows the clinicians to intervene early to avert further worsening.

A system and method for assessing the risk of clinical events based on heart rate complexity according to the present disclosure are thus disclosed herein. Various features, aspects, and advantages of the present disclosure will become more apparent from the following detailed description of the embodiments of the present disclosure, by way of non-limiting examples only, along with the accompanying drawings.

Brief Description of the Drawings

Figure 1 A and Figure 1 B are schematic illustrations of a system for assessing risks of clinical events based on heart rate complexity of a subject, in accordance with embodiments of the present disclosure.

Figure 2 is a flowchart illustration of a computerized method system for assessing risks of clinical events based on heart rate complexity of a subject, in accordance with embodiments of the present disclosure. Figure 3 illustrates the average profile parameters of the patients involved in the clinical study of the system and method in accordance with embodiments of the present disclosure.

Figure 4 is a graph illustration of the results of the clinical study.

Figure 5 is a block diagram illustration of the technical architecture of a computer, in accordance with embodiments of the present disclosure.

Detailed Description

For purposes of brevity and clarity, descriptions of embodiments of the present disclosure are directed to a system and method for assessing the risk of clinical events based on heart rate complexity, in accordance with the drawings. While aspects of the present disclosure will be described in conjunction with the embodiments provided herein, it will be understood that they are not intended to limit the present disclosure to these embodiments. On the contrary, the present disclosure is intended to cover alternatives, modifications and equivalents to the embodiments described herein, which are included within the scope of the present disclosure as defined by the appended claims. Furthermore, in the following detailed description, specific details are set forth in order to provide a thorough understanding of the present disclosure. Flowever, it will be recognized by an individual having ordinary skill in the art, i.e. a skilled person, that the present disclosure may be practiced without specific details, and/or with multiple details arising from combinations of aspects of particular embodiments. In a number of instances, well-known systems, methods, procedures, and components have not been described in detail so as to not unnecessarily obscure aspects of the embodiments of the present disclosure.

In embodiments of the present disclosure, depiction of a given element or consideration or use of a particular element number in a particular figure or a reference thereto in corresponding descriptive material can encompass the same, an equivalent, or an analogous element or element number identified in another figure or descriptive material associated therewith. References to“an embodiment / example”,“another embodiment / example”,“some embodiments / examples”,“some other embodiments / examples”, and so on, indicate that the embodiment(s) / example(s) so described may include a particular feature, structure, characteristic, property, element, or limitation, but that not every embodiment / example necessarily includes that particular feature, structure, characteristic, property, element or limitation. Furthermore, repeated use of the phrase “in an embodiment / example” or “in another embodiment / example” does not necessarily refer to the same embodiment / example.

The terms“comprising”,“including”,“having”, and the like do not exclude the presence of other features / elements / steps than those listed in an embodiment. Recitation of certain features / elements / steps in mutually different embodiments does not indicate that a combination of these features / elements / steps cannot be used in an embodiment.

As used herein, the terms“a” and“an” are defined as one or more than one. The use of 7” in a figure or associated text is understood to mean“and/or” unless otherwise indicated. The term“set” is defined as a non-empty finite organization of elements that mathematically exhibits a cardinality of at least one (e.g. a set as defined herein can correspond to a unit, singlet, or single-element set, or a multiple-element set), in accordance with known mathematical definitions. The recitation of a particular numerical value or value range herein is understood to include or be a recitation of an approximate numerical value or value range.

As used herein, the terms“component”,“module,”“system”,“interface� ��, and the like are generally intended to refer to a computer-related entity, either hardware, a combination of hardware and software, software, or software in execution. For example, a component or a module may be, but is not limited to being, a process running on a processor, a processor, an object, an executable, a thread of execution, a program, and/or a computer. By way of illustration, both an application running on a controller and the controller can be a component / module. One or more components / modules may reside within a process and/or thread of execution. A component / module may be localized on one computer and/or distributed among a plurality of computers.

In representative or exemplary embodiments of the present disclosure, as shown in Figure 1A and Figure 1 B, there is a system 100 for assessing risks of clinical events based on heart rate complexity (FIRC) of a subject 1 10. In many embodiments described herein, the subject 1 10 may be a human subject or a patient. Flowever, in other embodiments, the subject 1 10 may be a non-human subject or a living animal such as pet dogs and cats.

From the physics theory of complex systems, it is observed that a healthy complex living system exhibits a certain universal pattern, and that it is measurable in physiological data or parameters such as heart rate variability (FIRV). FIRV is defined as the physiological phenomenon of variation in the time interval between heartbeats, and is measured by the variation in the beat-to-beat or R-R interval. Early studies into FIRV found some evidence in cardiology for this, in that deviation from this pattern is correlated to ageing or cardiac illnesses. This is thought to reflect disequilibrium in the autonomic nervous system (sympathetic versus parasympathetic) and measurements of FIRC allows determination of the degree of perturbation from a healthy state.

As discussed in the background above, known studies on FIRC did not find widespread application possibly because they predate the advent of advanced home monitoring, diagnostics, and wearable technology. Flowever, much progress has been made to overcome many of these technological challenges in recent years and it is now feasible to apply these concepts to cardiac disease. The system 100 described herein presents a novel approach to FIRC analysis that provides advantages such as the ability to monitor ICD patients for impending clinic events in real time.

The system 100 includes a physiological sensing device 120 connected or connectable to the subject 1 10 for measuring physiological data from the subject 1 10. The physiological sensing device 120 may be a heart monitoring device such as an implantable cardiac device, implantable cardioverter defibrillator (ICD), implantable loop recorder, pacemaker, or electrocardiogram (ECG) monitor. The physiological sensing device 120 may be a clinical device used for longer-term external electrocardiogram (ECG) or heart rate monitoring, such as an ambulatory ECG recorder or a consumer wearable (e.g. an external wearable device) with heart rate monitoring capabilities. In many embodiments, the physiological sensing device 120 includes a heart rate sensor 122 for measuring heart rate data of the subject 1 10, and optionally includes a motion sensor 124 for measuring motion data of the subject 1 10. The physiological sensing device 120 further includes a database 126 for storing data, e.g. the heart rate data / motion data, measured by the physiological sensing device 120 from the subject 1 10.

The system 100 further includes an electronic device 130 communicable with the physiological sensing device 120. The electronic device 130 includes a processor for executing program instructions, such as to perform a computerized method for assessing risks of clinical events based on HRC of the subject 1 10. The electronic device 130 further includes a database 132 for storing data, e.g. the heart rate data / motion data, measured by and communicated from the physiological sensing device 120.

In some embodiments as shown in Figure 1A, the electronic device 130 is a computer device 130a of the subject 1 10. Some non-limiting examples of the computer device 130a include a desktop device, mobile device, personal digital assistant (PDA), mobile phone, smartphone, tablet, phablet, and laptop computer. The computer device 130a may be communicable with the physiological sensing device 120 via wireless / contactless communication protocols such as Bluetooth and Wi-Fi. For example, the physiological sensing device 120 and the computer device 130a may share the same Wi-Fi network and data is communicable therebetween. The physiological sensing device 120, such as if it is an external wearable device, may also be communicable with the computer device 130a via wired communications such as USB.

In one embodiment, the physiological sensing device 120 stores the physiological data on the database 126 before communicating the physiological data to the computer device 130a, such as upon request by the subject 1 10, for storing on the database 132. In another embodiment, there is an active communication link between the physiological sensing device 120 and the computer device 130a such that the physiological data is automatically shared with the computer device 130a immediately or shortly after measurement thereof.

In some embodiments as shown in Figure 1 B, the electronic device 130 is a remote server 130b and the system 100 further includes a computer device 140 of the subject 1 10. This computer device 140 is similar to the computer device 130a described above with reference to Figure 1A. The computer device 140 functions as an intermediary device between the physiological sensing device 120 and the remote server 130b, such that the physiological sensing device 120 is communicable with the remote server 130b via the computer device 140. The computer device 140 may be communicable with the physiological sensing device 120 via wireless / contactless communication protocols such as Bluetooth and Wi-Fi. The computer device 140 may be communicable with the remote server 130b across a wide area network (WAN) or the Internet. The computer device 140 may include a database 142 for storing data, e.g. the heart rate data / motion data, measured by and communicated from the physiological sensing device 120.

The remote server 130b is a computer server that is located remotely away from the computer device 140. The remote server 130b is a physical or cloud data processing system on which a server program runs. The server may be implemented in hardware or software, or a combination thereof. The server includes computers, laptops, mini- computers, mainframe computers, any non-transient and tangible machines that can execute a machine-readable code, cloud-based servers, distributed server networks, and a network of computer systems.

In one embodiment, the physiological sensing device 120 communicates the physiological data to the computer device 140 for storing on the database 142. The computer device 140 subsequently communicates the physiological data to the remote server 130b at a later time for storing on the database 132. In another embodiment, there is an active communication link between the computer device 140 and the remote server 130b such that the physiological data is automatically shared with the remote server 130b. With reference to Figure 2, there is shown a computer-implemented or computerized method 200 for assessing risks of clinical events based on HRC of the subject 1 10. The method 200 is implemented on and performed by the electronic device 130, i.e. the computer device 130a or remote server 130b, which comprises various modules / components, including the processor for executing program instructions, for performing various steps of the method 200.

A software application may be installed and executed on the electronic device 130 to perform the method 200. The software application may be implemented on Java and the electronic device 130 has suitable specification to execute the software application. For example, the electronic device 130 has a 1 GFIz processor, 256 GB RAM, and 1 GB hard disk space. The subject 1 10 may need to setup his/her profile to include profile parameters such as gender, date of birth, age, weight, height, and health background. The subject profile parameters may be used to determine reference conditions for the risk assessment.

The method 200 includes a step 202 of receiving, from the physiological sensing device 1 10, a physiological dataset measured from the subject 1 10, the physiological dataset including a set of heart rate data measured by the heart rate sensor 122. The physiological dataset optionally includes a set of motion data measured by the motion sensor 124. The method 200 further includes a step 204 of generating a heart rate time series from the set of heart rate data. The heart rate time series is defined as the consecutive R-R intervals on ECG records.

The method further includes a step 206 of calculating a set of entropy data, such as multi-scale entropy (MSE) values, from the heart rate time series. The method 200 further includes a step 208 of calculating an FIRC index from the entropy data. FIRC indices are parameters or scores that can be used to assess the risks of clinical events to the subject 1 10. The method 200 further includes a step 210 of comparing the FIRC index against predefined reference data. The predefined reference data may be arbitrarily defined and/or derived from the reference conditions of the subject 1 10 and/or a population of subjects 1 10. The method 200 further includes a step 212 of determining, from said comparing in the step 210, if the HRC index is indicative of increased risks of clinical events to the subject 1 10. If yes, the step 212 proceeds to a step 214 of generating a risk signal in response to the determination from said comparing that the HRC index is indicative of increased risks of clinical events to the subject 1 10. Conversely, if the HRC index is not indicative, the step 212 proceeds to a step 216 of analyzing the next physiological dataset measured from the subject 1 10 and returns to the step 202.

The method 200 may further include presenting the risk signal to alert the subject 1 10 of the increased risks of clinical events. In one embodiment, the electronic device 130 is the computer device 130a having an audio and/or visual module for presenting the risk signal, such as in the form of visual graphics and sound alarms. In another embodiment, the electronic device 130 is the remote server 130b configured for communicating the risk signal, such as a message, to the computer device 140 to thereby present the risk alert or message to the subject 1 10.

The benchmark for optimal health can be defined through multi-scale entropy of a continuous heart rate time series. Entropy measurement of a heart rate time series can refer to sample entropy, approximate entropy, or other entropy definitions that measures the entropy data in a heart rate time series Multi-scale entropy refers to the entropy values at different levels of coarse-graining time scales.

In some embodiments, a R-R interval time series or heart rate time series generated from the set of heart rate data is represented by x_ _t , where t = 0, 1 , 2, ... , N. The heart rate time series x _t is transformed into a new heart rate time series by coarse graining as shown in Equation 1 below for a scale m that is larger than 1 . m represents the level of coarse graining of x _t and the coarse-grained heart rate time series is represented by y _t>m · The set of entropy data is generated from the coarse-grained heart rate time series y _t,m and includes the entropy values (represented by E _m ) calculated from the heart rate time series y _t,m . Notably, the entropy value at scale m = 1 is represented by Ei. [Equation 1 ]

The optimal healthy status is defined as the E _m value from a heart rate time series y _t,m having the same E _m value for all the integer values of m. This E _m value is represented by Emax and the Emax value can be calculated through numerical methods. For heart rate time series x _t , a subject 1 10 in ill health have E _m values that are consistently smaller than the Emax value at m > 10. Additionally, the amount of deviation in the E _m values is correlated with the severity of the illness and can be a measure of the risk of future adverse clinical events. Hence, the amount of deviation can be transformed into the HRC index to signify the degree of deviation from a healthy state. Accordingly, the HRC index can be calculated from the set of entropy data calculated from the heart rate time series yt _,m .

In one embodiment, the HRC index represented by I _t is calculated as the average of entropy values at scales m from 21 to 40 inclusive and weighted by a fixed value Emax, as shown in Equation 2 below. Approximate entropy may be used to calculate Emax with threshold standard deviation of 0.15 and data length of 2. The value of Emax is calculated to be 1 .8.

[Equation 2]

HRC indices obtained from the method 200 are compared to the GRACE score. The GRACE score is a well-validated cardiac risk scoring system to risk stratify patients diagnosed with acute coronary syndrome (ACS) to estimate their in-hospital and 6- month to 3-year mortality. A clinical study was performed on patients diagnosed with ACS to validate the HRC indices against the GRACE scores. Figure 3 illustrates a table 300 of the average profile parameters of the patients involved in the clinical study. Higher risk patients had more depressed HRC compared to lower risk patients, which was found to be most evident during waking hours and also in the first 24 hours after their presentation. Their HRC also remained largely the same while low risk patients had a decline in their HRC during their convalescence as inpatients. Figure 4 illustrates a graph 400 of the correlation and validation of the HRC indices against the GRACE scores. The R-squared or R ² value and the p value from the graph 400 shows that the HRC indices correlates well with the GRACE scores, suggesting that the HRC indices can be used to risk stratify patients with diagnosed ACS and potentially other diseases. Interestingly, unlike earlier studies, the correlation of the HRC indices with the GRACE scores was independent of left ventricular ejection fraction (LVEF).

In some embodiments, the HRC index I _t is calculated from the entropy data and one or more profile parameters of the subject 1 10. The subject profile parameters may include the subject’s age derived from the initial profile setup by the subject 1 10. In one embodiment, the multi-scale entropy (represent by MSE) and age (represented by Age) values are used together to calculate the HRC index l _t , as shown in Equation 3 below ai and bi represent coefficients which may be arbitrarily defined or calculated through numerical methods. Incorporating the logarithmic function of the subject’s age to calculate the HRC index I _t , improved the correlation from approximately 75% to approximately 80%.

[Equation 3]

MSE

k = CL X -——— - + b

log (Age)

In some embodiments, the physiological dataset includes the set of heart rate data and further includes a set of motion data measured by the motion sensor 124. The motion data may be used to reduce non-linearity in the heart rate data, such that the more useful non-linearity in the heart rate time series can be retrieved to improve the correlation and achieve higher prediction accuracies. The motion data can be used for activity context sensing to complement the heart rate data, since the heart rate data is variable in purely the time domain. The motion data allows for the detection of physical activity carried out by the subject 1 10. The motion data can be used to identify the heart rate data during certain periods, such as the resting period. Light activity such as walking would cause a linear trend in the heart rate data. The motion data can be used to remove the linear trend caused by the activity while retaining the linear trend intrinsic to the heart rate variability. In one embodiment, the activity level (represented by m _t ) is determined from the motion data and introduced to the heart rate time series x _t generated from the measured set of heart rate data to thereby transform x _t into a new heart rate time series (represented by yi), as shown in Equation 4 below. 3 ₂ represents a coefficient which may be arbitrarily defined or calculated through numerical methods. In another embodiment, convolutional kernels may be used on the motion data instead of a simple linear correlation with the coefficient 32 as shown in Equation 4.

[Equation 4]

y _t = x _t - a ₂ x m _t

In one embodiment, the entropy data is generated from the new heart rate time series y _t and the HRC index I _t is calculated from the entropy data, as described above in the method 200. In another embodiment, the HRC index l _t is calculated from motion data together with the multi-scale entropy (represent by MSE) and age (represented by Age) values to further improve its correlation and accuracy. The motion data is regressed into Equation 3 which is then transformed into Equation 5 below a represents a coefficient which may be arbitrarily defined or calculated through numerical methods.

[Equation 5]

MSE

h = X + b x m _t + c

log (Age)

The MSE values may change immediately after a change in the activity state of the subject 110. The amount of change in the MSE values can be an indicator of cardiac health and risk. For example, the motion data may indicate a sudden change in physical or activity state from rest to walking. There is a first or current MSE value represented by MSEi for the rest state, and a second or next MSE value represented by MSE2 for the walking state. The MSE difference MSE2 - MSE1 can be used as the indicator.

HRC analysis and calculation of the HRC indices can thus be modified to reflect changes in activity levels of the subject 1 10 during the awake periods and the resting / sleeping periods. The basal state of the human body of the subject 1 10 during the awake periods when the subject 1 10 is not engaged in extreme states of activity or inactivity is likely to best reflect the adaptability of physiological systems. The optimal balance of the physiological systems, or lack thereof, is most evident in the basal state during the awake periods. The basal state may be defined as the resting metabolic state of the body early in the morning after fasting for a minimum of 12 hours. During the resting / sleeping periods, many of the bodily functions are inactive or suppressed. Conversely, during intense physical exercise, the body’s metabolism is stretched well beyond its baseline level. The clinical study showed that the HRC index is significantly higher during the awake periods than the resting / sleeping periods. Thus, the HRC indices work optimally to indicate or predict increased risks of clinical events, such as cardiac risk, for subjects 1 10 who are awake. The motion data can be used to determine the physical status of the subject 1 10 and thereby identify the time periods for calculating the HRC indices.

In some embodiments, the set of heart rate data is divided into a plurality of subsets of heart rate data, each subset of heart rate data measured during a predefined time period. Consequently, the heart rate time series generated from the set of heart rate data may include a plurality of segments, each segment generated from one of the subsets of heart rate data. The division and segmentation would allow for calculation of the HRC indices for different time periods, as described further below.

The heart rate time series x _t generated from the measured set of heart rate data is cleaned to remove noise. The cleaned heart rate time series is divided into a plurality of subsets represented by X _k according to the time of the day during which the respective subset of heart rate data was measured. In one embodiment, the heart rate data is batched into 6-hour periods, such as 0000 to 0600 hours, 0600 to 1200 hours, 1200 to 1800 hours, and 1800 to 0000 hours, so that there are four subsets of the heart rate data per day. In another embodiment, a moving window approach is used every time the FIRC index is calculated. For example, when the FIRC index is calculated at time t, the subset X _k contains the heart rate data between time t hours and time t - 6 hours.

As stated above, the heart rate time series x _t is transformed into a new heart rate time series y _t,m by coarse graining where m represents the level of coarse graining of x _t. In one embodiment, the coarse graining is done mathematically through the average value of every set of m values, as shown in Equation 1 above. In another embodiment, a moving window approach is used to obtain more data points in the coarse-grained data, as shown in Equation 6 below.

[Equation 6]

The coarse-grained heart rate time series y _t,m is segmented into a plurality of segments according to the subsets Xk and the segments are represented by Yk, _m . For each subset Xk, the standard deviation of the respective heart rate time series segment Yk,m is also calculated and is represented by c¾.

The set of entropy data including the E _m values is generated from the coarse-grained heart rate time series y _t,m. Specifically, the set of entropy data includes a plurality of subsets of entropy data represented by Ek,m, each subset of entropy data calculated from one of the segments of the heart rate time series Yk, _m . In one embodiment, sample entropy is used to calculate the entropy data or entropy values Ek,m. For each coarse-grained heart rate time series segment Yk,m, two series of vectors Uk,m and Vk, _m are formed as shown in Equations 7 and 8 below. Additionally, parameters ut,m and v _t,m are denoted according to Equations 9 and 10 below so that Equations 7 and 8 can be transformed into Equations 11 and 12 below.

[Equation 7]

[Equation 9]

[Equation 12] c,m t,mj

In the vector series Uk,m, the distance between each pair of vectors is calculated, and the total number of pairs that have distances larger than 0.15 O _k is represented by Ak. Similarly, in the vector series Vk, _m , the distance between each pair of vectors is calculated, and the total number of pairs that have distances larger than 0.15 O _k is represented by Bk. The entropy values Ek,m are then calculated according to Equation 13 below.

[Equation 13]

The HRC indices are subsequently calculated for each heart rate time series segment Yk,m and the HRC indices are represented by . As shown in Equation 14 below and similar to Equation 2 above, each HRC index I is calculated as the average of the respective entropy values Ek,m at scales m from 21 to 40 inclusive and weighted by a fixed value E _max . Approximate entropy may be used to calculate E _max with threshold standard deviation of 0.15 and data length of 2. The value of E _max is calculated to be 1 .8. [Equation 14]

The correlation and prediction accuracy of the HRC indices may be improved by introducing other parameters such as the motion data and age of the subject 1 10. These parameters may be introduced to modify the heart rate data and/or modify the HRC indices .

In one embodiment, the motion data is used to exclude the sleeping / resting period from the heart rate data, since the awake period has higher correlation with cardiac risk than the sleeping / resting period. The motion data can also be used to reduce the non-linearity in the heart rate data. For example, the motion data or activity level m _t is introduced to transform the heart rate time series x _t into a new heart rate time series y _t , as shown in Equation 15 below. a3 represents a coefficient or normalization constant which may be arbitrarily defined or calculated through numerical methods. In another embodiment, convolutional kernels may be used on the motion data instead of a simple linear correlation with the coefficient a _å as shown in Equation 4.

[Equation 15]

₌ _ _

y ^t 1 + log(a ₃ x m _t + 1)

In another embodiment, an initial HRC index I _k is modified with the value of the subject’s age. For example, the initial HRC index is multiplied with the inverse of the logarithmic function of the subject’s age, as shown in Equation 16 below. The modified HRC index is represented by I _k . This modification improves the correlation of the modified HRC index I _k with the GRACE score.

[Equation 16] In another embodiment, the method 200 further includes calculating a composite index from the HRC index and the motion data, wherein the composite index is assistive in the determination that the HRC index is indicative of increased risks of clinical events to the subject 1 10, such as for prediction of cardiac risk. The composite index is represented by f(I _k , m _t ), where f is any function with two variables - in this case the HRC index I _k and the motion data m _t .

As stated above, the HRC indices are calculated for different time periods based on the respective heart rate time series segment Yk,m and respective heart rate data subset Xk. Accordingly, the HRC indices may be arranged chronologically and include a current HRC index and a previous HRC index k-i. The current HRC index is calculated for a first or current time period and the previous HRC index -i is calculated for a second or previous time period preceding the first time period. The first and second time periods may have identical durations so that the current and previous HRC indices are consistently calculated.

In one embodiment, the method 200 further includes determining a HRC index difference between HRC index and a sequential HRC index, such as the previous HRC index or the next HRC index. For example, the HRC index difference is between the current HRC index and the previous HRC index k-i. The previous HRC index k-i may be associated with a basal state of the subject 1 10 and the current HRC index I _k may be associated with an active state of the subject 1 10.

The HRC index difference can be used to indicate or predict increased risks of clinical events to the subject 1 10, specifically if there is a decrease from the previous HRC index -i to the current HRC index I _k . The decrease may be a sudden drop or a gradual decrease. The HRC index difference is compared to the predefined reference data, specifically a predefined difference reference. As an example, the HRC index difference is indicative of the increased risks if it is at least 10%, i.e. the previous HRC index -i is at least 10% more than the current HRC index Ik. Said indication would alert the subject 1 10 of possible impending adverse clinical events, such as risk of cardiac arrest. Thus, in the step 214, the risk signal is generated in response to a determination from that the HRC index difference is indicative of increased risks of clinical events to the subject 1 10.

Therefore, various embodiments herein describe the calculation of the HRC index using the method 200 wherein the HRC index can be used as an indicator of increased risks of clinical events to the subject 1 10. The HRC index serves as a diagnostic modality that adds a new layer of information to quantify a subject’s health state. The HRC index can determine the activity level and the balance of the autonomic nervous system, and allows clinicians to look for derangements in the physiology of the subject 1 10 that may predict impending significant adverse cardiovascular events. For example, the HRC index may allow clinicians to detect worsening heart failure in patients before it is clinically evident by looking for derangements in the autonomic nervous system function so that there can be early intervention to avert further worsening.

The system 100 and method 200 can be implemented with different types of physiological sensing devices 120 for various industrial applications and for assessing risks of various clinical events.

In one example, the physiological sensing device 120 is an ICD and the method 200 can be performed for early detection of deterioration or derangement of the autonomic nervous system function that can forewarn of impending clinical events or deterioration, such as if there is a marked deterioration or change in the HRC index. Early detection allows for timely intervention to avert an episode of tachyarrhythmia or even a shock. Early detection may also help to reduce the incidence of inappropriate shocks by helping to detect atrial fibrillation which also markedly changes the HRC index.

The system 100 can be used to complement the capabilities of contemporary home monitoring systems which some patients may have. Clinicians may remotely access the home monitoring systems to monitor the patients’ health status and remotely detect any deterioration in the patients’ health status. If there is any deterioration, the clinicians can be alerted quickly to the need for early intervention, thereby further reducing shocks for the patients. In some other examples of the physiological sensing device 120, the physiological sensing device 120 may be a pacemaker and the method 200 can be performed for detection of worsening heart function. The physiological sensing device 120 may instead be an implantable loop recorder and the method 200 can be performed for detection of worsening heart failure in patients who do not already have an ICD, such as patients with diastolic heart failure who do not qualify for ICD implantation. In another example, the physiological sensing device 120 may instead be an external long-term ECG monitors and the method 200 can be performed for detection of worsening heart function and diagnosis of abnormal sinus node function. In these examples, the calculated HRC index is advantageously used in conjunction with the physiological data measured by the physiological sensing device 120 for risk assessment of impending clinical events, particularly in cardiac diseases.

Figure 5 is a block diagram illustrating a technical architecture of a computer 500 in accordance with embodiments of the present disclosure. Some non-limiting examples of the computer 500 are the electronic device 130, computer device 130a, remote server 130b, and computer device 140. The computer 500 includes a processor / central processing unit (CPU) 502, memory devices 504, a database 506, a data communication module 508, and a physiological data module 510.

The processor 502 executes instructions, codes, computer programs, and/or scripts which it accesses from the memory devices 504. The processor 502 includes suitable logic, circuitry, and/or interfaces to execute such operations or steps. Some non- limiting examples of the processor 502 include an application-specific integrated circuit (ASIC) processor, a reduced instruction set computing (RISC) processor, a complex instruction set computing (CISC) processor, a field-programmable gate array (FPGA), and the like. While only one processor 502 is shown, multiple processors may be present. Thus, while instructions may be discussed as executed by a processor 502, the instructions may be executed simultaneously, serially, or otherwise executed by one or multiple processors 502 (e.g. in a multi-core configuration). The memory devices 504 may comprise storage devices (such as flash memory, disk drives, or memory cards), read-only memory (ROM), and random-access memory (RAM). The memory devices 504 store non-transitory instructions operative by the processor 502 to perform various operations or steps of the method 200 according to various embodiments of the present disclosure. The memory devices 504 may be referred to as computer-readable storage media and/or non-transitory computer- readable media. Non-transitory computer-readable media include all computer- readable media, with the sole exception being a transitory propagating signal per se.

The database 506 is any computer-operated hardware suitable for storing and/or retrieving data. Some non-limiting examples of the database 506 are the databases 126, 132, and 142. The database 506 may include multiple storage units such as hard disks and/or solid-state disks in a Redundant Array of Independent Disks (RAID) configuration. The database 506 may include, but is not limited to, a storage area network (SAN) and/or a network attached storage (NAS) system. The data communication module 508 is configured for communication with other computers 500.

The physiological data module 510 is configured to process the physiological data, such as the heart rate data and motion data, and to perform various calculations based on the physiological data to obtain the HRC index in accordance with the method 200. Various algorithms may be implemented in the physiological data module 510 for performing said calculations based on the physiological data and to obtain the HRC index.

In the foregoing detailed description, embodiments of the present disclosure in relation to a system and method for assessing the risk of clinical events based on HRC are described with reference to the provided figures. The description of the various embodiments herein is not intended to call out or be limited only to specific or particular representations of the present disclosure, but merely to illustrate non-limiting examples of the present disclosure. The present disclosure serves to address at least one of the mentioned problems and issues associated with the prior art. Although only some embodiments of the present disclosure are disclosed herein, it will be apparent to a person having ordinary skill in the art in view of this disclosure that a variety of changes and/or modifications can be made to the disclosed embodiments without departing from the scope of the present disclosure. Therefore, the scope of the disclosure as well as the scope of the following claims is not limited to embodiments described herein.