System and method for detecting a persistent change in health

Abstract

A system and method for detecting a persistent change in health following an acute infection. A set of physiological parameters, acquired over at least a period prior to an acute infection and a period following the acute infection, are obtained for a subject, and a period during which the subject had the acute infection is identified. The set of physiological parameters are processed to determine whether the subject has a persistent change in health following the acute infection.

Description

[0001] 2024PF00609

[0002] 1

[0003] SYSTEM AND METHOD FOR DETECTING A PERSISTENT CHANGE IN HEALTH

[0004] FIELD OF THE INVENTION

[0005] The invention relates to the field of health monitoring and more particularly to systems and methods for detecting persistent changes in health following acute infections such as SARS-CoV-2 (COVID-19).

[0006] BACKGROUND OF THE INVENTION

[0007] Acute infections can be followed by a persistent change in health, even after the acute phase of the infection has resolved. For instance, Postural Orthostatic Tachycardia Syndrome (POTS) is a condition that can occur following an acute infection such as SARS-CoV-2 (COVID-19). POTS is characterized by an abnormal increase in heart rate upon standing, which can cause symptoms such as dizziness, lightheadedness, fainting, and fatigue. The COVID-19 pandemic has resulted in an increase in cases of POTS and other post-infection persistent changes in health, which are together known as “Long COVID” and include symptoms such as fatigue, brain fog, shortness of breath and other chronic health issues.

[0008] Similarly, various post-infection syndromes, such as chronic fatigue syndrome (CFS) and post-viral arthritis, are known to arise after infections like Lyme disease or bacterial sepsis. These syndromes often present with symptoms such as fatigue, brain fog, joint pain, and other chronic health issues, which can significantly impair quality of life.

[0009] These conditions are typically diagnosed using methods such as the tilt table test and the active stand test. These methods are time-consuming, expensive, and require specialized medical equipment. Further, as the symptoms of these conditions can often vary significantly throughout the day or from day to day, symptoms that occur outside of diagnostic tests may be unrecorded, leading to underdiagnosis or misdiagnosis, and thus to delayed intervention to treat the condition. Underdiagnosis can also occur due to low patient compliance, as existing diagnostic methods can be invasive and uncomfortable for patients.

[0010] SUMMARY OF THE INVENTION

[0011] The invention is defined by the claims.

[0012] According to examples in accordance with an aspect of the invention, there is provided a processing system for detecting a persistent change in health following an acute infection, the processing system being configured to: obtain a set of physiological parameters for a subject, the set of physiological parameters having been acquired over a period of time including a period prior to an acute infection and a2024PF00609

[0013] 2

[0014] period following the acute infection; identify a time of acute infection for the subject during the period over which the set of physiological parameters were acquired; and process at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection.

[0015] The inventors have recognized that physiological parameters acquired over a period of time including pre-infection and post-infection can be used to detect persistent changes in health following an acute infection more reliably than existing diagnostic methods. The physiological parameters are acquired over a much longer period of time than existing diagnostic tests cover, reducing a likelihood that symptoms are missed. Further, the physiological parameters may be acquired non-invasively and outside clinical settings, providing a more accessible and cost-effective method of detecting persistent changes in health, as fewer clinical visits are required and the need for expensive equipment and specialized personnel is reduced.

[0016] In some examples, the step of processing at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection comprises determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection.

[0017] In some examples, determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection comprises: identifying a baseline subset of the set of physiological parameters, wherein the baseline subset comprises data values acquired during the period prior to the acute infection; and for each physiological parameter, processing the data values for the physiological parameter in the baseline subset to define an expected healthy range for the physiological parameter.

[0018] In this way, the determination of whether the subject has a persistent change in health following the acute infection takes into account typical values for each physiological parameter for the subject before the infection occurred, tailoring the determination of whether the subject has a persistent change in health to the subject. The expected healthy range for the physiological parameter may, for example, be an interpercentile range of the data values for the physiological parameter in the baseline subset. For instance, the expected healthy range may be a 5th to 95th interpercentile range.

[0019] In some examples, determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection further comprises: identifying a post-infection subset of the set of physiological parameters, wherein the post-infection subset comprises data values acquired during the period following the acute infection; and for each physiological parameter, identifying any data values for the physiological parameter in the post-infection subset that fall outside the expected healthy range to determine an out-of-bound percent metric for the physiological parameter.

[0020] A higher out-of-bound percent metric indicates a higher likelihood that typical values for the physiological parameter have changed for the subject, compared to before the acute infection.2024PF00609

[0021] 3

[0022] In some examples, determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection further comprises: obtaining a null distribution for each physiological parameter, wherein the null distribution for each physiological parameter is based on physiological data from a plurality of healthy individuals; and for each physiological parameter, comparing the out-of-bound percent metric for the physiological parameter to the null distribution for the physiological parameter.

[0023] A comparison of the out-of-bound percent metric with the null distribution enables physiological parameters having an out-of-bound percent metric similar to healthy individuals to be identified, thus reducing a false positive rate of detecting a persistent change in health.

[0024] In some examples, the set of physiological parameters comprises data values acquired during a period of acute infection; and determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection further comprises: identifying an infection subset of the set of physiological parameters, wherein the infection subset comprises data values acquired during the period of acute infection; and for each physiological parameter: comparing the data values for the physiological parameter in the infection subset to the data values for the physiological parameter in the baseline subset to identify a direction of change of the physiological parameter as a result of the acute infection; and identifying whether any data values for the physiological parameter in the post-infection subset that fall outside the expected healthy range differ from the expected healthy range in the direction identified as the direction of change of the physiological parameter as a result of the acute infection.

[0025] Comparing the direction of change in each physiological parameter from pre-infection to post-infection with the direction of change from pre-infection to during the infection reduces the false positive rate of detecting a persistent change in health. For instance, this comparison enables a postinfection change to a physiological parameter due to an improvement in fitness to be distinguished from a persistent change in health resulting from the acute infection.

[0026] In some examples, determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection comprises: defining an ordered list of physiological parameters for the subject that demonstrate a persistent change post-infection; removing, from the ordered list, any physiological parameter for which a correlation with another physiological parameter ranked higher in the ordered list exceeds a predetermined correlation threshold; and determining, as the number of physiological parameters for the subject that demonstrate a persistent change post-infection, the number of physiological parameters in the ordered list following the removal of any physiological parameter for which a correlation with another physiological parameter ranked higher in the ordered list exceeds the predetermined correlation threshold.

[0027] This reduces double-counting of correlated physiological parameters, improving a reliability of and reducing a false positive rate of detecting a persistent change in health. The ordered list may, for example, be ordered according to an amount of data available for each physiological parameter (with the physiological parameter for which the greatest amount of data is available ranked highest).2024PF00609

[0028] 4

[0029] In some examples, the step of processing at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection further comprises: determining a maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance; and in response to the number of physiological parameters for the subject that demonstrate a persistent change post-infection being equal to or greater than the determined maximum number, determining that the subject has a persistent change in health following the acute infection.

[0030] Determining that a subject has a persistent change in health only if the number of physiological parameters demonstrating a persistent change is equal to or greater than the maximum number that could demonstrate a persistent change by chance reduces the false positive rate of detecting a persistent change in health.

[0031] In some examples, the processing system is configured to determine the maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance by: defining a desired false-positive rate; defining a probability distribution expressing a probability of each physiological parameter demonstrating a persistent change post-infection; and using the probability distribution to determine, as the maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance, a maximum number of physiological parameters that could demonstrate a persistent change post-infection while resulting in a total probability of no more than the desired false-positive rate.

[0032] In this way, the determination of whether the subject has a persistent change in health may be tailored to a desired false positive rate.

[0033] In some examples, the processing system is further configured to, in response to determining that the subject has a persistent change in health following the acute infection, control a feedback unit to provide a user-perceptible output indicating that a persistent change in health has been detected.

[0034] In this way, the subject or a clinician may be alerted to the persistent change in health, enabling the subject to access appropriate treatment.

[0035] In some examples, the set of physiological parameters for the subject was acquired using a wearable device.

[0036] A wearable device enables continuous acquisition of the set of physiological parameters over a period of several hours or days.

[0037] In some examples, the set of physiological parameters for the subject comprises one or more of: at least one heart rate parameter, at least one heart rate variability parameter, at least one temperature parameter, at least one respiratory rate parameter, at least one blood pressure parameter and / or at least one blood oxygen saturation level parameter.

[0038] In some examples, the acute infection is a viral infection or a bacterial infection. In some examples, the acute infection is SARS-CoV-2 (also known as CO VID-19).2024PF00609

[0039] 5

[0040] According to examples in accordance with another aspect of the invention, there is provided a computer-implemented method for detecting a persistent change in health following an acute infection, the computer-implemented method comprising: obtaining a set of physiological parameters for a subject, the set of physiological parameters having been acquired over a period of time including a period prior to an acute infection and a period following the acute infection; identifying a time of acute infection for the subject during the period over which the set of physiological parameters were acquired; and processing at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection.

[0041] In some examples, there is also provided a computer program product comprising computer program code means which, when executed on a computing device having a processing system, cause the processing system to perform all of the steps of the computer-implemented method described above.

[0042] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiment(s) described hereinafter.

[0043] BRIEF DESCRIPTION OF THE DRAWINGS

[0044] For a better understanding of the invention, and to show more clearly how it may be carried into effect, reference will now be made, by way of example only, to the accompanying drawings, in which:

[0045] Fig. 1 illustrates a system for detecting a persistent change in health following an acute infection, according to an embodiment of the invention;

[0046] Fig. 2 illustrates a graph showing the resting heart rate of a subject over time;

[0047] Fig. 3 illustrates an example method for determining whether a physiological parameter demonstrates a persistent change following an acute infection;

[0048] Fig. 4 illustrates an example method for determining whether a subject has a persistent change in health following an acute infection; and

[0049] Fig. 5 illustrates a computer-implemented method for detecting a persistent change in health following an acute infection, according to an embodiment of the invention.

[0050] DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The invention will be described with reference to the Figures.

[0052] It should be understood that the detailed description and specific examples, while indicating exemplary embodiments of the apparatus, systems and methods, are intended for purposes of illustration only and are not intended to limit the scope of the invention. These and other features, aspects, and advantages of the apparatus, systems and methods of the present invention will become better understood from the following description, appended claims, and accompanying drawings. It should be2024PF00609

[0053] 6

[0054] understood that the Figures are merely schematic and are not drawn to scale. It should also be understood that the same reference numerals are used throughout the Figures to indicate the same or similar parts.

[0055] The invention provides a system and method for detecting a persistent change in health following an acute infection. A set of physiological parameters, acquired over at least a period prior to an acute infection and a period following the acute infection, are obtained for a subject, and a period during which the subject had the acute infection is identified. The set of physiological parameters are processed to determine whether the subject has a persistent change in health following the acute infection.

[0056] Fig. 1 illustrates a system 100 for detecting a persistent change in health following an acute infection, according to an embodiment of the invention. The system 100 comprises a processing system 110 and a wearable device 120. The processing system 110 is, itself, an embodiment of the invention.

[0057] A persistent change in health following an acute infection may be any negative physiological change, experienced by a subject who has had an acute infection, that the subject continues to experience after recovery from the acute infection (e.g. for a period of weeks, months or even years). A persistent change in health following an acute infection may indicate that a subject may be experiencing a post-infection syndrome (such as Postural Orthostatic Tachycardia Syndrome, chronic fatigue syndrome or post-viral arthritis). Thus, the system 100 may be used to alert a subject or their clinician to the possibility that the subject may be suffering from a post-infection syndrome.

[0058] The acute infection may be any infection with the potential to cause persistent changes in physiological parameters for which symptoms are typically experienced for a short period, e.g. a period of 1-14 days. In some examples, the acute infection may be a viral infection (e.g. SARS-CoV-2, influenza, dengue, mononucleosis, etc.). In some examples, the acute infection may be a bacterial infection (e.g. pneumonia, Lyme disease, sepsis, etc.). However, it is to be understood that that present disclosure is not limited to persistent changes in health following a viral or bacterial infection, and the system 100 may be used to detect a persistent change in health following any acute infection with the potential to cause persistent changes in physiological parameters capable of being monitored using a wearable device.

[0059] The processing system 110 is configured to obtain a set of physiological parameters for a subject. The physiological parameters are acquired over a period of time including a period prior to an acute infection and a period following the acute infection. In other words, each physiological parameter in the set of physiological parameters includes data values acquired during the period prior to the acute infection, and data values acquired during the period following the acute infection. In some examples, the set of physiological parameters may also include data values acquired during a period of acute infection.

[0060] The processing system 110 may be configured to determine whether the subject has a persistent change in health only if the available data values meet predetermined criteria. For example, the predetermined criteria may require that the data values acquired during the period prior to the acute infection span at least a first predefined minimum period that ends by a first predefined time before the time of acute infection. The first predefined minimum period and the first predefined time before the time2024PF00609

[0061] 7

[0062] of acute infection may depend on the acute infection experienced by the subject. For instance, in the case of a SARS-CoV-2 infection, the predetermined criteria may require that the data values acquired during the period prior to the acute infection span a period of at least 28 days ending at least 14 days before the time of acute infection (e.g. the date on which the subject first tests positive for SARS-CoV-2). Similarly, the predetermined criteria may require that the data values acquired during the period following the acute infection span at least a second predefined minimum period that starts no earlier than a second predefined time after the time of acute infection. The second predefined minimum period and the second predefined time after the time of acute infection may depend on the acute infection experienced by the subject. For instance, in the case of a SARS-CoV-2 infection, the predetermined criteria may require that the data values acquired during the period following the acute infection span a period of at least 28 days (e.g. a period of between 28 days and 12 weeks) starting at least 28 days after the time of acute infection. Where the set of physiological parameters includes data values acquired during a period of acute infection, the period of acute infection may span a predefined period including the time of acute infection. The predefined period may depend on the acute infection experienced by the subject. For instance, in the case of a SARS-CoV-2 infection, the period of acute infection may span a period of 14 days, e.g. starting 7 days before the date on which the subject first tests positive for SARS-CoV-2 and ending 7 days after the date on which the subject first tests positive for SARS-CoV-2. Suitable periods prior to the acute infection, during the acute infection and following the acute infection for other acute infections may be predefined based on clinical guidelines.

[0063] In some examples, the predetermined criteria may include at least one compliance criterion. For instance, the predetermined criteria may require that data was acquired for at least a minimum percentage of the dates in each of the period prior to the acute infection and the period following the acute infection (and, in some examples, in the period of acute infection). The minimum percentage of dates may depend on the type of acute infection. In some examples, the minimum percentage of dates may be different for different physiological parameters, e.g. depending on a stability of the physiological parameter and the data reliability. For instance, a higher minimum percentage of dates may be used for a physiological parameter with a high variability (e.g. heart rate), while a lower minimum percentage of dates may be used for a more stable physiological parameter (e.g. blood oxygen saturation level). In some examples, the minimum percentage of dates may have a value between 50% and 70% (e.g. 60%).

[0064] In some examples, the set of physiological parameters may include only data values acquired during periods of sleep of the subject, as acquiring physiological data during sleep typically provides stable and high-quality measurements. In Fig. 1, the physiological parameters are obtained by receiving physiological data 125 from the wearable device 120, and processing the physiological data to obtain the set of physiological parameters. In other examples, the processing system 110 may obtain the set of physiological parameters by receiving the set of physiological parameters directly from the wearable device.2024PF00609

[0065] 8

[0066] In some examples, the physiological parameters may be acquired by using the wearable device to monitor the subject over a continuous period of time that includes the period prior to the acute infection, the period of acute infection and the period following the acute infection. In other examples, the wearable device may be configured to periodically (e.g. every few months or once a year) acquire physiological data, each time acquiring the physiological data for at least the first predefined minimum period, to obtain data values for the physiological parameters in the period prior to the acute infection. This periodic acquisition of physiological data accounts for any changes over time to the physiological parameters for the subject in a healthy state. In response to the subject experiencing an acute infection, physiological data may be acquired from the second predefined time after the time of acute infection to obtain data values for the physiological parameters in the period following the acute infection (for instance, in response to the processing system 110 identifying a time of acute infection, as described below, the processing system may control the wearable device to acquire data for at least a second predefined minimum period from the second predefined time after the time of acute infection).

[0067] The device(s) used to obtain the set of physiological parameters may be capable of measuring one or more of: a heart rate, a heart rate variability, a temperature, a respiratory rate, a blood pressure and / or a blood oxygen saturation level. Thus, the set of parameters may comprise one or more of: at least one heart rate parameter, at least one heart rate variability parameter, at least one temperature parameter, at least one respiratory rate parameter, at least one blood pressure parameter and / or at least one blood oxygen saturation level parameter. In some examples, the set of parameters may comprise at least one heart rate parameter. In some examples, the set of parameters may comprise at least one heart rate parameter and at least one heart rate variability parameter.

[0068] The at least one heart rate parameter may comprise one or more of: a minimum heart rate, a maximum heart rate, a mean heart rate, a standard deviation of heart rate, a skewness of heart rate, a kurtosis of heart rate, a hyperskewness of heart rate, a hypertailedness of heart rate, a 1st percentile of heart rate, a 5th percentile of heart rate, a 10th percentile of heart rate, a 20th percentile of heart rate, a 25th percentile of heart rate, a 30th percentile of heart rate, a 40th percentile of heart rate, a 50th percentile of heart rate, a 60th percentile of heart rate, a 70th percentile of heart rate, a 75th percentile of heart rate, an 80th percentile of heart rate, a 90th percentile of heart rate, a 95th percentile of heart rate and / or a 99th percentile of heart rate, each determined per day or per sleep session.

[0069] The at least one heart rate variability parameter may comprise one or more of: a minimum heart rate variability, a maximum heart rate variability, a mean heart rate variability, a standard deviation of heart rate variability, a skewness of heart rate variability, a kurtosis of heart rate variability, a hyperskewness of heart rate variability, a hypertailedness of heart rate variability, a 1st percentile of heart rate variability, a 5th percentile of heart rate variability, a 10th percentile of heart rate variability, a 20th percentile of heart rate variability, a 25th percentile of heart rate variability, a 30th percentile of heart rate variability, a 40th percentile of heart rate variability, a 50th percentile of heart rate variability, a 60th2024PF00609

[0070] 9

[0071] percentile of heart rate variability, a 70th percentile of heart rate variability, a 75th percentile of heart rate variability, an 80th percentile of heart rate variability, a 90th percentile of heart rate variability, a 95th percentile of heart rate variability and / or a 99th percentile of heart rate variability, each determined per day or per sleep session.

[0072] The at least one temperature parameter may comprise one or more of: a minimum temperature, a maximum temperature, a mean temperature, a standard deviation of temperature, a skewness of temperature, a kurtosis of temperature, a hyperskewness of temperature, a hypertailedness of temperature, a 1st percentile of temperature, a 5th percentile of temperature, a 10th percentile of temperature, a 20th percentile of temperature, a 25th percentile of temperature, a 30th percentile of temperature, a 40th percentile of temperature, a 50th percentile of temperature, a 60th percentile of temperature, a 70th percentile of temperature, a 75th percentile of temperature, an 80th percentile of temperature, a 90th percentile of temperature, a 95th percentile of temperature and / or a 99th percentile of temperature, each determined per day or per sleep session.

[0073] The at least one respiratory rate parameter may comprise one or more of: a minimum respiratory rate, a maximum respiratory rate, a mean respiratory rate, a standard deviation of respiratory rate, a skewness of respiratory rate, a kurtosis of respiratory rate, a hyperskewness of respiratory rate, a hypertailedness of respiratory rate, a 1st percentile of respiratory rate, a 5th percentile of respiratory rate, a 10th percentile of respiratory rate, a 20th percentile of respiratory rate, a 25th percentile of respiratory rate, a 30th percentile of respiratory rate, a 40th percentile of respiratory rate, a 50th percentile of respiratory rate, a 60th percentile of respiratory rate, a 70th percentile of respiratory rate, a 75th percentile of respiratory rate, an 80th percentile of respiratory rate, a 90th percentile of respiratory rate, a 95th percentile of respiratory rate and / or a 99th percentile of respiratory rate, each determined per day or per sleep session.

[0074] The at least one blood pressure parameter may comprise one or more of: a minimum blood pressure, a maximum blood pressure, a mean blood pressure, a standard deviation of blood pressure, a skewness of blood pressure, a kurtosis of blood pressure, a hyperskewness of blood pressure, a hypertailedness of blood pressure, a 1st percentile of blood pressure, a 5th percentile of blood pressure, a 10th percentile of blood pressure, a 20th percentile of blood pressure, a 25th percentile of blood pressure, a 30th percentile of blood pressure, a 40th percentile of blood pressure, a 50th percentile of blood pressure, a 60th percentile of blood pressure, a 70th percentile of blood pressure, a 75th percentile of blood pressure, an 80th percentile of blood pressure, a 90th percentile of blood pressure, a 95th percentile of blood pressure and / or a 99th percentile of blood pressure, each determined per day or per sleep session.

[0075] The at least one blood oxygen saturation level parameter may comprise one or more of: a minimum blood oxygen saturation level, a maximum blood oxygen saturation level, a mean blood oxygen saturation level, a standard deviation of blood oxygen saturation level, a skewness of blood oxygen saturation level, a kurtosis of blood oxygen saturation level, a hyperskewness of blood oxygen saturation level, a hypertailedness of blood oxygen saturation level, a 1st percentile of blood oxygen saturation2024PF00609

[0076] 10

[0077] level, a 5th percentile of blood oxygen saturation level, a 10th percentile of blood oxygen saturation level, a 20th percentile of blood oxygen saturation level, a 25th percentile of blood oxygen saturation level, a 30th percentile of blood oxygen saturation level, a 40th percentile of blood oxygen saturation level, a 50th percentile of blood oxygen saturation level, a 60th percentile of blood oxygen saturation level, a 70th percentile of blood oxygen saturation level, a 75th percentile of blood oxygen saturation level, an 80th percentile of blood oxygen saturation level, a 90th percentile of blood oxygen saturation level, a 95th percentile of blood oxygen saturation level and / or a 99th percentile of blood oxygen saturation level, each determined per day or per sleep session.

[0078] In some examples, data values for each physiological parameter may only be determined for days / sleep sessions for which a minimum amount of data is available.

[0079] In Fig. 1, the wearable device 120 is a smartwatch; however, the set of physiological parameters may be obtained using any device or devices capable of acquiring physiological data, including other wearable devices, such as a smart ring or a patch, and any suitable non-wearable device.

[0080] In some examples, more than one device may be used to obtain the set of physiological parameters. For instance, a first subset of physiological data for the subject may be acquired by a first device and a second subset of physiological data for the subject may be acquired by a second device. One or more types of physiological data may belong to both the first subset and the second subset. For example, a smartwatch may be used to measure heart rate and respiratory rate, and a smart ring may be used to measure temperature, heart rate and heart rate variability; the heart rate measurements acquired by the smartwatch may then be combined with the heart rate measurements acquired by the smart ring to obtain at least one heart rate parameter.

[0081] The processing system 110 is further configured to identify a time of acute infection for the subject during the period over which the set of physiological parameters were acquired. Any suitable technique may be used to identify the time of acute infection. For example, the time of acute infection may be based on a user input indicating a date on which the subject tested positive for a particular acute infection or received a clinical diagnosis of the acute infection. In other examples, an infection prediction model may be used to identify the time of acute infection. For instance, the processing system 110 may be configured to provide the set of physiological parameters (or the physiological data used to obtain the set of physiological parameters) to an infection prediction model. The infection prediction model may then process the set of physiological parameters (or the physiological data) to determine the time of acute infection, and provide the time of acute infection to the processing system 110.

[0082] The processing system 110 is then configured to process at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection. The determination as to whether the subject has a persistent change in health following the acute infection may comprise a comparison between the physiological parameters in the period prior to the acute infection with the physiological parameters in the period following the acute infection.2024PF00609

[0083] 11

[0084] The inventors have recognized that a persistent change in the health of a subject following an acute infection may result in or correlate with a persistent change in physiological parameters for the subject.

[0085] An example of this is shown in Fig. 2, which illustrates a graph 200 showing the resting heart rate of a subject over time. The period of time shown on graph 200 includes a period prior to a SARS-CoV-2 infection (prior to ti ), a period of infection (between ti and t2) and a period following the SARS-CoV-2 infection (after t2). The graph 200 shows a persistent change in the resting heart rate of the subject in the period following the SARS-CoV-2 infection compared to the period prior to the SARS-CoV-2 infection.

[0086] In some examples, the processing system 110 may be configured to determine whether the subject has a persistent change in health following the acute infection by determining, for each physiological parameter, whether the physiological parameter demonstrates a persistent change following the acute infection.

[0087] Fig. 3 illustrates an example method 300 for determining whether a physiological parameter demonstrates a persistent change following the acute infection. The method 300 may be performed for each physiological parameter in the set of physiological parameters.

[0088] The method 300 begins at step 310, at which an expected healthy range is defined for the physiological parameter. The expected healthy range for the physiological parameter may be defined using a baseline subset of the set of physiological parameters that comprises data values acquired during the period prior to the acute infection, by processing the data values for the physiological parameter in the baseline subset to define the expected healthy range for the physiological parameter. The processing system 100 may, for example, be configured to identify a baseline subset (e.g. containing data values for all physiological parameters in the set) before performing the method 300 for each physiological parameter, or the step 310 of identifying the expected healthy range for the physiological parameter may comprise a sub-step of identifying a baseline subset for the physiological parameter.

[0089] The baseline subset may, for example, comprise all data values acquired during the period prior to the acute infection for all physiological parameters in the set, all data values acquired during the period prior to the acute infection for the physiological parameter for which the method 300 is being performed, all data values acquired during the period prior to the acute infection for all physiological parameters in the set that meet one or more conditions, or all data values acquired during the period prior to the acute infection for the physiological parameter for which the method 300 is being performed that meet one or more conditions. The one or more conditions may, for example, include a compliance condition; for instance, a data value may only be included in the baseline subset if physiological data relating to the physiological parameter was acquired for a least a predefined minimum period on the date on which the data value was acquired (if this was not already taken into account when obtaining the set of physiological parameters).2024PF00609

[0090] 12

[0091] The expected healthy range may, for example, be an interpercentile range of the data values in the baseline subset. In some examples, the expected healthy range may be a 5th to 95th interpercentile range; however, any other interpercentile range may be used, depending on a desired balance between specificity and sensitivity. For instance, the expected healthy range may be the interdecile range or the interquartile range.

[0092] At step 320, an out-of-bound percent metric is determined for the physiological parameter. The out-of-bound percent metric may be a percentage of data values acquired in the period following the acute infection that fall outside the expected healthy range for the physiological parameter. The out-of-bound percent metric may be determined using a post-infection subset of the set of physiological parameters that comprises data values acquired in the period following the acute infection, by identifying any data values for the physiological parameter in the post-infection subset that fall outside the expected healthy range to determine the percentage of data values for the physiological parameter in the post-infection subset that fall outside the expected healthy range. The processing system 100 may, for example, be configured to identify a post-infection subset (e.g. containing data values for all physiological parameters in the set) before performing the method 300 for each physiological parameter, or the step 320 of determining the out-of-bound percent metric for the physiological parameter may comprise a sub-step of identifying a post-infection subset for the physiological parameter.

[0093] The post-infection subset may, for example, comprise all data values acquired in the period following the acute infection for all physiological parameters in the set, all data values acquired in the period following the acute infection for the physiological parameter for which the method 300 is being performed, all data values acquired in the period following the acute infection for all physiological parameters in the set that meet one or more conditions, or all data values acquired in the period following the acute infection for the physiological parameter for which the method 300 is being performed that meet one or more conditions. The one or more conditions may, for example, include a compliance condition; for instance, a data value may only be included in the post-infection subset if physiological data relating to the physiological parameter was acquired for a least a predefined minimum period on the date on which the data value was acquired (if this was not already taken into account when obtaining the set of physiological parameters).

[0094] At step 330, the out-of-bound percent metric is compared to a threshold value. In some examples, the threshold value may have a predefined value, which may depend on a desired balance between specificity and sensitivity. In some examples, the predefined value may be between 85% and 95%; for instance, the threshold value may be 90%.

[0095] In some examples, the value of the threshold value may be determined using a null distribution for the physiological parameter (i.e. the method 300 may further comprise a step of obtaining a null distribution for the physiological parameter and a step of processing the null distribution to determine the threshold value). The null distribution for the physiological parameter comprises data values for the physiological parameter acquired from a plurality of healthy individuals. It is to be2024PF00609

[0096] 13

[0097] understood that, in the context of the present disclosure, a “healthy individual” is an individual who has not suffered from the acute infection at any point during the period over which the data values for the individual included in the null distribution were acquired, and who has tested negative for the acute infection during the period over which the data values for the individual included in the null distribution were acquired, and who has remained free from a predefined set of symptoms throughout this period. Thus, for each healthy individual, the null distribution includes data values acquired prior to testing negative for the acute infection and data values acquired following testing negative for the acute infection. Preferably, the data values acquired prior to testing negative for the acute infection span at least the first predefined minimum period, and the data values acquired following testing negative for the acute infection span at least the second predefined minimum period. The null distribution for the physiological parameter may be acquired under the same or similar conditions as the acquisition of the physiological parameter for the subject; for instance, if the physiological parameter for the subject is based on physiological data acquired during sleep, the physiological parameter for each healthy individual may be based on physiological data acquired during sleep.

[0098] The threshold value may, for example, be determined by determining an out-of-bound percent metric for each healthy individual represented in the null distribution. An out-of-bound percent metric for each healthy individual may be determined by processing the data values acquired prior to testing negative for the acute infection to define an expected healthy range for the individual, using the same method as used to define the expected healthy range for the subject, and determining the percentage of data values acquired following testing negative for the acute infection that fall outside the expected healthy range for the individual.

[0099] A predetermined percentile of the out-of-bound percent metric for the null distribution (i.e. across all healthy individuals in the null distribution) may then be identified; this value may be used as the threshold value. The predetermined percentile may, for example, be at least the 90th percentile; for instance, the predetermined percentile may be the 95th percentile.

[0100] In some examples, the threshold value may be determined using both a predefined value and a null distribution. For instance, the threshold value may be determined as the higher of a predefined value and a predetermined percentile of the out-of-bound percent metric for the null distribution (e.g. the threshold value may be the higher of 90% and the 95th percentile of the out-of-bound percent metric for the null distribution). In some examples, a comparison with the predetermined percentile of the out-of-bound percent metric for the null distribution may be made only if the out-of-bound percent metric for the physiological parameter for the subject exceeds a predefined value (e.g. the out-of-bound percent metric for the physiological parameter for the subject may be compared to the 95th percentile of the out-of-bound percent metric for the null distribution only if the out-of-bound percent metric for the physiological parameter for the subject is greater than 90%).2024PF00609

[0101] 14

[0102] In response to the out-of-bound percent metric failing to exceed the threshold value, it is determined that the physiological parameter does not demonstrate a persistent change following the acute infection.

[0103] In response to the out-of-bound percent metric exceeding the threshold value, the method proceeds to step 340, at which an expected direction of change for the physiological parameter is identified. The expected direction of change of the physiological parameter is the direction in which the physiological parameter is expected to change following the acute infection, compared to the period prior to the acute infection, if the subject experiences a persistent change in health following the acute infection (i.e. whether the change in the physiological parameter is expected to be an increase or a decrease compared with the period prior to the acute infection).

[0104] In some examples, where the set of physiological parameters includes data values acquired during the period of acute infection, the expected direction of change may be identified using the data values acquired during the period of acute infection. This allows an expected direction of change to be identified regardless of whether the effect of the acute infection on the physiological parameter is known (or whether the effect of the acute infection on the physiological parameter is consistent between different subjects).

[0105] In other words, the expected direction of change may be identified by identifying an infection subset of the set of physiological parameters comprising data values acquired during the period of acute infection, and comparing the data values for the physiological parameter in the infection subset to the data values for the physiological parameter in the baseline subset to identify, as the expected direction of change, a direction of change of the physiological parameter as a result of the acute infection.

[0106] The processing system 100 may, for example, be configured to identify an infection subset (e.g. containing data values for all physiological parameters in the set) before performing the method 300 for each physiological parameter, or the step 340 of identifying an expected direction of change for the physiological parameter may comprise a sub-step of identifying an infection subset for the physiological parameter.

[0107] The infection subset may, for example, comprise all data values acquired in the period of acute infection for all physiological parameters in the set, all data values acquired in the period of acute infection for the physiological parameter for which the method 300 is being performed, all data values acquired in the period of acute infection for all physiological parameters in the set that meet one or more conditions, or all data values acquired in the period of acute infection for the physiological parameter for which the method 300 is being performed that meet one or more conditions. The one or more conditions may, for example, include a compliance condition; for instance, a data value may only be included in the infection subset if physiological data relating to the physiological parameter was acquired for a least a predefined minimum period on the date on which the data value was acquired (if this was not already taken into account when obtaining the set of physiological parameters).2024PF00609

[0108] 15

[0109] Any suitable comparison between the data values for the physiological parameter in the infection subset and the data values for the physiological parameter in the baseline subset may be used to identify the expected direction of change. In some examples, the comparison may comprise a comparison of the mean value for the physiological parameter in the infection subset with the mean value for the physiological parameter in the baseline subset. In some examples, the expected direction of change may be identified using a measure of effect size. For instance, the expected direction of change may be identified by computing an adjusted deflection, e.g. using Equation 1:

[0110]

[0111] where Meaninfectionis the mean value for the physiological parameter in the infection subset, MeanbaseUneis the mean value for the physiological parameter in the baseline subset, oinfectionis the standard deviation for the physiological parameter in the infection subset, and O'baseline is the standard deviation for the physiological parameter in the baseline subset. A positive adjusted deflection value indicates an increase in the physiological parameter during the period of acute infection, compared to the period prior to the acute infection, while a negative adjusted deflection value indicates a decrease in the physiological parameter.

[0112] In other examples, if the effect of the acute infection on the physiological parameter is known, the expected direction of change of the physiological parameter may be predefined. The expected direction of change of the physiological parameter may be predefined using prior clinical knowledge of the effect of the acute infection. Alternatively, the expected direction of change of the physiological parameter may be predefined using physiological data from a population of positive cases for the acute infection (i.e. using the technique as described above).

[0113] At step 350, it is determined whether the change in the physiological parameter postinfection is in the same direction as the expected direction of change. This may, for example, be determined by identifying the direction of change for the physiological parameter in the period following the acute infection, compared with the period prior to the acute infection, for instance, by computing the adjusted deflection as described above. In some examples, it may be determined that the change in the physiological parameter post-infection is in the same direction as the expected direction of change in response to a predefined percentage (e.g. at least 50%) of the data values in the period following the acute infection that fall outside the expected healthy range having the expected direction of change. For instance, if the expected direction of change is an increase, it may be determined that the change in the physiological parameter post-infection is in the same direction as the expected direction of change in response to a predefined percentage of the data values in the period following the acute infection that fall outside the expected healthy range being greater than the expected healthy range.2024PF00609

[0114] 16

[0115] In response to the direction of change of the physiological parameter post-infection differing from the expected direction of change, it is determined that the physiological parameter does not demonstrate a persistent change following the acute infection.

[0116] In response to the direction of change of the physiological parameter post-infection corresponding to the expected direction of change, it is determined that the physiological parameter demonstrates a persistent change following the acute infection.

[0117] Thus, by performing the method 300 for each physiological parameter in the set of physiological parameters, it may be determined, for each physiological parameter, whether the physiological parameter does or does not demonstrate a persistent change following the acute infection. It is to be understood that the steps of the method 300 are not necessarily performed in the order shown in Fig. 3; for instance, the direction of change of the physiological parameter post-infection may be compared to the expected direction of change prior to comparing the out-of-bound percent metric to a threshold value (and the comparison of the out-of-bound percent metric with the threshold value may be performed only if the direction of change of the physiological parameter post-infection corresponds to the expected direction of change).

[0118] Any other suitable method may be used to determine, for each physiological parameter, whether the physiological parameter does or does not demonstrate a persistent change following the acute infection. For instance, in some examples, only a comparison of the out-of-bound percent metric with a threshold value may be used to identify physiological parameters demonstrating a persistent change following the acute infection (i.e. the comparison of direction of change of the physiological parameter post-infection with the expected direction of change may be omitted), while in other examples, only a comparison of direction of change of the physiological parameter post-infection with the expected direction of change may be used to identify physiological parameters demonstrating a persistent change following the acute infection (i.e. the comparison of the out-of-bound percent metric with a threshold value may be omitted). In yet other examples, both the comparison of the out-of-bound percent metric with a threshold value and the comparison of direction of change of the physiological parameter postinfection with the expected direction of change may be performed for the physiological parameter, and it may be determined that the physiological parameter demonstrates a persistent change in response to at least one of: the out-of-bound percent metric exceeding the threshold value; and / or the direction of change of the physiological parameter post-infection corresponding to the expected direction of change (i.e. a persistent change in the physiological parameter may be detected if only one of these conditions is met). The method used to determine whether the physiological parameter does or does not demonstrate a persistent change following the acute infection may depend on a desired balance between specificity and sensitivity. In yet other examples, a mean value for the physiological parameter in the baseline subset may be compared with a mean value for the physiological parameter in the post-infection subset to determine whether the physiological parameter does or does not demonstrate a persistent change following the acute infection; for instance, it may be determined that the physiological parameter2024PF00609

[0119] 17

[0120] demonstrates a persistent change if difference between the mean value for the physiological parameter in the baseline subset and the mean value for the physiological parameter in the post-infection subset exceeds a threshold.

[0121] Whether the physiological does or does not demonstrate a persistent change following the acute infection may be used to determine whether the subject has a persistent change in health. For instance, having performed the method 300 for each physiological parameter in the set of physiological parameters, the processing system 110 may be configured to determine whether the subject has a persistent change in health following the acute infection by carrying out the method 400 illustrated in Fig.

[0122] 4.

[0123] The method 400 begins at step 410, at which the number S of physiological parameters demonstrating persistent change post-infection is determined.

[0124] In some examples, the number S of physiological parameters demonstrating a persistent change post-infection may simply be the total number of physiological parameters for which it has been determined that a persistent change is demonstrated (e.g. by performing the method 300 described above with reference to Fig. 3).

[0125] In other examples, the step 410 of determining the number S of physiological parameters demonstrating persistent change post-infection may comprise a filtering step to reduce double-counting between highly correlated physiological parameters and / or to remove physiological parameters for which the number of data values available is low.

[0126] For instance, the step 410 of determining the number S of physiological parameters demonstrating a persistent change post-infection may comprise defining an ordered list of physiological parameters for the subject that demonstrate a persistent change in the period following the acute infection. The ordered list may comprise all physiological parameters that have been found to demonstrate a persistent change following the acute infection (e.g. using the method 300).

[0127] The ordered list may, for example, be ordered according to the amount of data available for each physiological parameter, with a physiological parameter for which more data values are available being ranked higher on the ordered list. In some examples, other factors may additionally or alternatively be used to order the list. For instance, physiological parameters known to be particularly affected by acute infection may be prioritized in the ordered list; for example, in the case of SARS-CoV-2, heart rate and heart rate variability parameters may be prioritized in the ordered list, even if these parameters have low amounts of available data.

[0128] Having defined the ordered list, a correlation may be determined for each physiological parameter on the ordered list with each other physiological parameter on the ordered list. Any physiological parameter for which a correlation with another physiological parameter ranked higher in the ordered list exceeds a predetermined correlation threshold may then be removed from the ordered list. In other words, for a pair of highly correlated physiological parameters, the lower ranking parameter of the2024PF00609

[0129] 18

[0130] pair is removed from the ordered list. In this way, the effect of double counting is reduced, while ensuring that the fdtered list includes as much useful data as possible.

[0131] Any suitable measure of correlation may be used to determine the correlation between physiological parameters on the ordered list. In some examples, the measure of correlation may be a Spearman correlation, in order to identify any physiological parameters having a monotonic relationship with one another. The predetermined correlation threshold may depend on the measure of correlation used. In some examples, the predetermined correlation threshold may have a value of at least 0.3, preferably, at least 0.5; for example, a physiological parameter may be removed from the ordered list in response to having a Spearman correlation with a higher ranked physiological parameter of 0.8 or higher.

[0132] The number of physiological parameters in the ordered list following the removal of any physiological parameter for which a correlation with another physiological parameter ranked higher in the ordered list exceeds the predetermined correlation threshold (i.e. the number of physiological parameters in the fdtered list) may then be used as the number S of physiological parameters demonstrating persistent change post-infection. Alternatively (e.g. if compliance was not taken into account when obtaining the set of physiological parameters), an additional fdtering step may be performed to remove any physiological parameter for which the number of data values (or for which the number of dates for which a minimum number of data values is available) fails to exceed a predetermined compliance threshold; the number of physiological parameters in the ordered list following this additional fdtering step may then be used as the number S of physiological parameters demonstrating persistent change post-infection.

[0133] At step 420, a threshold number K of physiological parameters may be determined. The threshold number K may represent a maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance. The maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance may, for example, be determined by defining a desired false positive rate, defining a probability distribution expressing a probability of each physiological parameter demonstrating a persistent change postinfection, and using the probability distribution to determine, as the maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance, a maximum number of physiological parameters that could demonstrate a persistent change post-infection while resulting in a total probability of no more than the desired false-positive rate.

[0134] The desired false-positive rate may be predefined or defined via a user input. In some examples, the desired false-positive rate may have a predefined value of 5%.

[0135] Any suitable probability distribution may be used to express the probability of each physiological parameter demonstrating a persistent change post-infection. For instance, a binomial distribution may be used to express the probability, as set out in Equation 2:

[0136]

[0137] 2024PF00609

[0138] where F is the total number of physiological parameters in the set (or the total number of physiological parameters for which a minimum amount of data is available, if compliance is not taken into account at the stage of obtaining the set of physiological parameters), S is the number of physiological parameters demonstrating a persistent change post-infection, p is the desired false-positive rate, and P (S | F, p) is the probability that S out of the F physiological parameters demonstrates a persistent change given a false-positive rate p .

[0139] The maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance (and thus the threshold number K) could then, for a desired false-positive rate of 5%, be determined using Equation 3:

[0140] P(X < K\F,p) < 0.05 (3)

[0141] In other words, the threshold number K may be calculated as:

[0142] Ko.os =

[0143]

[0144] 0.05 (4)

[0145] At step 430, the number S of physiological parameters demonstrating persistent change post-infection may be compared to the threshold number K.

[0146] In response to the number S of physiological parameters demonstrating persistent change post-infection being greater than or equal to the threshold number K, it may be determined that the subject has a persistent change in health following the acute infection.

[0147] In response to the number S of physiological parameters demonstrating persistent change post-infection being less than the threshold number K, it may be determined that the subject does not have a persistent change in health following the acute infection.

[0148] In an exemplary embodiment of the SARS-CoV-2 infection, the system established three predetermined periods with respect to the test date: the first period of a baseline window defined from day -42, wherein minus denotes the time span before the acute infection, to day -14, the second period of a testing window from day -7 to day +6, wherein “+” denotes the time span after the acute infection, and a third period of a study window from day +28 to day +56. Physiological monitoring during sleep was used to improve signal quality and stability, and 34 base physiological parameters were aggregated by 23 nightly statistics, yielding 782 per-night features (the set of physiological parameters with associated nightly statistics). Compliance was enforced per-parameter by requiring at least 60% nightly coverage in both baseline and study windows (approximately 17 of 28 nights in each window), which ensures that2024PF00609

[0149] 20

[0150] downstream statistics are computed over adequate data. The system quantified infection-acquired directional deflection by comparing testing and baseline windows using an adjusted standardized effect size, where an increase or decrease was inferred for each feature. Further, computing the adjusted deflection using Equation 1 and assigning a feature direction (expected direction of the physiological parameter change in the given statistical context) sign per feature based on cohort-level trends. In the study window, the system built individualized 90% confidence intervals from the baseline for each available feature and computed an out-of-bound percentage, requiring that more than 90% of nightly values fall outside the interval in the direction consistent with the feature direction learned from the acute period. To ensure observed deviations exceeding natural variation, the out-of-bound rate for each feature was compared against a null distribution from a curated SARS-negative cohort that remained asymptomatic in baseline, testing, and study windows; features exceeding the 95th percentile of the null were retained. To reduce dependence between features and control the number of tests, the retained set in this example was pruned by iteratively eliminating co-triggered features (physiological parameters) with a population Spearman correlation of 0.8 or higher, thereby producing a subset of approximately independent detectors. For each subject, the number of available independent features F and the number triggered S were evaluated against a binomial model with p=0.05 to compute the maximal K in accordance to Equation 3. A user was classified as exhibiting persistent physiological change when S>K, which sets a subject-level threshold tied to the number of available tests while preserving a 5% familywise false positive rate under the null. C

[0151] Concrete evidence from the disclosed embodiment demonstrated a prevalence of persistent changes in 33 of 349 SARS-positive subjects who met the 60% per-parameter compliance threshold, corresponding to 9.4%. A Monte Carlo analysis using 1,000 resamplings of SARS-negative cohorts of matched size yielded, on average, 270.7 compliant subjects and 5.3 detections, corresponding to an average prevalence of 1.9% with a standard deviation of 0.7%, which substantiates that the 9.4% rate in positives exceeds what is expected by chance given the null distribution. The used dataset comprised 12,698 enrolled participants, of whom 4,614 had at least one SARS test, with 760 positives and 3,854 negatives; after exclusions for baseline symptoms and compliance, 663 SARS-positive and 2,513 SARS-negative users formed the analytic base for the broader analysis, with 349 SARS-positive subjects and approximately 270 SARS-negative subjects per simulation satisfying the 60% compliance requirement across windows and features.

[0152] The results of the method practiced in the disclosed above embodiment was supported by repeated physiological patterns in detected subjects. Across the 33 detected cases, an elevation in nightly resting heart rate metrics persisted for multiple months relative to individualized baselines. On average, the resting heart rate was elevated by about 7 beats per minute (BPM), representing approximately a 13% increase over the averaged population baseline, and heart rate variability (HRV) decreased with nightly Root Mean Square of Successive Differences (RMSSD) dropping from approximately 65 ms to 45 ms. Directional consistency across infection and study windows was validated by infection-acquired2024PF00609

[0153] 21

[0154] deflection analyses showing that upper-tail temperature and higher percentiles of respiratory rate increased significantly during the acute window in SARS-positive subjects relative to negatives, confirming the expected febrile and tachypneic responses to infection. Feature recurrence analysis among detected subjects showed that lower percentiles of nightly heart rate (e.g., HR_pct_20) were frequently triggered in the study window, indicating sustained increases in resting heart rate; at least one heart rate related parameter triggered in all 33 detected cases. HRV features, especially RMSSD percentile statistics, repeatedly showed persistent suppressions consistent with reduced parasympathetic tone. These persistent post-acute deflections were not observed in the SARS-negative cohort under the same compliance and thresholding rules.

[0155] Longitudinal cohort-level traces illustrate the contrast between detected persistent cases, acute-only SARS-positives not flagged as persistent, and SARS-negatives. When normalizing nightly features to individualized z-scores based on baseline means and standard errors, the SARS-negative group shows no sustained deviations across the observation horizon, whereas the detected persistent group exhibits a pronounced and protracted elevation in normalized resting heart rate following day 0 that remains above baseline for several months. The acute-only positive group reverts more quickly toward baseline. In complementary fashion, nightly median RMSSD in the persistent group remains moderately but chronically suppressed post day 0, while the acute-only group shows smaller and more transient deviations.

[0156] The disclosed evidence also supports epidemiologic associations enabled by the objective label produced by the disclosed herein method. A logistic regression across the SARS-positive cohort indicates that the 50-64-year age group has a statistically significant association with detection of persistent change with an odds ratio of 3.22 (p=0.027). While not meeting the 5% significance threshold, the model shows trends that males had higher odds (OR-3.9. p~0.1), the heaviest weight group had increased odds (OR-2.0. p~0.1), and prior vaccination reduced the odds (OR-0.4. p~0.1 ). The vaccination finding is directionally aligned with external literature and is supported by vaccination timing distributions in the cohort (average vaccination about 221 days prior to testing).

[0157] Device and signal provenance further support generalizability and feasibility. The cohort used commercially available devices — a Garmin Fenix 6 or Vivoactive 4 and an Oura ring — with 98.6% wearing Garmin watches and 89.8% wearing Oura rings, with 88.4% wearing both. Features are derived from sleep episodes as identified by the Oura hypnogram, including heart rate, inter-beat interval, respiratory rate derived from IBI, skin temperature, and SpO2, plus a library of HRV features (e.g., RMSSD, LF / HF, baroreflex sensitivity, multiscale entropy) summarized by nightly descriptors such as means and percentiles. The individualized-baseline framework and null calibration ensure portability across devices and firmware, as the detection decision depends on within-subject changes relative to baseline and on cross-subject null distributions constructed from test-confirmed negatives under identical acquisition and compliance rules.2024PF00609

[0158] 22

[0159] The disclosed methodology has several technical advantages. The individualized 90% confidence interval baselining and sustained out-of-bound testing across a predefined study window may replace symptom dependence with a strictly objective criterion, which can reduce misclassification due to reporting bias. The dual-cohort null model built from SARS-negative, asymptomatic subjects can provide a quantitative guardrail that screens out natural variation and seasonality, as demonstrated by the 1.9% null prevalence found in Monte Carlo resampling. The feature-direction constraint learned from the acute infection window can ensure that only physiologically plausible, illness-consistent deviations are counted in the persistent label, thereby reducing spurious detections. While the de-correlation and binomial post-hoc fusion produces subject-level detections with controlled false positive rates even when dozens of features are available, which is not taught by the symptom-centric or single-metric methods.

[0160] The disclosed evidence demonstrates that the claimed system and method enable detection of persistent post-acute physiological changes with statistical rigor, individualized baselines, null calibration, and multi-feature fusion, all of which are supported by concrete cohort-level prevalence, subject-level trajectories, and feature-level infection-direction analyses.

[0161] Returning to Fig. 1, in some examples, the system 100 may further comprise a feedback unit 130. The processing system 110 may be further configured to, in response to determining that the subject has a persistent change in health following the acute infection, control a feedback unit to provide a user-perceptible output indicating that a persistent change in health has been detected.

[0162] In Fig. 1, the feedback unit 130 is a display device configured to provide a user-perceptible output indicating that a persistent change in health has been detected in the form of a textual display and / or an image. However, any suitable user-perceptible output for indicating that a persistent change in health has been detected (e.g. one or more of: an image, a light, a textual display, a sound and / or a vibration); as the skilled person will appreciate, the type of user-perceptible output will depend on a type of feedback unit provided in the system 100.

[0163] Fig. 5 illustrates a computer-implemented method 500 for detecting a persistent change in health following an acute infection, according to an embodiment of the invention.

[0164] The computer-implemented method 500 begins at step 510, at which a set of physiological parameters for a subject are obtained. The set of physiological parameters are acquired over a period of time including a period prior to an acute infection and a period following the acute infection. The set of physiological parameters may, for example, be acquired using any of the techniques described above.

[0165] At step 520, a time of acute infection for the subject during the period over which the set of physiological parameters were acquired is identified. The time of acute infection may, for example, be identified based on a user input, or by providing the set of physiological parameters to an infection prediction model.

[0166] At step 530, at least the set of physiological parameters for the subject is processed to determine whether the subject has a persistent change in health following the acute infection. Whether the2024PF00609

[0167] 23

[0168] subject has a persistent change in health following the acute infection may, for example, be determined using any of the techniques described above.

[0169] In some examples, the computer-implemented method 500 may further comprise a step 540 of, in response to determining that the subject has a persistent change in health following the acute infection, controlling a feedback unit to provide a user-perceptible output indicating that a persistent change in health has been detected.

[0170] It will be understood that the disclosed methods are computer-implemented methods. As such, there is also proposed a concept of a computer program comprising code means for implementing any described method when said program is run on a processing system.

[0171] The skilled person would be readily capable of developing a processing system for carrying out any herein described method. Thus, each step of a flow chart may represent a different action performed by a processing system, and may be performed by a respective module of the processing system.

[0172] As discussed above, the system makes use of a processing system to perform the data processing. The processing system can be implemented in numerous ways, with software and / or hardware, to perform the various functions required. The processing system typically employs one or more microprocessors that may be programmed using software (e.g. microcode) to perform the required functions. The processing system may be implemented as a combination of dedicated hardware to perform some functions and one or more programmed microprocessors and associated circuitry to perform other functions.

[0173] Examples of circuitry that may be employed in various embodiments of the present disclosure include, but are not limited to, conventional microprocessors, application specific integrated circuits (ASICs), and field-programmable gate arrays (FPGAs).

[0174] In various implementations, the processing system may be associated with one or more storage media such as volatile and non-volatile computer memory such as RAM, PROM, EPROM, and EEPROM. The storage media may be encoded with one or more programs that, when executed on one or more processing systems and / or controllers, perform the required functions. Various storage media may be fixed within a processing system or controller may be transportable, such that the one or more programs stored thereon can be loaded into a processing system.

[0175] Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure and the appended claims. In the claims, the word “comprising” does not exclude other elements or steps, and the indefinite article “a” or “an” does not exclude a plurality.

[0176] Functions implemented by a processing system may be implemented by a single processing system or by multiple separate processing units which may together be considered to constitute a “processor”. Such processing units may in some cases be remote from each other and communicate with each other in a wired or wireless manner.2024PF00609

[0177] 24

[0178] The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0179] A computer program may be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems.

[0180] If the term “adapted to” is used in the claims or description, it is noted the term “adapted to” is intended to be equivalent to the term “configured to”. If the term “arrangement” is used in the claims or description, it is noted the term “arrangement” is intended to be equivalent to the term “system”, and vice versa.

[0181] Any reference signs in the claims should not be construed as limiting the scope.

Claims

2024PF0060925CLAIMS:

1. A processing system (110) for detecting a persistent change in health following an acute infection, the processing system being configured to:obtain a set of physiological parameters for a subject, the set of physiological parameters having been acquired over a period of time including a period prior to an acute infection and a period following the acute infection;identify a time of acute infection for the subject during the period over which the set of physiological parameters were acquired; andprocess at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection.

2. The processing system (110) of claim 1, wherein the step of processing at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection comprises determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection.

3. The processing system (110) of claim 2, wherein determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection comprises:identifying a baseline subset of the set of physiological parameters, wherein the baseline subset comprises data values acquired during the period prior to the acute infection; andfor each physiological parameter, processing the data values for the physiological parameter in the baseline subset to define an expected healthy range for the physiological parameter.

4. The processing system (110) of claim 3, wherein determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection further comprises:identifying a post-infection subset of the set of physiological parameters, wherein the post-infection subset comprises data values acquired during the period following the acute infection; and for each physiological parameter, identifying any data values for the physiological parameter in the post-infection subset that fall outside the expected healthy range to determine an out-of-bound percent metric for the physiological parameter.2024PF00609265. The processing system (110) of claim 4, wherein determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection further comprises:obtaining a null distribution for each physiological parameter, wherein the null distribution for each physiological parameter is based on physiological data from a plurality of healthy individuals; andfor each physiological parameter, comparing the out-of-bound percent metric for the physiological parameter to the null distribution for the physiological parameter.

6. The processing system (110) of claim 4 or 5, wherein:the set of physiological parameters comprises data values acquired during a period of acute infection; anddetermining the number of physiological parameters for the subject that demonstrate a persistent change post-infection further comprises:identifying an infection subset of the set of physiological parameters, wherein the infection subset comprises data values acquired during the period of acute infection; andfor each physiological parameter:comparing the data values for the physiological parameter in the infection subset to the data values for the physiological parameter in the baseline subset to identify a direction of change of the physiological parameter as a result of the acute infection; andidentifying whether any data values for the physiological parameter in the post-infection subset that fall outside the expected healthy range differ from the expected healthy range in the direction identified as the direction of change of the physiological parameter as a result of the acute infection.

7. The processing system (110) of any of claims 2 to 6, wherein determining the number of physiological parameters for the subject that demonstrate a persistent change post-infection comprises:defining an ordered list of physiological parameters for the subject that demonstrate a persistent change post-infection;removing, from the ordered list, any physiological parameter for which a correlation with another physiological parameter ranked higher in the ordered list exceeds a predetermined correlation threshold; anddetermining, as the number of physiological parameters for the subject that demonstrate a persistent change post-infection, the number of physiological parameters in the ordered list following the removal of any physiological parameter for which a correlation with another physiological parameter ranked higher in the ordered list exceeds the predetermined correlation threshold.2024PF00609278. The processing system (110) of any of claims 2 to 7, wherein the step of processing at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection further comprises:determining a maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance; andin response to the number of physiological parameters for the subject that demonstrate a persistent change post-infection being equal to or greater than the determined maximum number, determining that the subject has a persistent change in health following the acute infection.

9. The processing system (110) of claim 8, wherein the processing system is configured to determine the maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance by:defining a desired false-positive rate;defining a probability distribution expressing a probability of each physiological parameter demonstrating a persistent change post-infection; andusing the probability distribution to determine, as the maximum number of physiological parameters that could demonstrate a persistent change post-infection by chance, a maximum number of physiological parameters that could demonstrate a persistent change post-infection while resulting in a total probability of no more than the desired false-positive rate.

10. The processing system (110) of any of claims 1 to 9, wherein the processing system is further configured to, in response to determining that the subject has a persistent change in health following the acute infection, control a feedback unit (130) to provide a user-perceptible output indicating that a persistent change in health has been detected.

11. The processing system ( 110) of any of claims 1 to 10, wherein the set of physiological parameters for the subject was acquired using a wearable device (120).

12. The processing system (110) of any of claims 1 to 11, wherein the set of physiological parameters for the subject comprises one or more of: at least one heart rate parameter, at least one heart rate variability parameter, at least one temperature parameter, at least one respiratory rate parameter, at least one blood pressure parameter and / or at least one blood oxygen saturation level parameter.

13. The processing system (110) of any of claims 1 to 12, wherein the acute infection is a viral infection or a bacterial infection.2024PF006092814. A computer-implemented method (500) for detecting a persistent change in health following an acute infection, the computer-implemented method comprising:obtaining a set of physiological parameters for a subject, the set of physiological parameters having been acquired over a period of time including a period prior to an acute infection and a period following the acute infection;identifying a time of acute infection for the subject during the period over which the set of physiological parameters were acquired; andprocessing at least the set of physiological parameters for the subject to determine whether the subject has a persistent change in health following the acute infection.

15. A computer program product comprising computer program code means which, when executed on a computing device having a processing system, cause the processing system to perform all of the steps of the computer-implemented method (500) according to claim 14.

Images