System and method for measuring cardiopulmonary health condition of subject
The sensor unit analyzes the kinetic energy data flow during exercise periods and non-exercise periods, calculates the average kinetic energy period and sequential statistics, and converts it into a linear VO2 function, solving the economic and accuracy of the lung health measurement in various environments in the existing technology center. The problem of this achieves efficient cardiopulmonary health assessment in various scenarios.
Patent Information
- Application Number
- CN202380029629.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-02-10
- Filing Date
- 2023-02-10
- Publication Date
- 2024-11-08
AI Technical Summary
The prior art is difficult to provide an economical, accurate and efficient method of measuring cardiopulmonary health status in a variety of environments, especially maximum VO2 measurements, especially in resource-limited sports centers, developing countries or space vehicles.
Using a sensor unit including a motion sensing module, a cardiac impact schema and a cardiac shock schema, the average kinetic energy period and sequential statistics are calculated by analyzing the kinetic energy data flow during exercise periods and non-exercise periods, and converting them into a linear VO2 function to estimate the Maximum VO2 measure of subjects.
It realizes providing an economical and accurate measurement of cardiopulmonary health in different environments with an error of less than 10%, suitable for a variety of exercise equipment and scenarios, including sports centers and spacecraft.
Smart Images

Figure CN118922129A_ABST
Abstract
Description
[0001] field
[0002] Systems and methods are provided for measuring maximal oxygen consumption of a subject.
[0003] background
[0004] Cardiorespiratory fitness represents an important aspect of a subject's physical condition and is a key indicator of a subject's medical condition. Measurement of cardiorespiratory fitness is useful in many different situations, such as improving the performance of elite athletes, monitoring heart patients undergoing rehabilitation programs, and monitoring the cardiovascular system of astronauts who are unconditioned in space. The gold standard measure for cardiorespiratory fitness assessment is the maximum oxygen uptake rate a person can withstand, more commonly known as the VO2 max measurement. Obtaining an accurate VO2 max measurement requires a maximal effort spirometry test. The experiment involves complex gas exchangers and heavy machinery, requires expertise, is therefore costly, and is not entirely feasible in all environments (e.g., sports centers, developing countries, local clinics, space vehicles).
[0005] The aim is to provide an affordable, lightweight, more efficient to operate solution to the problem of cardiorespiratory fitness measurement, which provides a high level of measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0006] Figure 1 Exemplary placement of motion sensing modules on the posterior (panel A) and anterior (panel B), placement of a ballistocardiograph (panel C) and placement of a seismocardiograph (panel D) on a subject are shown.
[0007] Figure 2A An exemplary schematic diagram of the device and a sensor unit including a motion sensing module, a ballistocardiograph and a shockcardiograph is shown.
[0008] Figure 2B An exemplary schematic diagram of the device and a sensor unit including a ballistocardiograph and a shockcardiograph is shown.
[0009] Figure 3 In panel A, alternating exertion periods (S) and non-exercise periods (N) within an exercise period (E) are shown; panel B shows an example of a channel output from a sensor unit.
[0010] Figure 4is an exemplary schematic diagram of the method described herein, in which a mean period set of kinetic energy (SPMKE) is converted to a linear VO2 function (LVO2F).
[0011] Figure 5 and Figure 4 Similarly, the conversion of the mean kinetic energy period set (SPMKE) into a linear VO2 function (LVO2F) via a set of order statistics is shown.
[0012] Figure 6 In panel A, mean kinetic energy periods (PMKE) are shown, each mean kinetic energy (MKE) being determined from a different time window (W) of the kinetic energy data stream (KEDS) during a non-exercise period (N1). Panel B depicts mean kinetic energy periods (PMKE) plotted as a graph of mean kinetic energy (MKE) versus time during a non-exercise period (N). Panel C shows the conversion of the set of mean kinetic energy periods (SPMKE) into a VO2 function (VO2F) for the subject.
[0013] Figure 7 A version of the method is depicted in which each mean kinetic energy period (PMKE) is converted to an intermediate linear function (ILF), an order statistic (OS) is then determined for each ILF, and the set of order statistics (SOS) for the exercise period is converted to a linear VO2 function (LVO2F).
[0014] Figures 8A to 8D Describes in more detail Figure 7 Schematic diagram: Fig. 8A is the mean kinetic energy period (PMKE) during a non-exercise period; Figure 8B It is the mean kinetic energy period (PMKE) of the non-exercise period converted to an intermediate linear function; Figure 8C is the order statistic (1st quartile, median, 3rd quartile) of the intermediate linear function of the non-exercise period (N); Fig.8D is the set of order statistics (SOS) for the exercise period (E) converted into a linear VO2 function with respect to the subject.
[0015] Fig. 9 A version of the method is described in which an order statistic (OS) is determined for each mean kinetic energy period (PMKE), the set of order statistics for the exercise period (SOS) is then converted to an intermediate linear function (ILF), and the intermediate linear function (ILF) is then converted to a linear VO2 function (LVO2F) with respect to the subject.
[0016] Figures 10A to 10E Describes in more detail Fig. 9 Schematic diagram: Fig. 10A is the mean kinetic energy period (PMKE) during a non-exercise period; Fig. 10Bis the order statistic (1st quartile, median, 3rd quartile) of each PMKE in the non-exercise period (N); Fig. 10C is the order statistic (OS) (median) of PMKE for the exercise period (E); Fig. 10D is the transformation of the set of order statistics into an intermediate linear function (ILF) with respect to the exercise period (E); Fig.10E The intermediate linear function (ILF) of the exercise period (E) was converted to a linear VO2 function for the subject.
[0017] Fig.11A Exemplary channel outputs of a sensor unit including two motion sensing modules are shown.
[0018] Fig. 11B Exemplary channel outputs of a sensor unit including two motion sensing modules, one of which is an epicardiograph and the other is a ballistocardiograph, are shown.
[0019] Fig.12 Progression of median (circles) and quartiles (scale) of SOS (white) and VO2 (black) for a given subject during the E1 protocol.
[0020] Fig.13 is a scatter plot of estimated and measured VO2max values. Here, the estimated values are based on the prior knowledge of the maximum workload achieved and the value of the correction factor "d". The identification line is drawn with a solid black line, and the 10% and 20% error bars are drawn with dashed and dotted lines, respectively.
[0021] Overview
[0022] A system (100) for determining a cardiopulmonary fitness measure of a subject (50) is provided herein, the system (100) comprising a processing unit (160) configured to:
[0023] - receiving a stream (120) of signals output by a sensor unit (110), said sensor unit (110) comprising one or more motion sensing modules MSM (111),
[0024] - determining a kinetic energy data stream KEDS of each non-exercise period (N) of an exercise period (E) for gradually increasing the oxygen consumption VO2 of the subject from the signal stream (120), wherein the exercise period (E) comprises a plurality of alternating exertion segments (S) and non-exercise periods (N),
[0025] -For each non-exercise period (N):
[0026] - obtaining a mean kinetic energy MKE as a time-averaged kinetic energy value from the time window (W) of the kinetic energy data stream KEDS;
[0027] - determining a mean kinetic energy period PMKE, said mean kinetic energy period PMKE comprising a plurality of MKEs of different time windows (W) within said non-exercise period (N);
[0028] - determining the cardiorespiratory fitness measure of the subject from a set of PMKEs SPMKE containing a plurality of PMKEs during the exercise period (E).
[0029] The cardiorespiratory fitness measure of the subject can be determined by transforming the SPMKE in the following manner:
[0030] -Converting the SPMKE into a linear VO2 function (LVO2F) representing the subject's oxygen consumption VO2.
[0031] The conversion of the SPMKE to the LVO2F may include the step of converting each PMKE of the SPMKE into a scalar value representing the non-exercise period (N), wherein the scalar value is an order statistic.
[0032] The conversion of each PMKE of the SPMKE to a scalar value can be performed in the following way:
[0033] - indirectly for the PMKE, by first converting the MKE values within the PMKE into an intermediate linear function (ILF) and then sorting the values within the ILF to obtain the order statistic, or
[0034] - Directly for the PMKE, by value sorting the MKE values within the PMKE to obtain the order statistic.
[0035] The same order statistics may be determined for each non-exercise period (N), at least some of these order statistics (OS) forming a set of order statistics SOS for the exercise period (E), and the SOS being converted to the subject's LVO2F.
[0036] The conversion of the SOS to the LVO2F can be performed in the following manner:
[0037] -directly to the SOS, or
[0038] - indirectly to the SOS, by first converting the order statistic (OS) value within the SOS to an intermediate linear function (ILF), and then converting the intermediate linear function (ILF) to the LVO2F of the subject.
[0039] The order statistics (OS) may be medians or quantiles, and the order statistics (OS) in each set of order statistics (SOS) are all medians or all quantiles.
[0040] The exercise session (E) may cause the subject to reach exhaustion, preferably complete exhaustion.
[0041] The time window (W) may be a time period of w seconds, and the value of w is greater than the duration of one cardiac cycle of the subject.
[0042] The value of w may be the same during each non-exercise period (N) and during the exercise period (E).
[0043] The cardiorespiratory fitness measure of the subject (50) may be a VO2max measure of the subject (50).
[0044] One of the MSMs (111) may be a shockcardiograph SCG (114), and optionally, the other MSM (111) may be a ballistocardiograph BCG (112).
[0045] There is further provided a computer-implemented method for determining a measure of cardiopulmonary fitness of a subject (50), the computer-implemented method comprising:
[0046] - receiving a stream (120) of signals output by a sensor unit (110), said sensor unit (110) comprising one or more motion sensing modules (111),
[0047] - determining a kinetic energy data stream KEDS of each non-exercise period (N) of an exercise period (E) for gradually increasing the oxygen consumption VO2 of the subject from the signal stream (120), wherein the exercise period (E) comprises a plurality of alternating exertion segments (S) and non-exercise periods (N),
[0048] -For each non-exercise period (N):
[0049] - obtaining a mean kinetic energy (MKE) as a time-averaged kinetic energy value from a time window (W) of the kinetic energy data stream KEDS;
[0050] - determining a plurality of mean kinetic energies (MMKE) for different time windows (W);
[0051] - determining an order statistic (OS) from said plurality of mean kinetic energies (MMKE),
[0052] - determining a set of order statistics (SOS) comprising at least some of said order statistics (OS) determined from a plurality of non-exercise periods (N) of said exercise period (E);
[0053] - determining the cardiorespiratory fitness measure of the subject from the set of order statistics (SOS).
[0054] The computer-implemented method may include the limitations defined herein, particularly above.
[0055] There is further provided a computer program or a computer program product having instructions which, when executed by a computing device or system, cause the computing device or system to perform the computer-implemented method as defined herein.
[0056] Details
[0057] Before describing the present system and method of the present invention, it should be understood that the present invention is not limited to the specific system and method or combination described, because these systems and methods and combinations can certainly be varied. It should also be understood that the terminology used herein is not intended to be limiting, because the scope of the present invention will be limited only by the appended claims.
[0058] As used herein, the singular form "a," "an," and "the" include both singular and plural referents unless the content clearly dictates otherwise.
[0059] As used herein, the terms "comprising," "comprises," and "comprised of" are synonymous with "including," "includes," or "containing," "contains," and are inclusive or open-ended and do not exclude additional, unrecited members, elements, or method steps. It should be understood that the terms "comprising," "comprises," and "comprised of" as used herein include the terms "consisting of," "consists," and "consists of."
[0060] The descriptions of numerical ranges by endpoints include all numbers and fractions subsumed within the respective ranges as well as the indicated endpoints.
[0061] As used herein, the term "about" or "approximately", when referring to a measurable value such as a parameter, amount, duration, etc., is meant to include variations of + / -10% or less, preferably + / -5% or less, more preferably + / -1% or less, and even more preferably + / -0.1% or less of the specified value, as long as these variations are suitable for performing the disclosed invention. It should be understood that the value referred to by the modifier "about" or "approximately" itself is also specifically and preferably disclosed.
[0062] Although the term "one or more" or "at least one", e.g., one or more or at least one member of a group of members, is itself explicit, by further example, the term includes the specific examples to any one of the members, or to any two or more of the members, such as, for example, any ≥3, ≥4, ≥5, ≥6 or ≥7 of the members, etc., and up to all of the members.
[0063] The term "each" refers to a single value or group of values in a set. It does not necessarily mean every single value or every single value in the set.
[0064] All references cited in this specification are hereby fully incorporated by reference. In particular, the teachings of all references specifically mentioned herein are incorporated by reference.
[0065] Unless otherwise defined, all terms used in the disclosure of the present invention, including technical and scientific terms, have the meanings commonly understood by one of ordinary skill in the art to which the present invention belongs. By further guidance, term definitions are included to better understand the teachings of the present invention.
[0066] In the following paragraphs, different aspects of the invention will be defined in more detail. Each aspect so defined may be combined with any other aspect or aspects, unless explicitly stated to the contrary. In particular, any feature indicated as being preferred or advantageous may be combined with any other feature or features indicated as being preferred or advantageous.
[0067] Throughout the specification, references to "one embodiment" or "an embodiment" mean that a particular feature, structure, or characteristic associated with the embodiment is included in at least one embodiment of the present invention. Therefore, the phrases "in one embodiment" or "in an embodiment" that appear in different places throughout this specification do not necessarily all refer to the same embodiment, but may refer to the same embodiment. Further, particular features, structures, or characteristics may be combined in any suitable manner in one or more embodiments, as will be apparent to those skilled in the art from this disclosure. Further, although some embodiments described herein include some features included in other embodiments but not other features, the combination of features of different embodiments should be within the scope of the present invention and form different embodiments, as will be understood by those skilled in the art. For example, in the appended claims, any of the claimed embodiments may be used in any combination.
[0068] In the present description of the invention, reference is made to the accompanying drawings forming a part of the description, in which specific embodiments in which the invention may be implemented are shown by way of illustration only. The reference numerals in brackets or in bold attached to the various elements illustrate these elements by way of example only and are not intended to limit the various elements. Unless otherwise indicated, all figures and drawings in this document are not to scale and are selected for the purpose of illustrating different embodiments of the invention. In particular, the dimensions of the various components are described in illustrative terms only, and unless so indicated, no relationship between the dimensions of the various components should be inferred from the drawings.
[0069] It is to be understood that other embodiments may be utilized and structural or logical changes may be made without departing from the scope of the present invention. Therefore, the following detailed description is not to be taken in a limiting sense, and the scope of the present invention is defined by the appended claims.
[0070] A system (100) and method for determining a measure of cardiopulmonary fitness of a subject (50), in particular, maximum oxygen consumption (VO2 max). Figure 2A and Figure 2B shown.
[0071] The method or system comprises a processing unit (160) configured to receive a stream (120) of signals output by a sensor unit (110). The sensor unit (110) comprises one or more motion sensing modules (MSM) (111), each of which is configured to detect body vibrations caused by cardiovascular activity in the body. The sensor unit (110) may comprise: a motion sensing module (MSM) (111) as a shock cardiograph (SCG) (114), and optionally, a motion sensing module (MSM) (111) as a ballistocardiograph (BCG) (112), and optionally, one or more electrocardiograph (ECG) electrodes.
[0072] The method or processing unit (160) determines a kinetic energy data stream KEDS of each non-exercise period (N) of an exercise period (E) for gradually increasing the oxygen consumption VO2 of the subject from the signal stream (120), wherein the exercise period (E) includes a plurality of alternating exertion segments (S) and non-exercise periods (N).
[0073] The method or processing unit (160) is further configured to, for each non-exercise period (N):
[0074] - obtaining a mean kinetic energy MKE as a time-averaged kinetic energy value from the time window (W) of the kinetic energy data stream KEDS;
[0075] - Determining a mean kinetic energy period PMKE containing a plurality of MKEs for different time windows (W).
[0076] The method or processing unit (160) is further configured to obtain a set of PMKEs (SPMKE), wherein the SPMKE comprises a plurality of PMKEs (mean kinetic energy periods) obtained during an exercise period (E). A measure of the cardiorespiratory fitness of the subject can be determined from the set of PMKEs (SPMKE) because the SPMKE indicates a time-varying rate of oxygen consumption (VO2) for the exercise period (E). Steps such as Figure 4 In particular, the subject's VO2max can be determined by the oxygen consumption rate (VO2) corresponding to the point in time at which the subject reaches or is predicted to reach exhaustion.
[0077] A measure of a subject's cardiorespiratory fitness is a measure of the subject's physical condition. In particular, it is a measure of the subject's maximum rate of oxygen consumption (by volume) under maximal exertion. In other words, the more oxygen a subject is able to consume, the more oxygen reaches the muscle cells, allowing the subject to withstand longer and / or more intense exercise. If the increase in oxygen consumption rate is not sufficient to meet the increase in power (such as resistance in a bicycle workout), then maximum oxygen uptake can be detected. This means that the body is no longer able to increase its oxygen consumption and enters the anaerobic phase (energy consumption in the absence of oxygen). VO2 max is a measure of cardiorespiratory fitness.
[0078] The exercise session (E) performed by the subject contains a plurality of alternating exertion periods (S) and non-exercise periods (N). The exercise session (E) gradually increases the subject's VO2. An example of the exercise session (E) is shown in Figures 3 to 5 , 7 and 9; there are multiple exertion segments (S) and subsequent rest periods (non-exercise (N) periods). The exercise period (E) gradually increases the subject's heart rate, thereby increasing oxygen consumption (VO2). The exercise period (E) brings the subject to an exhaustion level. The subject can continue to exercise until they can no longer continue (complete exhaustion). The subject can continue to exercise until they reach a predetermined exhaustion level, such as 16 to 19 on the Borgscale (Rate of Perceived Exertion (RPE) known in the art). The duration of the exercise period (E) is determined by the subject's exhaustion level.
[0079] The exertion segment (S) may involve the use of a stationary exercise machine, such as a bicycle, elliptical trainer, or treadmill.
[0080] The intensity of the exertion segment (S) is set to increase the heart rate and breathing rate. The intensity of the exertion segment (S) can be between 50-300W, that is, the subject consumes this amount of energy during the exertion segment (S). For each exertion segment (S) in the entire exercise period (E), the intensity of the exertion segment (S) can be the same. As the exercise period (E) progresses, the intensity of the exertion segment (S) can gradually increase.
[0081] The length of the force segment (S) may be 30 to 180 seconds, more preferably 60 to 180 seconds. As the exercise period (E) progresses, the length of the force segment (S) may gradually increase. Preferably, the length of the force segment (S) is the same for each force segment (S) in the entire exercise period (E).
[0082] The first exertion segment (S) may be a warm-up period designed to slowly increase the heart rate and breathing rate. The intensity of the warm-up period may be lower than that of the subsequent exertion segments (S). The length of the warm-up period may be longer than that of the subsequent exertion segments (S).
[0083] The number of exertion segments (S) within an exercise period (E) may vary from subject to subject, as described elsewhere, and is determined by the subject's level of exhaustion. Subjects may continue exercising until they can no longer continue (total exhaustion), or until they reach a predetermined level of exhaustion, such as 16 to 19 on the Borg Scale (Rate of Perceived Exertion (RPE)).
[0084] The non-exercise period (N) is a period of inactivity. During the non-exercise period (N), the subject can sit, stand or lie down. The non-exercise period (N) usually follows the exertion period (S); in this case, it is a rest period or recovery period.
[0085] The length of the non-exercise period (N) may be 20 to 60 seconds, more preferably 30 to 45 seconds. Preferably, for each non-exercise period (N) in the entire exercise period (E), the length of the non-exercise period (N) is the same.
[0086] The period prior to the first exertion segment (S) is a baseline period during which baseline measurements are taken of the subject; this may or may not serve as a non-exercise period (N) to determine cardiorespiratory fitness, particularly VO2max.
[0087] The period after the first effort period (warm-up period) may or may not be used as a non-exercise period (N) to determine a cardiopulmonary fitness measure, in particular VO2max. Preferably, it is used as a non-exercise period (N) to determine a cardiopulmonary fitness measure, in particular VO2max.
[0088] The sensor unit (110) comprises one or more motion sensing modules (MSM) (111). It may further optionally comprise one or more electrocardiograph (ECG) electrodes.
[0089] The sensor unit (110) outputs a signal stream (120) from one or more motion sensors in one or more motion sensing modules (MSM) (111). The ECG output may or may not be part of the signal stream (120). Figure 2A An example of a sensor unit (110) is shown, the sensor unit (110) comprising at least one motion sensing module (MSM) (111a) and optionally two further motion sensing modules (MSM) (111b, 111c), which sends a signal stream (120) to a processing unit (160). In particular, the sensor unit (110) outputs a signal stream (120), the signal stream (120) comprising one or more channel outputs generated by the motion sensing module (MSM). The channel output reports linear motion or rotational motion (acceleration or velocity) in one direction (x, y, z). Fig.11AAn embodiment is shown of a sensor unit (110) and possible multiple channel outputs (111a-a to f) of a first motion sensing module (MSM) (111a) and possible multiple channel outputs (111b-a to f) of a second motion sensing module (MSM) (111b) that may be present in a signal stream (120). Figure 3 B shows exemplary channel outputs derived from BCG (Lin y, Rot x) and SCG (Lin z and Rot y) and ECG during a non-exercise period (N).
[0090] The motion sensing module (MSM) (111) detects body vibrations caused by cardiovascular activity in the body. The body vibrations originate from the heart and propagate through tissue structures, including the vascular system and bones, in the body. Therefore, the body vibrations repeat with each heartbeat, showing a certain pattern. The body vibrations are detected by mechanical transmission to the motion sensing module (MSM) (111). The motion sensing module (MSM) (111) is discrete. Discrete means that it occupies or detects movement from discrete areas of the body. When there are multiple motion sensing modules (MSM), they can be placed at different discrete locations on the body. Typically, the motion sensing module (MSM) (111) includes an outer (rigid) housing that receives body vibrations, which are transmitted to one or more motion sensors placed in a fixed relationship with the housing.
[0091] Non-limiting examples of the motion sensing module (MSM) (111) include a shockcardiograph (SCG) (114) and a ballistocardiograph (BCG) (112) as described elsewhere herein.
[0092] The motion sensing module (MSM) (111) outputs one or more channel outputs (e.g., 111a-a to f or 111b-a to f), each channel output reporting a different type of motion (linear or rotational) and / or a different direction (x, y, z).
[0093] The motion sensing module (MSM) (111) is configured to be placed at discrete locations on a subject, typically on the skin, or on a material in contact with the skin (e.g., clothing). Figure 1A and 1B show examples of discrete locations (a to s). Primary examples of discrete locations include around the subject's center of mass (or gravity) (e.g., as used in BCG), lumbar region (lordotic curve, between the second and third lumbar vertebrae) (e.g., as used in BCG), chest (e.g., as used in SCG), sternum, upper sternum (manubrium) (e.g., as used in SCG), sternum body (corpus sternal) (e.g., as used in SCG). Further examples of discrete locations include head (top, back, forehead), neck, shoulder, arm, wrist, hand, finger, leg, thigh, calf ankle, foot (upper or bottom), toe. Figure 1 Shown in A are motion sensing modules (MSMs) (111) located at different exemplary discrete locations: a-head, b-cervical region, c-thoracic region, d-upper arm (shown on the left), e-lumbar region (lordotic flexion), f-wrist (shown on the left), g-fingers (shown on the left), h-thigh (shown on the right), i-calf (shown on the right), j-ankle (shown on the right), k-sole (shown on the left). Figure 1 Shown in B are motion sensing modules (MSM) (111) located at different exemplary discrete locations: 1-forehead, m-shoulder, n-upper sternum (manubrium), o-body of sternum (sternal part), p-wrist, q-thigh, r-calf, s-toe. Although the motion sensing module (MSM) (111) is exemplarily shown on only one side of the body (e.g., the left side), it should be understood that it can be alternately or additionally placed on the other side (e.g., the right side).
[0094] The motion sensing module (MSM) (111) includes one or more motion sensors for detecting body movement. The one or more motion sensors may be included in the housing of the motion sensing module (MSM) (111). The motion sensor detects linear motion or rotational motion. Each motion sensor may generate one or more channel outputs. The sampling rate may be between 40 Hz and 1000 Hz. The motion sensing module (MSM) may include a motion sensor, which is an accelerometer for measuring linear acceleration. The accelerometer may be a one, two or three-dimensional accelerometer for measuring linear acceleration in one, two or three different directions (1, 2, or 3 linear degrees of freedom (1, 2, or 3DOF)). Depending on the number of degrees of freedom measured, the motion sensing module (MSM) (111) may generate one, two or three channel outputs. Preferably, the motion sensing module (MSM) (111) includes a three-dimensional accelerometer for measuring linear acceleration in three different directions (3 linear degrees of freedom (3DOF)); the three-dimensional accelerometer preferably generates three channel outputs ( Fig.11A, 111a-a to -c or 111b-a to -c), each linear dimension being measured corresponds to a channel output.
[0095] The motion sensing module (MSM) (111) may alternatively or additionally include a motion sensor that is a gyroscope for measuring rotational motion. The gyroscope may be a one, two or three-dimensional gyroscope that measures rotational motion in one, two or three different directions (1, 2 or 3 linear degrees of freedom (1, 2 or 3DOF)). The motion sensing module (MSM) (111) may produce one, two or three channel outputs depending on the number of degrees of freedom being measured. Preferably, the motion sensing module (MSM) (111) includes a three-dimensional gyroscope for measuring rotational motion around three different axes (3 linear degrees of freedom (3DOF)); the three-dimensional gyroscope preferably produces three channel outputs ( Fig. 11B , 111a-d to -f or 111b-d to -f), each measured rotational dimension corresponds to a channel output.
[0096] For measurements of the motion sensing module (MSM) (111), the x-axis can be the lateral axis (horizontal axis: from the left side of the subject to the right side of the subject); the y-axis can be the longitudinal axis (caudal-cranial axis, from the feet to the head), and the z-axis can be the anterior-posterior axis (ventral-dorsal axis) (see Figure 1 A and 1B).
[0097] Preferably, the motion sensing module (MSM) (111) comprises at least two motion sensors, one being a one, two or three-dimensional (preferably three-dimensional) accelerometer and the other being a one, two or three-dimensional (preferably three-dimensional) gyroscope. Preferably, the motion sensing module (MSM) (111) produces two or more channel outputs, producing one, two or three (preferably three) channel outputs from the accelerometer ( Fig.11A , 111a-a to -c or 111b-a to -c), generating one, two or three (preferably three) channel outputs from the gyroscope ( Fig.11A , 111a-d to f or 111b-d to f).
[0098] The motion sensing module (MSM) may further include a mechanism for attachment to the subject's body, such as a band or strap (e.g., for attachment to the chest, head or waist, wrist, etc.), adhesive, clothing (e.g., hat, sweatband, vest, gloves, shorts, socks, shoes), or a wearable device (e.g., suspenders, watches, glasses, rings, hats). It may be incorporated into a health tracker (e.g., chest strap, watch, ankle strap).
[0099] The motion sensing module (MSM) may include inputs for one or more electrocardiograph (ECG) electrodes for measuring ECG signals. The motion sensing module (MSM) may include one or more interface modules for communication, for example, wireless or wired data transmission of the one or more channel outputs. Data transmission may be performed between a motion sensing module (MSM) and a smartphone, tablet computer or other computing device. Data transmission may be performed between a motion sensing module (MSM) and one or more further motion sensing modules (MSM).
[0100] One example of a motion sensing module (MSM) (111) is a shock cardiograph (SCG) (114). The shock cardiograph (SCG) is configured to be placed on the chest of a subject, for example, on the sternum, upper sternum (manubrium), or body of the sternum (sternal part). Figure 1 A shows the placement of an epicardiocardiograph (114) at a discrete location on the subject, either on the upper sternum (manubrium) (n) or the body of the sternum (sternum) (o). Figure 1 D shows an exemplary placement of an epicardiocardiograph (114) at a discrete location on the subject, namely, the upper sternum (manubrium). Figure 2B A shockcardiograph (114) is shown as part of the sensor unit (110) and sends signals to the processing unit (160). The shockcardiograph measures the movement of the heart, such as the contraction of the heart and the ejection of blood into the aorta, such as the opening and closing of the aortic valve. The shockcardiograph (SCG) outputs one or more channel outputs, each channel output reporting a different movement (linear or rotational) and / or a different direction (x, y, z).
[0101] The shockcardiograph (SCG) includes one or more motion sensors for detecting body movement. The motion sensors detect linear or rotational motion. Each motion sensor can produce one or more channel outputs. The sampling rate can be between 40 Hz and 1000 Hz.
[0102] The seismocardiograph (SCG) may include a motion sensor that is an accelerometer for measuring linear acceleration. The accelerometer may be a one, two, or three-dimensional accelerometer for measuring linear acceleration in one, two, or three different directions (1, 2, or 3 linear degrees of freedom (1, 2, or 3DOF)). Depending on the number of degrees of freedom measured, the seismocardiograph may produce one, two, or three channel outputs. Preferably, the seismocardiograph includes a three-dimensional accelerometer for measuring linear acceleration in three different directions (3 linear degrees of freedom (3DOF)); the three-dimensional accelerometer preferably produces an output three channel outputs ( Fig. 11B , 115a-c), each measured linear dimension corresponds to a channel output.
[0103] The seismocardiograph (SCG) may alternatively or additionally include a motion sensor that is a gyroscope for measuring rotational motion. The gyroscope may be a one, two or three-dimensional gyroscope for measuring rotational motion in one, two or three different directions (1, 2 or 3 linear degrees of freedom (1, 2 or 3DOF)). The seismocardiograph may produce one, two or three channel outputs depending on the number of degrees of freedom measured. Preferably, the seismocardiograph includes a three-dimensional gyroscope for measuring rotational motion in three different axes (3 linear degrees of freedom (3DOF)); the three-dimensional gyroscope preferably produces three channel outputs (Figure 11, 115d-f), one channel output for each measured rotational dimension.
[0104] For epicardiograph measurements, the x-axis can be the transverse axis (horizontal axis: from the subject's right side to the subject's left side); the y-axis can be the longitudinal axis (caudal-cranial axis, from the feet to the head); and the z-axis can be the anterior-posterior axis (ventral-dorsal axis, from the back to the abdomen) (see Figure 1 B).
[0105] Preferably, the seismocardiograph (SCG) comprises at least two motion sensors, one being a one, two or three-dimensional (preferably three-dimensional) accelerometer and the other being a one, two or three-dimensional (preferably three-dimensional) gyroscope. Preferably, the seismocardiograph produces two or more channel outputs, one, two or three (preferably three) channel outputs from the accelerometer ( Fig. 11B , 115a-c), and generate one, two or three (preferably three) channel outputs from the gyroscope ( Fig. 11B , 115d-f).
[0106] The seismocardiograph may further include a mechanism for attachment to the subject's body, such as a belt mounted on a body-worn, adhesive gel pad. The seismocardiograph may include an input for one or more electrocardiograph (ECG) electrodes for measuring ECG signals. The seismocardiograph may include one or more interface modules for communication, such as wireless or wired data transmission for the one or more channel outputs. Data transmission may be performed between the seismocardiograph and a smart phone, tablet computer or other computing device. Data transmission may be performed between the seismocardiograph and a ballistocardiograph.
[0107] One example of a motion sensing module (MSM) (111) is a ballistocardiograph (BCG) (112). The ballistocardiograph (BCG) is configured to be placed on a subject, about the subject's center of mass (or gravity). Figure 1 B shows the ballistocardiograph (112) placed at a discrete location on the subject, namely the dorsal lumbar region (lordotic flexion) (e). Figure 1C shows an exemplary placement of a ballistocardiograph (112) at a discrete location on a subject, namely the dorsal lordotic curvature (between the second and third lumbar vertebrae). The ballistocardiograph measures whole body motion. The ballistocardiograph (BCG) outputs one or more channel outputs (113a-f), each channel output reporting a different motion type (linear or rotational) and / or a different direction (x, y, z).
[0108] The ballistocardiograph (BCG) includes one or more motion sensors for detecting body movement. The motion sensor detects linear motion or rotational motion. Each motion sensor can produce one or more channel outputs. The sampling rate can be between 40 Hz and 1000 Hz. The ballistocardiograph (BCG) may include a motion sensor, which is an accelerometer for measuring linear acceleration. The accelerometer can be a one, two or three-dimensional accelerometer for measuring linear acceleration in one, two or three different directions (1, 2, or 3 linear degrees of freedom (1, 2, or 3DOF)). Depending on the number of degrees of freedom measured, the ballistocardiograph can produce one, two or three channel outputs. Preferably, the ballistocardiograph includes a three-dimensional accelerometer for measuring linear acceleration in three different directions (3 linear degrees of freedom (3DOF)); the three-dimensional accelerometer preferably produces three channel outputs ( Fig. 11B , 113a-c), each measured linear dimension corresponds to a channel output.
[0109] The ballistocardiograph (BCG) may alternatively or additionally include a motion sensor which is a gyroscope for measuring rotational motion. The gyroscope may be a one, two or three dimensional gyroscope for measuring rotational motion in one, two or three different directions (1, 2 or 3 linear degrees of freedom (1, 2 or 3DOF)). The ballistocardiograph may produce one, two or three channel outputs depending on the number of degrees of freedom being measured. Preferably, the ballistocardiograph includes a three dimensional gyroscope for measuring rotational motion around three different axes (3 linear degrees of freedom (3DOF)); the three dimensional gyroscope preferably produces three channel outputs ( Fig. 11B , 113d-f), each measured rotational dimension corresponds to a channel output.
[0110] For ballistocardiograph (BCG) measurements, the x-axis can be the transverse axis (horizontal axis: from the subject's left side to the subject's right side); the y-axis can be the longitudinal axis (caudal-cranial axis, from the feet to the head), and the z-axis can be the anterior-posterior axis (ventral-dorsal axis, from the abdomen to the back) (see Figure 1 A).
[0111] Preferably, the ballistocardiograph (BCG) comprises at least two motion sensors, one being a one, two or three-dimensional (preferably three-dimensional) accelerometer and the other being a one, two or three-dimensional (preferably three-dimensional) gyroscope. Preferably, the ballistocardiograph produces two or more channel outputs, one, two or three (preferably three) channel outputs from the accelerometer ( Fig. 11B , 113a-c), generating one, two or three (preferably three) channel outputs from the gyroscope ( Fig. 11B , 113d-f).
[0112] The ballistocardiograph (BCG) may further include a mechanism for attachment to the subject's body, such as a belt mounted to the body, an adhesive gel pad. The ballistocardiograph may include an input for one or more electrocardiograph (ECG) electrodes for measuring ECG signals. The ballistocardiograph may include one or more interface modules for communication, such as wireless or wired transmission of the one or more channel outputs. Data transmission may be performed between the ballistocardiograph and a smart phone, tablet computer or other computing device.
[0113] The sensor unit (110) may be configured to generate a stream (120) of signals from a shockcardiograph (SCG) and optionally from a ballistocardiograph (BCG). The sensor unit (100) signal stream (120) may include:
[0114] - one or more channel outputs (115a-c) from one or more accelerometer axes (3 linear (x, y, z) axes) of the SCG; and / or
[0115] - one or more channel outputs (115d-f) from one or more gyroscopic axes (3 rotational (x, y, z) axes) of the SCG; and / or
[0116] - one or more channel outputs (113a-c) from one or more accelerometer axes (3 linear (x, y, z) axes) of the BCG; and / or
[0117] - One or more channel outputs (113d-f) from one or more gyro axes (3 rotational (x, y, z) axes) of the BCG.
[0118] The sensor unit (100) signal stream (120) may include a channel output (115b) from the y-linear axis accelerometer axis of a shockcardiograph (SCG). The sensor unit (100) signal stream (120) may include channel outputs (115a-c) from three accelerometer axes (all 3 linear (x, y, z) axes) of a ballistocardiograph (BCG).
[0119] The sensor unit (100) signal stream (120) may include a channel output (113b) from the y-linear axis accelerometer axis of a ballistocardiograph (BCG). The sensor unit (100) signal stream (120) may include channel outputs (113a-c) from three accelerometer axes (all 3 linear (x, y, z) axes) of a ballistocardiograph (BCG).
[0120] It has been discovered that a cardiopulmonary fitness measure, in particular, a VO2max value, can be determined from only one shockcardiograph (SCG) motion axis (e.g., y-linear axis, one channel output (115b)) or one ballistocardiograph (BCG) motion axis (e.g., y-linear axis, one channel output (113b)), however, more motion axes may be used to consider more optimized cardiopulmonary fitness measures, in particular, VO2max values.
[0121] The sensor unit (110) may include (only) one motion sensing module (MSM) (111) for the convenience of the subject. For ease of wearing, the sensor unit (110) may include (only) a shock cardiograph (SCG) (114). In this case, the shock cardiograph (SCG) includes at least one motion sensor; the motion sensor may be a one, two or three-dimensional (preferably three-dimensional) accelerometer, or a one, two or three-dimensional (preferably three-dimensional) gyroscope.
[0122] The sensor unit (110) may include both a ballistocardiograph (BCG) and a shockcardiograph (SCG). In this case, the ballistocardiograph (BCG) includes at least one motion sensor, which is a one, two or three-dimensional (preferably three-dimensional) accelerometer, or a one, two or three-dimensional (preferably three-dimensional) gyroscope. The shockcardiograph (SCG) may include at least one motion sensor, which is a one, two or three-dimensional (preferably three-dimensional) accelerometer, or a one, two or three-dimensional (preferably three-dimensional) gyroscope. The sensor unit (110) may generate two or more channel outputs:
[0123] - one, two or three (preferably three) from BCG accelerometers, and / or one, two or three (preferably three) from BCG gyroscopes, and / or
[0124] - One, two or three (preferably three) from SCG accelerometers, and / or one, two or three (preferably three) from SCG gyroscopes.
[0125] An example of a sensor unit (110) that includes both SCG and BCG is the "Kino" created by HeartKinetics of Gosselies, Belgium. It consists of two small modules that independently acquire the subject's SCG and BCG signals, as well as a 1-lead ECG (one lead includes electrodes of opposite polarity).
[0126] The processing unit (160) includes circuitry configured to perform the method of the present invention. Typically, the circuitry includes a processor and a memory. The processing unit (160) can be implemented in a computing device such as a desktop PC, a laptop, a dedicated programmable controller as a collection of connected computing devices. The processing unit can be provided in part or in whole by a processor and a memory, which are arranged in the housing of the sensor unit (110), or the ballistocardiograph (112) or the shockcardiograph (114), or in the housing of a separate unit. The processing unit (160) can be configured to perform one or more of the methods or parts thereof described herein.
[0127] The processing unit (160) comprises or is a part of a standard computer system, such as an Intel Architecture IA-32based computer system 2, and can implement programmed instructions of one or more software modules stored on a non-volatile (e.g., hard disk or solid state drive) memory associated with the corresponding computer system. However, it is apparent that at least some of the steps of any of the described processes may alternatively be performed in part or in whole as one or more dedicated hardware components, such as gate configuration data of one or more field programmable gate arrays (FPGAs), or, for example, as an application specific integrated circuit (ASIC).
[0128] A kinetic energy data stream (KEDS) is determined for each non-exercise period (N) from the signal stream (120). The kinetic energy data stream (KEDS) is determined from the sensor unit (100) signal stream (120) on one or more output channels. The determination of the KEDS for each non-exercise period (N) is schematically shown in Figure 4 , 5 , 7, 9. The kinetic energy data stream (KEDS) can be determined in real time when the subject is in a non-exercise period (N) of the exercise period (E), or can be determined later from a record of a sensor unit (100) signal stream (120) during the exercise period (E). The kinetic energy data stream (KEDS) includes a plurality of discrete kinetic energy values (DKEV), each discrete kinetic energy value (DKEV) corresponding to a different time point or block of the non-exercise period (N).
[0129] The kinetic energy data stream (KEDS) may be determined by one or more of the following:
[0130] - linear kinetic energy along one or more axes (such as x, y or z);
[0131] - rotational kinetic energy about one or more axes (such as x, y or z);
[0132] - one or more parameters related to linear kinetic energy and / or rotational kinetic energy obtained from the sensor unit (110).
[0133] The sensor unit (100) signal stream (120) can be converted into a kinetic energy data stream (KEDS), and the conversion is performed according to the method described in Hossein et al. in "Accurate Detection of Dobutamine-induced Haemodynamic Changes by Kino-Cardiography: A Randomised Double-Blind Placebo-Controlled Validation Study", Nature Scientific Reports (2019) 9: 10479 (https: / / doi.org / 10.1038 / s41598-019-46823-3).
[0134] The linear kinetic energy (K) of a single linear axis ((ax), which can be x, y or z) at time (t) Lin ) can be determined according to the following equation:
[0135] K Lin(ax)(t) =0.5[mv ax(t) 2 ],in
[0136] -K Lin(ax)(t) is the linear kinetic energy along the axis (ax) at time (t), in joules;
[0137] -v ax(t) is the linear velocity along the axis (ax) at time (t) measured by the MSM (e.g. BCG or SCG) in m / s;
[0138] -m is the mass of the subject in kg.
[0139] When multiple linear axes (x, y, and z) have multiple channel outputs, the channel outputs are summed, that is:
[0140] Linear kinetic energy (K) in multiple linear axes (x, y, and z) at time (t) Lin ) can be determined according to the following equation:
[0141] K Lin(x,y,z)(t) =0.5[m((v x(t) 2 )+(v y(t) 2 )+(V z(t) 2 ))]
[0142] -K Lin(x,y,z)(t) is the linear kinetic energy along the axis (x, y, z) at time (t), in joules;
[0143] -v x(t) is the linear velocity along the axis (x) at time (t) measured by the MSM (e.g. SCG or BCG) in m / s;
[0144] -v y(t) is the linear velocity along the axis (y) at time (t) measured by the MSM (e.g. SCG or BCG) in m / s;
[0145] -v z(t) is the linear velocity along the axis (z) at time (t) measured by the MSM (e.g. SCG or BCG) in m / s;
[0146] -m is the mass of the subject in kg.
[0147] If two of the three axes are used, remove the unused axis from the equation.
[0148] The rotational kinetic energy (K) at time (t) about a single linear axis ((ax), which can be x, y, or z) Rot ) can be determined according to the following equation:
[0149] K Rot(ax)(t) =0.5[I (ax) .W (ax)(t) 2 ],in:
[0150] -K Rot(ax)(t) is the rotational kinetic energy around the axis (ax) at time (t), in joules;
[0151] -I (ax) is the moment of inertia about the axis (ax), in kg.m 2The moment of inertia can be the moment of inertia of the sensor unit, or the moment of inertia of the body. The model of the human body moment of inertia is calculated at the center of mass along the anatomical axis (I(ax)), which takes into account the height and weight of the subject (see Matsuo, A., Ozawa, H., Goda, K. & Fukunaga, T. Moment of inertia of whole body using an oscillating table in adolescent boys. J. Biomech. 28, 219-223 (1995).)
[0152] -W (ax)(t) is the angular velocity about an axis (ax) measured at time (t) by an MSM (such as SCG or BCG) in rad / s.
[0153] When there are multiple channels outputting rotations around multiple linear axes (x, y, and z), the channel outputs are summed, that is:
[0154] The rotational kinetic energy (K) of a rotation about multiple linear axes (x, y, and z) at time (t) Rot ) can be determined according to the following equation:
[0155] K Rot(x,y,z)(t) =0.5(I (x) (W (x)(t) ) 2 +I (y) (W (y)(t) ) 2 +I (z) (W (z)(t) ) 2 )
[0156] -K Rot(x,y,z)(t) is the rotational kinetic energy around the axis (x, y, z) at time (t), in joules;
[0157] -I(x), I(x), I(x) are the moments of inertia around the axes (x), (y), (z), respectively, in kg.m 2The moments of inertia may be those of the sensor unit, or those of the body. The model of the human body's moments of inertia is calculated along the anatomical axes (I(x), I(y), I(z)) at the center of mass, taking into account the height and weight of the subject (see also Matsuo, A., Ozawa, H., Goda, K. & Fukunaga, T. Moment of inertia of whole body using anoscillating table in adolescent boys. J. Biomech. 28, 219-223 (1995).)
[0158] -W (x)(t) , W (y)(t) , W (z)(t) are the angular velocities around the axes (x), (y), and (z), respectively, measured at time (t) by the MSM (such as SCG or BCG), in rad / s.
[0159] If two of the three axes are used, remove the unused axis from the equation.
[0160] The kinetic energy data stream (KEDS) may be determined from one or more parameters related to linear kinetic energy and / or related to rotational kinetic energy (ie one or more related parameters) obtained from the sensor unit (110).
[0161] An example of a relevant parameter is the amplitude (also referred to as magnitude) of one or more channel outputs (e.g., 111a-a to f or 111b-a to f) of a motion sensing module (MSM) (111) of the sensor unit (110). In particular, an example of a relevant parameter is the amplitude expressed in terms of acceleration, velocity, displacement, or force. Another example is the signal energy of one or more channel outputs (e.g., 111a-a to f or 111b-a to f) of a motion sensing module (MSM) (111) of the sensor unit (110). In particular, an example of a relevant parameter is the signal energy expressed in terms of acceleration, velocity, displacement, or force.
[0162] The relevant parameter as amplitude is the norm of the signal from one or more channel outputs (e.g., 111a-a to f or 111b-a to f) of the motion sensing module (MSM) (111) of the sensor unit (110). This signal can be an acceleration vector, a velocity vector, a displacement vector or a force vector. The unit of the acceleration vector is preferably m / s 2 The unit of the velocity vector is preferably m / s. The unit of the displacement vector is preferably m. The unit of the force vector is preferably N (Newton).
[0163] The magnitude of the VoM at time (t) about a single linear axis ((ax), which can be x, y, or z) can be determined according to the following equation:
[0164] A Lin(ax)(t) =sqrt[VoM ax(t) 2 ],in
[0165] - VoM is a motion vector, the motion vector being one of a linear acceleration vector, a linear velocity vector, a linear displacement vector or a linear force vector;
[0166] -A Lin(ax)(t) is the magnitude of VoM in the VoM unit along the axis (ax) at time (t);
[0167] -VoM ax(t) is the VoM measured by an MSM (eg, SCG or BCG) along axis (ax) at time (t) in a VoM unit.
[0168] When the VoM has multiple channel outputs for multiple linear axes (x, y, and z), the channel outputs are summed, that is:
[0169] The amplitude of the multiple linear axes (x, y, and z) of the VoM at time (t) can be determined according to the following equation: Lin(x,y,z)(t) =sqrt[((VoM x(t) 2 )+(VoM y(t) 2 )+(VoM z(t) 2 ))],in
[0170] - VoM is a motion vector, which is one of a linear acceleration vector, a linear velocity vector, a linear displacement vector or a linear force vector.
[0171] -A Lin(x,y,z)(t) is the magnitude of VoM in the VoM unit along the axis (x, y, z) at time (t);
[0172] -VoM x(t) is the VoM measured by the MSM (e.g., SCG or BCG) at time (t) along axis (x) in the VoM unit;
[0173] -VoM y(t) is the VoM measured by an MSM (e.g., SCG or BCG) at time (t) along axis (y) in the VoM unit;
[0174] -VoM z(t) is the VoM measured by the MSM (e.g., SCG or BCG) along the axis (z) at time (t) in the VoM unit;
[0175] If two of the three axes are used, remove the unused axis from the equation.
[0176] The magnitude of the VoM at time (t) about a single linear axis ((ax), which can be x, y, or z) can be determined according to the following equation:
[0177] A Rot(ax)(t) =sqrt[VoM ax(t) 2 ],in
[0178] - VoM is a motion vector, the motion vector being one of a rotational acceleration vector, a rotational velocity vector, a rotational displacement vector or a rotational force vector;
[0179] -A Rot(ax)(t) is the magnitude of VoM in the VoM unit around the axis (ax) at time (t);
[0180] -VoM ax(t) is the VoM measured by an MSM (eg, SCG or BCG) at time (t) around an axis (ax) in a VoM unit.
[0181] When the VoM has multiple channel outputs for multiple linear axes (x, y, and z), the channel outputs are summed, that is:
[0182] The magnitude of the VoM at time (t) around multiple linear axes (x, y, and z) can be determined according to the following equations:
[0183] A Rot(x,y,z)(t) =sqrt[((VoM x(t) 2 )+(VoM y(t) 2 )+(VoM z(t) 2 ))],in
[0184] - VoM is a motion vector, the motion vector being one of a rotational acceleration vector, a rotational velocity vector, a rotational displacement vector or a rotational force vector;
[0185] -A Rot(x,y,z)(t) is the magnitude of VoM in the VoM unit around the axis (x, y, z) at time (t);
[0186] -VoM x(t) is the VoM measured by an MSM (e.g., SCG or BCG) in VoM units around axis (x) at time (t);
[0187] -VoM y(t)is the VoM measured by an MSM (e.g., SCG or BCG) in VoM units around axis (y) at time (t);
[0188] -VoM z(t) is the VoM measured by an MSM (eg, SCG or BCG) at time (t) around an axis (z) in VoM units.
[0189] If two of the three axes are used, remove the unused axis from the equation.
[0190] The relevant parameter as signal energy is the area under the squared magnitude (or amplitude) of the signal output from one or more channels (e.g., 111a-a to f or 111b-a to f) of the motion sensing module (MSM) (111) of the sensor unit (110). This signal can be an acceleration vector, a velocity vector, a displacement vector or a force vector. The unit of the acceleration vector is preferably m / s 2 The unit of the velocity vector is preferably m / s. The unit of the displacement vector is preferably m. The unit of the force vector is preferably N (Newton).
[0191] The signal energy of the VoM at time (t) about a single linear axis ((ax), which can be x, y or z) can be determined according to the following equation:
[0192] E Lin(ax)(t) =Integral(A Lin(ax)(t) 2 )
[0193] -E Lin(ax)(t) is the energy of the VoM in the VoM unit along the axis (ax) at time (t);
[0194] -A Lin(ax)(t) is the magnitude of VoM in the VoM unit along the axis (ax) at time (t);
[0195] - VoM is a motion vector, which is one of a linear acceleration linear vector, a linear velocity vector, a linear displacement vector or a linear force vector.
[0196] The signal energy of the VoM at time (t) with respect to multiple linear axes (x, y'z) can be determined according to the following equation:
[0197] E Lin(x,y,z)(t) =Integral(A Lin(x,y,z)(t) 2 ),in
[0198] -E Lin(x,y,z)(t) is the energy of the VoM in the VoM unit along the axis (x, y, z) at time (t);
[0199] -A Lin(x,y,z)(t)is the magnitude of VoM in the VoM unit along the axis (x, y, z) at time (t);
[0200] - VoM is a motion vector, which is one of a linear acceleration vector, a linear velocity vector, a linear displacement vector or a linear force vector.
[0201] If two of the three axes are used, remove the unused axis from the equation.
[0202] The signal energy at time (t) about a single linear axis ((ax), which can be x, y or z) of the VoM can be determined according to the following equation:
[0203] E Rot(ax)(t) =Integral(A Rot(ax)(t) 2 ),
[0204] -E Rot(ax)(t) is the energy of VoM in VoM unit around axis (ax) at time (t);
[0205] -A Rot(ax)(t) is the magnitude of VoM in the VoM unit around the axis (ax) at time (t);
[0206] - VoM is a motion vector, the motion vector being one of a rotational acceleration vector, a rotational velocity vector, a rotational displacement vector or a rotational force vector;
[0207] The signal energy of the VoM around multiple linear axes (x, y, and z) at time (t) can be determined according to the following equation:
[0208] E Rot(x,y,z)(t) =Integral(A Rot(x,y,z)(t) 2 ),in
[0209] -E Rot(x,y,z)(t) is the energy of VoM in VoM unit around axis (x, y, z) at time (t);
[0210] -A Rot(x,y,z)(t) is the magnitude of VoM in the VoM unit around the axis (x, y, z) at time (t);
[0211] - VoM is a motion vector, which is one of a rotational acceleration vector, a rotational velocity vector, a rotational displacement vector or a rotational force vector.
[0212] If two of the three axes are used, remove the unused axis from the equation.
[0213] Obtain a mean kinetic energy (MKE), which is a time-averaged kinetic energy value from a time window (W) of the kinetic energy data stream. The time window (W) is a time period of w seconds. The value of w is less than the duration of a non-exercise period (N). Typically, w is greater than the duration of one cardiac cycle of the subject. The value of w may be in the range of 2 to 10 seconds. The value of w is preferably the same during each non-exercise period (N). The value of w is preferably the same during each non-exercise period (N) and during an exercise period (E).
[0214] The average kinetic energy (MKE) may be determined from the time window (W) of the kinetic energy data stream (KEDS) by any method to calculate the average of the discrete kinetic energy values (DKEV) in a given time window. For example, the average kinetic energy (MKE) may be determined by the integration of the discrete kinetic energy values (DKEV) measured within the time window (W) of the kinetic energy data stream (KEDS). For example, the average kinetic energy (MKE) may be determined by the time average of the discrete kinetic energy values (DKEV) measured within the time window (W) of the kinetic energy data stream (KEDS).
[0215] By determining each of a plurality of average kinetic energies (MKE) at different starting time points within a non-exercise period (N), a mean kinetic energy period (PMKE) is determined within each non-exercise period (N). Some mean kinetic energy periods (PMKE) may overlap but have different starting time points. Preferably, the first mean kinetic energy (MKE) is determined at the beginning of the non-exercise period (N), the last mean kinetic energy (MKE) is determined at the end of the non-exercise period (N), and the intermediate mean kinetic energy (MKE) is determined in regular intervals between the first MKE and the last MKE; the plurality of MKEs constitute the mean kinetic energy period PMKE. Within the non-exercise period (N), the time difference between consecutive time windows (W) and the resulting mean kinetic energy (MKE) may be constant or different. For example, the start times of consecutive time windows (W) may differ by 0.5 to 2 seconds.
[0216] Figure 6 An example of multiple average kinetic energies (MKE) (MKEt0, MKEt1, MKEt2, MKEtp) determined from different time windows (Wt0, Wt1, Wt2, Wtp) during a non-exercise period (N) is shown in A. The starting time (t0, t1, t2, tp) of each time window (Wt0, Wt1, Wt2, Wtp) is incrementally greater than the starting time (t1, t2, tp) of the previous time window (Wt1, Wt2, Wtp). The MKE values at different time points (t0 to t7) within the non-exercise period (N) are shown in a graph as shown in FIG. Figure 6As shown in B, multiple MKE values form a mean kinetic energy period (PMKE). The subject's oxygen consumption rate (VO2) can be determined from the mean kinetic energy period set (SPMKE) of the exercise period (E) ( Figure 6 C).
[0217] During the exercise period (E), a plurality of PMKEs (mean kinetic energy periods) are obtained, one PMKE for each non-exercise period, which is referred to as a PMKE set (SPMKE). The determination of the PMKE for each non-exercise period (N) within the exercise period (E) (the SPMKE) is schematically as follows: Figure 7 and 9 as shown in .
[0218] A measure of the subject's cardiopulmonary fitness can be determined from a set of PMKEs (SPMKE). The SPMKE represents the change in oxygen consumption rate (VO2) over time during an exercise period (E). From the change in oxygen consumption rate (VO2) over time, a maximum VO2 value can be determined, which is a measure of cardiopulmonary fitness, as described elsewhere.
[0219] In particular, the SPMKE can be converted into a linear VO2 function (LVO2F) for a subject, wherein the VO2 function (LVO2F) represents a time function of the subject's oxygen consumption rate (VO2) during an exercise period. The determination of the LVO2F of the SPMKE during an exercise period (E) is schematically shown as follows: Figure 4 , 5 , 7 and 9; Figure 7 and 9 The other transformation steps of SPMKE are shown in detail.
[0220] In order to convert SPMKE to LVO2F, each PMKE in the PMKE set (SPMKE) is converted into a scalar value (single value) representing a non-exercise period (N), wherein the scalar value is an order statistic. Thus, a set of order statistics (SOS) for the exercise period (E) is obtained, and the linear VO2 function (LVO2F) of the subject can be determined based on the progression of the order statistics (OS) of the SOS over time. Figure 5 The use of a set of order statistics (SOS) to determine LVO2F from SPM KE within an exercise period (E) is schematically shown in FIG.
[0221] An order statistic (OS) is determined for each non-exercise period (N). The order statistic (OS) can be the median, or another quantile (eg, the 1st and 3rd quartiles).
[0222] To generate the order statistics (OS) for each non-exercise period (N), each PMKE in the set of PMKEs (SPMKE) is converted to a scalar value (single value). This conversion can be performed as:
[0223] - indirectly for PMKE (by first converting the MKE value within PMKE to another function, such as an intermediate linear function) (e.g. Figure 7 , 8A to 8D), or
[0224] - Directly for PMKE (by sorting the MKE values within PMKE) (e.g. Fig. 9 , 10A to 10E)
[0225] In the case of converting PMKE indirectly to an order statistic (OS), the average kinetic energy period (PMKE) within the non-exercise period (N) is converted to an intermediate linear function (ILF); this can be achieved using a scaling law. For example, each average kinetic energy (MKE) from the average kinetic energy period (PMKE) can be plotted on a graph (y-axis: MKE, x-axis: time), and a curve fit can be performed on it. The fitted curve is converted to an intermediate straight line (the intermediate linear function) by a scaling law. The intermediate linear function can be represented by a graph or by a mathematical equation. The scaling law can be any function that linearizes the curve. In particular, the scaling law can use the cube root of the average kinetic energy (MKE). In particular, the scaling law can use the heart rate (HR) of the subject. An example of a scaling law includes the equation HR*CubicRoot(MKE / HR).
[0226] Fig. 8A A graph showing the change in MKE value (y-axis) from PMKE over time (x-axis, time interval (t0 to t6) within the non-exercise period (N)) during the non-exercise period (E); Figure 8B In the example, the mean kinetic energy period (PMKE) has been converted to an intermediate linear function (ILF) (y-axis: ILF(t), x-axis: time) by means of a scaling law. The intermediate linear function is then filled with blocks of ILF values from non-exercise periods. The ILF values are sorted in order of value (e.g., largest to smallest and vice versa). From the sorted ILF values, the median (median) or quantile can be determined. Figure 8C An example of determining order statistics (OS) (median, 1st and 3rd quartiles) from ILF values ranked (lowest to highest) within an ILF bin determined for a non-exercise period (N) is shown. An order statistic (OS) is determined for each non-exercise period (N).
[0227] In the case of directly converting PMKE to an order statistic, each mean kinetic energy (MKE) within the mean kinetic energy period (PMKE) within the non-exercise period (N) is sorted by value (e.g., largest to smallest, or vice versa). From the sorted mean kinetic energy (MKE) by value, the median (median) or quantile can be determined. Fig. 10B An example of determining an order statistic (OS) (median, 1st and 3rd quartiles) from the mean kinetic energy (MKE) within a mean kinetic energy period (PMKE) determined for a non-exercise period (N) sorted by value (lowest to highest) is shown. An order statistic (OS) is determined for each non-exercise period (N).
[0228] The same order statistics (median or quantile) are determined for each non-exercise period (N), and at least some of these order statistics (OS) form a set of order statistics (SOS) for the exercise period (E). For example, the set of order statistics (SOS) may include only the median order statistic (rather than a mixture of median and quantile order statistics).
[0229] The set of order statistics (SOS) preferably includes only order statistics (OS) that show the linear progress of the subject over time. The order statistics (OS) included in the set of order statistics (SOS) can be determined by the following steps: including the first three order statistics (OS) related to the first three non-exercise periods (N) (excluding the first measurement performed before the first exercise period), and then gradually adding subsequent order statistics, and evaluating whether the new set of order statistics (SOS) has a better linear correlation coefficient (e.g., Pearson correlation coefficient) than the previous set of order statistics (SOS). Whenever the addition of a new order statistic causes the linear correlation coefficient to decrease, the set of order statistics (SOS) is considered to be complete. Typically, the last (in terms of time) order statistic (OS) is not included in the set of order statistics (SOS) because the subject reaches a non-linear point. Advantageously, measuring cardiopulmonary fitness, especially the maximum VO2 based on submaximal effort, is more comfortable (less exhausting) for the subject. Cardiopulmonary fitness measures, especially the maximum VO2, can be determined by the set of order statistics (SOS). As described elsewhere herein, cardiorespiratory fitness measures, particularly VO2max, can be determined by converting a set of order statistics (SOS) into a subject's VO2 function (VO2F).
[0230] In order to convert SOS to LVO2F, a scaling function is used. The scaling function is applied to the progression of the order statistic (OS) of SOS over time, thereby obtaining the subject's linear VO2 function (LVO2F) using the scaling function. The scaling function will be described in more detail below.
[0231] Once the subject's linear VO2 function (LVO2F) is obtained, the value on the linear VO2F curve at the point corresponding to the actual (or estimated) maximum load is determined. This can be based on the observation of a plateau of the order statistic (OS). Alternatively, it can be based on the point on the linear VO2 function (LVO2F) corresponding to the actual (or estimated) maximum heart rate.
[0232] This estimate can be corrected by a constant ('d'). The constant (d) corrects for the subject's recovery during the non-exercise period, which would not occur if the measurement was taken during the exercise period (S).
[0233] This conversion of SOS to LVO2F by means of a proportional function can be performed in the following manner:
[0234] -Directly targeting SOS( Figure 7 , 8D ),or
[0235] - indirectly for SOS (by first transforming the OS values within SOS to another function, such as an intermediate linear function) Fig. 9 , 10D , 10E).
[0236] In the case of direct conversion of SOS to LVO2F, a scaling function is applied to the SOS. For example, the OS values of the SOS can be plotted as the first straight line on a graph (y-axis: OS(t), x-axis: time) and the SOS converted to LVO2F by a scaling function. The VO2 function (VO2F) can be represented graphically or by a mathematical equation. Fig.8D In FIG. 5 , the set of order statistics (SOS) of the exercise period (E) has been converted into the VO2 function (VO2F) by a scaling function.
[0237] In the case of converting SOS indirectly to LVO2F, the set of order statistics (SOS) can first be converted to an intermediate linear function (ILF) of the exercise period (E); this can be achieved using scaling laws. For example, the set of order statistics (SOS) can be plotted on a graph (y-axis: order statistic value, x-axis: time), and a curve fit can be performed on it. The fitted curve is converted to a first straight line (linear function) by scaling laws. The linear function can be represented by a graph or by a mathematical equation. Fig. 10C A graph showing the median (y-axis) versus time (x-axis, non-exercise period (N1 to N7)) during the exercise period (E); Fig. 10DIn , the set of order statistics (SOS) has been transformed into a linear function (LF) by a scaling law (y-axis: ILF(t), x-axis: time). The scaling law can be any function that linearizes the curve. In particular, the scaling law can use the cube root of the order statistic (OS). In particular, the scaling law can use the heart rate (HR) of the subject. An example of a scaling law includes the equation HR*CubicRoot(OS / HR).
[0238] The intermediate linear function (ILF) is then converted to a VO2 function (VO2F) with respect to the subject; this can be achieved using a scaling function. The scaling function will be described in more detail below. The VO2 function (VO2F) represents the subject's oxygen consumption as a function of time during an exercise period (E). For example, the intermediate linear function (ILF) can be plotted as a first straight line on a graph (y-axis: LF(t), x-axis: time), and the LF is converted to a second straight line (VO2 function) by a scaling function. The intermediate linear function (ILF) is converted to a VO2 function (second straight line) by a scaling function. The VO2 function (VO2F) can be represented graphically or by a mathematical equation. In Fig. 10D In the example, the set of order statistics (SOS) of the exercise period (E) has been transformed into a linear function (LF) by means of a scaling law, and the intermediate linear function (ILF) has been transformed into a VO2 function (VO2F) by means of a scaling function ( Fig.10E ).
[0239] As previously described, the set of order statistics (SOS) or intermediate linear function (LF) is converted to the subject's VO2 function (VO2F); this can be achieved using a scaling function.
[0240] The scaling function may include a step of multiplication by a constant (Q). The constant (Q) may be fixed for each subject; it may be different for different subjects.
[0241] An example of a proportionality function includes the equation Q(LF). An example of a function for determining LVO2F is:
[0242] Y(VO2F)=Q(LF)+d
[0243] The value of the constant Q can be determined by measuring a calibration subject using spirometry. After the first measurement, the value of Q and other VO2max measurements can be referenced for the same subject. Alternatively, the value of Q is based on parameters from the subject (height, weight, body composition, aortic cross-sectional area, blood composition). As described elsewhere herein, the value of d corrects for recovery of the subject during non-exercise periods, which would not occur if the measurement was performed during the exercise period (S).
[0244] When referring to determining a kinetic energy data stream (KEDS) for each non-exercise period (N) from the signal stream (120), it will be appreciated that one or more kinetic energy data streams (KEDS) may be generated depending on whether one or more channel outputs are used.
[0245] For example, when only one channel output (e.g., BCG Lin(y), 113b) of the signal stream (120) is used, only one kinetic energy data stream (KEDS) can be determined from the one channel (e.g., BCG Lin(y), 113b). When multiple (two or more) channel outputs (e.g., BCG Lin(x), 113a; BCG Lin(y), 113b; BCG Lin(z), 113c) are used, only one kinetic energy data stream (KEDS) can be determined by combining the channel outputs. For example, one kinetic energy data stream (KEDS) as a single BCG linear kinetic energy (KLin(BCG)) can be determined by combining the BCG channel outputs BCG Lin(x), 113a; BCG Lin(y), 113b; BCG Lin(z), 113c. Other combinations of channel outputs are contemplated: combining KLin(BCG) with: KLin(SCG), and / or combining with KRot(BCG), and / or combining with KRot(SCG). It is contemplated that a kinetic energy data stream (KEDS) may be determined by combining any two of the plurality of channel outputs (113a-f; 115a-f). Exemplary equations for generating a KEDS from a channel output or multiple channel outputs are described elsewhere.
[0246] Therefore, the subsequent steps of determining the mean kinetic energy period (PMKE), order statistics, conversion using scaling laws and proportional functions to obtain VO2 functions are performed using only one channel output (e.g. BCG Lin(y), 113b) from the sensor unit (100) signal stream (120).
[0247] Multiple kinetic energy data streams (KEDS) may be determined from multiple (two or more) channel outputs (e.g., BCG Lin(x), 113a; BCG Lin(y), 113b; BCG Lin(z), 113c). For example, one KEDS may be determined from each of the multiple channel outputs (e.g., one KEDS is determined for each of BCG Lin(x), 113a; BCG Lin(y), 113b; BCG Lin(z), 113c).
[0248] When multiple KEDS are determined, subsequent steps of determining a mean kinetic energy period (PMKE), order statistics, and conversion using scaling laws and proportional functions to derive a VO2 function may be performed on each of the different channel outputs (113a-f; 115a-f) to obtain multiple VO2Fs, one VO2F for each channel output (113a-f; 115a-f). The steps following the generation of multiple VO2Fs may combine the channel output conversions, such as by summing or averaging, to obtain a single function or value, from which a cardiopulmonary fitness measure, particularly a maximum VO2 value, may be ultimately calculated.
[0249] For example, multiple VO2Fs, each generated from a different channel output (113a-f; 115a-f), can be combined, such as by averaging (weighted or unweighted), to obtain a single VO2F. Cardiopulmonary fitness measures, in particular, VO2max values can be determined by combining VO2Fs.
[0250] For example, multiple cardiopulmonary fitness measures, particularly VO2max values, each generated from a different channel output (113a-f; 115a-f)), can be combined, such as by averaging (weighted or unweighted), to obtain a single cardiopulmonary fitness measure, particularly, a single VO2max value.
[0251] The processing steps of the method are performed in vitro.
[0252] The system (100) may include a sensor unit (110).
[0253] Also provided is a computer-implemented method for determining a measure of cardiopulmonary fitness of a subject, in particular maximal oxygen consumption (VO2max), the method comprising steps performed by a processing unit (160) as described elsewhere herein.
[0254] In addition, a computer program or a computer program product is provided, the computer program or the computer program product having instructions, which, when a computing device or a system executes the instructions, will cause the computing device or the system to execute the computer-implemented method.
[0255] Furthermore, a computer-readable medium storing the computer program or computer program product is provided.
[0256] Additionally provided is a data stream representing the computer program or computer program product. Example
[0257] Sixteen healthy subjects (8 males and 8 females) were recruited. The demographic data of the test subjects are shown in Table 1.
[0258] Table 1: Subject demographics represented by median [first quartile; third quartile] for each group (male / female).
[0259]
[0260] The subjects performed a standard VO2max protocol using bicycle spirometry (ESt), as well as an exercise session (E1) as described herein, consisting of alternating exertion segments (S1) and non-exercise periods (N1). Both protocols are described in Table 2 and included Kino recordings in parallel with conventional gas exchange analysis.
[0261] Table 2: Summary of the two solutions
[0262]
[0263] The results of VO2max given by spirometry (see Table 3) showed that the test protocol (E1) gave similar results to the standard protocol (ESt). The Wilcoxon matching test indeed gave a p-value of 1.00, strongly supporting the equivalence of the two protocols.
[0264] Table 3: Results of VO2max measured by spirometry for male and female subjects in protocols E0 and E1. Results are given as median [first quartile; third quartile]
[0265]
[0266] To estimate the VO2max value based on Kino recordings, we chose to focus on the Kinetic Energy Flow (KED) calculated based on the linear axis of the BCG:
[0267] K Lin(x,y,z)(t) =0.5[m((V x(t) 2 )+(v y(t) 2 )+(V z(t) 2 ))]
[0268] During each non-exercise period (N), the mean kinetic energy (MKE) was calculated by integrating the KEDS signal over a 5-second window W and normalized by the length of the window. The mean kinetic energy (MKE) was calculated by moving the window every 1 second within a given non-exercise period (N) and yielding the mean kinetic energy period (PMKE) for each non-exercise period (N).
[0269] The PMKE of each non-exercise period (N) is converted to a linear function (LF) using the following scaling law: LF = HR.CubicRoot(MKE / HR)
[0270] where HR is the heart rate of the subject in the relevant time-window W.
[0271] An LF is determined for each PMKE and, therefore, for each non-training period (N1). The LF of the non-training period (N1) is populated by a block of LF values. An order statistic (OS) is extracted from the block of LF values, the extraction being given by the median of the block of LF values within the LF. For the training period (E1), a set of order statistics (SOS) is obtained, the SOS comprising a plurality of order statistics (medians), each order statistic being obtained from a block of LF values. An example of a set of order statistics (SOS) for the training period (E1) is Fig.12 as shown in .
[0272] exist Fig.12 In the figure, it is apparent that in the linear portion of the curve (the first part), the order statistic increases linearly with workload, and in the second part of the line (towards the end), it becomes nonlinear with respect to workload. By considering the first three non-exercise periods (N1) (excluding the baseline) and calculating the Spearman linear correlation coefficient, we obtain Fig.12 The duration of the linear portion on the SOS curve of . Then, the subsequent non-exercise period (N1) is gradually increased, and the Spearman linear correlation coefficient is calculated again. The linear portion is considered to be completed when a decrease in the Spearman coefficient is observed compared to the previous step.
[0273] For each of the SOS points located in the linear portion, the ratio between the median value of the measured VO2 and OS was determined. The median of these ratios gave the proportionality factor Q. The SOS curve was then converted to a VO2F line by multiplying each value by Q and fitting to a line.
[0274] The VO2F line values were assessed when the subjects reached their maximum workload, providing a preliminary estimate of VO2max.
[0275] The VO2max value is further optimized to correct for a slight underestimation, since it is based on measurements taken during a non-exercise period (N), during which the subject has had some time to recover (depending on the duration of the non-exercise period and the subject's level of fitness). From the spirometry data, a correction factor 'd' is determined for each subject. Optionally, this correction factor is found to be positively correlated with the estimated VO2max value, and can therefore be estimated using a linear relationship.
[0276] The results of VO2max estimation for known maximal workload and correction factor “d” are given in Table 4. These estimates are very accurate, with relative errors mostly below 10%. Fig.13The scatter plot of the estimated and measured values of VO2max is also given in Figure 1. These points are indeed close to the identity line, and this is verified by calculating the linear correlation coefficient: R2 = 0.8914 (p value is 6.6310 -4 ).
[0277] Table 4: VO2max measured by spirometry and Kinmax measured by Kinmax given the maximum workload achieved and correction factor "d". o Comparison of estimated VO2max.
[0278]
[0279] We also verified the impact of missing input information by estimating the maximum workload based on measured maximum heart rate and based on an estimated maximum heart rate (220-age). Each estimation resulted in a decrease in the accuracy of the estimate, but the classification of subjects according to their global fitness level was still very good (results not shown).
[0280] Additional experiments (results not shown) showed that the proportionality factor Q was constant for a given subject.
Claims
1. A system (100) for determining a cardiopulmonary fitness measure of a subject (50), the system (100) comprising a processing unit (160), the processing unit (160) being configured to: - receiving a stream (120) of signals output by a sensor unit (110), said sensor unit (110) comprising one or more motion sensing modules MSM (111), - determining a kinetic energy data stream KEDS of each non-exercise period (N) of an exercise period (E) for gradually increasing the oxygen consumption VO2 of the subject from the signal stream (120), wherein the exercise period (E) comprises a plurality of alternating exertion segments (S) and non-exercise periods (N), -For each non-exercise period (N): - obtaining a mean kinetic energy MKE as a time-averaged kinetic energy value from the time window (W) of the kinetic energy data stream KEDS; - determining a mean kinetic energy period PMKE, said mean kinetic energy period PMKE comprising a plurality of MKEs of different time windows (W) within said non-exercise period (N); - determining the cardiorespiratory fitness measure of the subject from a set of PMKEs SPMKE containing a plurality of PMKEs during the exercise period (E).
2. The system (100) of claim 1, wherein the cardiopulmonary fitness measure of the subject is determined by transforming the SPMKE in the following manner: -Converting the SPMKE into a linear VO2 function (LVO2F) representing the subject's oxygen consumption VO2.
3. The system (100) according to claim 2, wherein the conversion of the SPMKE to the LVO2F comprises the following steps: converting each PMKE of the SPMKE into a scalar value representing the non-exercise period (N), wherein the scalar value is an order statistic.
4. The system (100) according to claim 3, wherein the conversion of each PMKE of the SPMKE to a scalar value is performed in the following manner: - indirectly for the PMKE, by first converting the MKE values within the PMKE into an intermediate linear function (ILF) and then sorting the values within the ILF to obtain the order statistic, or - Directly for the PMKE, by value sorting the MKE values within the PMKE to obtain the order statistic.
5. The system (100) of claim 4, wherein the same order statistics are determined for each non-exercise period (N), at least some of these order statistics (OS) forming a set of order statistics SOS for the exercise period (E), and the SOS are converted into the subject's LVO2F.
6. The system (100) according to claim 5, wherein the conversion of the SOS to the LVO2F is performed by: -directly to the SOS, or - indirectly to the SOS, by first converting the order statistic (OS) value within the SOS to an intermediate linear function (ILF), and then converting the intermediate linear function (ILF) to the LVO2F of the subject.
7. The system (100) according to any one of claims 1 to 6, wherein the order statistics (OS) are medians or quantiles, and the order statistics (OS) within each set of order statistics (SOS) are all medians or all quantiles.
8. The system (100) according to any one of claims 1 to 7, wherein the exercise period (E) brings the subject to exhaustion, preferably to complete exhaustion.
9. The system (100) according to any one of claims 1 to 8, wherein the time window (W) is a time period of w seconds, and the value of w is greater than the duration of one cardiac cycle of the subject.
10. The system (100) of claim 9, wherein the value of w is the same during each non-exercise period (N) and during an exercise period (E).
11. The system (100) according to any one of claims 1 to 10, wherein the cardiorespiratory fitness measure of the subject (50) is a maximum VO2 measure of the subject (50).
12. The system according to any one of claims 1 to 11, wherein one of the MSMs (111) is a shockcardiograph, SCG (114), and optionally the other MSM (111) is a ballistocardiograph, BCG (112).
13. A computer-implemented method for determining a measure of cardiopulmonary fitness of a subject (50), the computer-implemented method comprising: - receiving a stream (120) of signals output by a sensor unit (110), said sensor unit (110) comprising one or more motion sensing modules (111), - determining a kinetic energy data stream KEDS of each non-exercise period (N) of an exercise period (E) for gradually increasing the oxygen consumption VO2 of the subject from the signal stream (120), wherein the exercise period (E) comprises a plurality of alternating exertion segments (S) and non-exercise periods (N), -For each non-exercise period (N): - obtaining a mean kinetic energy (MKE) as a time-averaged kinetic energy value from a time window (W) of the kinetic energy data stream KEDS; - determining a plurality of mean kinetic energies (MMKE) for different time windows (W); - determining an order statistic (OS) from said plurality of mean kinetic energies (MMKE), - determining a set of order statistics (SOS) comprising at least some of said order statistics (OS) determined from a plurality of non-exercise periods (N) of said exercise period (E); - determining the cardiorespiratory fitness measure of the subject from the set of order statistics (SOS).
14. A method according to claim 13, comprising a restriction as defined in any one of claims 2 to 11.
15. A computer program or a computer program product having instructions which, when executed by a computing device or system, cause the computing device or system to execute the computer-implemented method of claim 13 or 14.
