WO2021148921A1

WO2021148921A1 - A medical system and method using a pair of gloves equipped with physiological sensors

Info

Publication number: WO2021148921A1
Application number: PCT/IB2021/050332
Authority: WO
Inventors: Dennis Majoe; David Atienza Alonso
Original assignee: Ecole Polytechnique Federale de Lausanne EPFL
Current assignee: Ecole Polytechnique Federale de Lausanne EPFL
Priority date: 2020-01-21
Filing date: 2021-01-18
Publication date: 2021-07-29
Anticipated expiration: 2022-07-21

Abstract

A medical system comprising a first wearable device configured to be worn by a first hand of a user, a second wearable device configured to be worn by a second hand of the user, a first sensor device arranged on the first wearable device, the first sensor device configured to measure a first physiological parameter from a living being, a second sensor device arranged on the second wearable device, the second sensor device configured to measure a second physiological parameter from a living being, and a data processing device configured to obtain data from the first sensor device and the second sensor device, and configured to process the data to obtain information characterizing a health and body status of the living being.

Description

A MEDICAL SYSTEM AND METHOD USING A PAIR OF GLOVES EQUIPPED WITH

PHYSIOLOGICAL SENSORS

CROSS REFERENCE TO RELATED APPLICATIONS

(0001) The present patent application claims priority to International patent application with the Serial No. PCT/IB2020/050438 that was filed on January 21, 2020, the entire contents thereof herewith incorporated by reference.

FIELD OF THE INVENTION

(0002) The present invention is directed to the field of wearable diagnosis and measurement systems and devices, and methods of performing diagnosis and measurement on living beings by the user of the wearable system, for example by the use of a pair or wearable gloves to be worn by the user, and also directed to the field of wearable computing and mobile data processing for determining a health or body status of a living being. BACKGROUND

(0003) The field of wearable computing has recently seen rapid growth because products have been made available, such as wearable lifestyle products that provide a new interesting dimension the lives of the users. However, recent trends have shown that many of the wearable computing features have been integrated into smartphones as free applications, such that the original wearable computing device has become obsolete and has no more commercial value. Generally, user are used to wear computing devices as a watch, a bracelet, or a ring, for example for sports tracking, smart phone accessory, and health measurement applications, that also include a fashion element or status statement, but it has been shown that users prefer not to wear other types of wearable computing devices, especially if they do not have a fashion element to it.

(0004) However, when a person is diagnosed with illness, the living being is more inclined or ready to wear a computing device, mostly on the basis that the wearable device will provide for some measurements or therapeutic function that will, in some way, lead to healing or at least provide for data for the healing and curing process. From a market point of view, the future for commercially sustainable wearable technology may he only where the potential user see a distinct advantage in using them. Currently, different types of wearable computing devices have shown up on the market place, other than a smartphone or other type of computing devices that can be placed in a pocket of a clothing item, such as items that can at least partially integrate to the form of a living being, including smart eyeglasses, gloves worn on the hands, a cap worn on the head, a belt around the waist, and a cuff around the elbows or knees. These devices can be worn for a short time by a living being, and can be easily taken off at any given moment in time.

(0005) Wearable computing devices that need to be worn for a longer period of time, they need to benefit the user in a somewhat substantial way, and preferably also need to be very comfortable or even barely noticeable when worn by the user. For example, we suggest that the benefit is that the wearable computing device will improve different capabilities of the user or protect them along some dimension, and in many cases this is achieved by augmenting their senses. While the field of wearables is very wide, in the field of wearable devices for medical or therapeutic purposes, for example one that are worn by the hand of a user or medical person, there are only few devices and systems that have been proposed, for relatively simple tasks.

(0006) For example, in U.S. Patent Publication No. 2014/0330087, this patent publication herewith incorporated by reference in its entirety, a medical device and method has been proposed, having a body configured to secure to a hand of a user and having a finger covering, and having at least one sensor positioned on the body, including one or more sensors positioned at a distal end of the at least one finger covering for obtaining data characterizing one or more physiological conditions of a patient. This reference is herewith incorporated by reference in its entirety. The medical device is used for determining at least one of blood pressure, blood flow and heart rate characteristics of the patient.

(0007) As another example, in U.S. Patent Publication No. 2017/0000370, this patent publication herewith incorporated by reference in its entirety, an Electrocardiogram (EKG, sometimes referred to as ECG) system has been proposed, consisting of at least one glove or glove pair, this reference herewith incorporated by reference in its entirety. The at least one glove or glove pair can include at least one electrode. The at least one electrode is configured to detect two or more electrical potentials at two or more different of surfaces of a subject.

At least two electrical potentials can be detected at different times. Moreover, the glove- based ECG system further includes a controller configured to control all or some components of the ECG system 100.

(0008) In yet another example, Chinese Patent No. CN105455788A, a measuring glove has been provided, the measuring glove including a first sensing unit and a second sensing unit, the first sensing unit is used for measuring a first physiological signal, and the second sensing unit is used for measuring a second physiological signal, when the first sensing unit acts, the second sensing unit stops acting, when the second sensing unit acts, the first sensing unit stops acting. According to this measuring glove, detection on multiple different physiological parameters can be achieved.

(0009) However, all of these systems lack certain specific functionality for more complex measurement tasks, and health and body status diagnosis, other than simple measurements. Accordingly, despite all the advancements in the field of wearable devices, and specifically in the field of wearable devices for medical or therapeutic purposes, there is strong desire to provide for advanced and more sophisticated solutions and methods.

SUMMARY (00010) According to one aspect of the present invention, a medical system is provided. Preferably, the medical system includes a first wearable device configured to be worn by a first hand of a user, a second wearable device configured to be worn by a second hand of the user, and a first sensor device arranged on the first wearable device, the first sensor device configured to measure a first physiological parameter from a living being. Moreover, preferably, the medical system further includes a second sensor device arranged on the second wearable device, the second sensor device configured to measure a second physiological parameter from a living being, and a data processing device configured to obtain data on the first and second physiological parameters from the first sensor device and the second sensor device, and configured to process the data to obtain information characterizing a health and body status of the living being.

(00011) According to another aspect of the present invention, a method for determining an electrocardiogram from a living being is provided, comprising the steps of measuring a superposed electrocardiogram signal, the superposed electrocardiogram signal including an electrocardiogram signal of the living being and an electrocardiogram signal from an operator, subjecting the superposed electrocardiogram signal to a neural network to extract the electrocardiogram signal of the living being, wherein the neural network has been trained by using an error function that minimizes the error of the electrocardiogram signal of the living being.

(00012) The above and other objects, features and advantages of the present invention and the manner of realizing them will become more apparent, and the invention itself will best be understood from a study of the following description with reference to the attached drawings showing some preferred embodiments of the invention.

BRIEF DISCUSSION OF THE SEVERAF VIEWS OF THE DRAWINGS (00013) The accompanying drawings, which are incorporated herein and constitute part of this specification, illustrate the presently preferred embodiments of the invention, and together with the general description given above and the detailed description given below, serve to explain features of the invention.

(00014) FIGs. 1A shows an exemplary and schematic view of the medical system 100, including a first and second wearable device 10, 20 in the form of a pair of gloves that can communicate with each other via wireless communication ports 60, 70, and FIG. IB shows a simplified schematic overview of a medical system 100 that is communicating with an external data processing system 200 for data analysis and recording;

(00015) FIG. 2 shows an exemplary and schematic view of a block representation of medical system 100 with different data input, output and data processing elements;

(00016) FIGs. 3A and 3B show a simplified and exemplary schematic view of medical system 100, showing patient L/P and user U, and a schematic representation of first and second wearable devices 10, 20 with the electrodes 42, 52.1, 52.2, wearable devices 10, 20 worn by user U, with FIG. 3A showing a schematic representation of first and second wearable devices 10, 20 with the electrodes 42, 52.1, 52.2 depicting user U and patient L/P, and FIG. 3B showing an exemplary representation of medical system 100 including a first and second wearable device 10, 20, with an electrode 42 of first wearable device 10 being in contact with a skin portion of the living being or patient P and the user U at one hand, and a second electrode 52 of the second wearable device 20 being in contact with a skin portion of the living being or patient P and the user U at the other hand, to measure a superposed or combined electrocardiogram (ECG) signal of both P and U, and FIG. 3B (00017) Herein, identical reference numerals are used, where possible, to designate identical elements that are common to the figures. Also, the images are simplified for illustration purposes and may not be depicted to scale. DESCRPTION OF THE SEVERAL EMBODIMENTS

(00018) FIG. 1A shows an exemplary and schematic view of the medical system 100, including a first and second wearable device 10, 20 in the form of an exemplary pair of soft gloves that can be worn on each hand of a user U, for example a nurse, medical staff personnel, operator, physician that will be treating or diagnosing or testing a living being L, for example a human patient or diagnosed person. Wearable devices 10, 20 need not be in the form of a pair of gloves, but include and are not limited to a plurality of finger or thumb coverlets, fingerless hand protectors, partial gloves covering not all of the five (5) fingers, wrist covers, and can include a combination of a glove and another device that only covers the hand of user U partially. In the variant shown, first wearable device 10 is configured for the left hand of a user U, and second wearable device 20 is configured for the right hand of the user U, having finger coverings for all five fingers of both hands. First and second wearable device 10, 20 form part of medical system, and can communicate with each other via wireless communication ports 60, 70 and communication antennas 62, 72 associated thereto, each wearable device 10, 20 having a plurality of sensors 40, 50. At least one of the first and second wearable devices 10, 20 includes a data processing device 30, for example a microcontroller, microprocessor, data processing unit or other type of data processing device, that is associated with memory and other hardware elements for receiving or otherwise obtaining data from the sensors 40, 50, for example directly from the respective wearable device 10, 20 that includes data processing device 30, in the variant shown wearable device 20, and via the wireless data link 75 by wireless communication ports 60, 70 from the other wearable device 10, 20, in the variant shown wearable device 10. In this respect, system can include a master wearable device 20 with data processing device 30, and a slave wearable device 10. However, it is also possible that both wearable devices 10, 20 are each equipped with a data processing device 30. (00019) FIG. IB show the integration of medical system 100 with the two wearable devices 10, 20 that are worn on each hands of a user U, and showing a living being L as the patient or person to be examined or tested. Medical system 100 can connect and exchange data with an external data processing system 200, for example via a data communication link that between one of the first or second wearable devices 10, 20 via communication port 222, and communication port 220 that is in operative connection with data processing system 200, for example a server, remote data processing device, personal computer, Macintosh computer, or other type of data processing device. Data processing system 200 can have a computer screen or monitor 210 and one or more input devices such as but not limited to keyboard, touchscreen, mouse, trackpad, for data input, and screen 210 can be used to visualize data to a health professional. Also, data processing system 200 can be used for recording of the different data.

(00020) Preferably, both first and second wearable device 10, 20 will share a certain number of sensor devices 40, 50 that measure the same type of physiological signal, for example a first physiological parameter, but can also have some sensor devices 40, 50 that measure different types of physiological signals, thereby providing for different measurements from one hand to the other hand of the user, for example to measure a second physiological parameter that is different from the first one.

(00021) For example, both first and second wearable devices 10, 20 can include five or less electrodes as sensors 41, 51 at at least some of the finger tips of each one of the wearable devices 10, 20, each electrode galvanically separated from each other, all of these electrodes being in operative connection with data processing device 30, either directedly or via wireless data link 75. These electrodes 41, 51 are places such that the user place his fingers in a natural way to the skin of a living being whilst wearing the first and second wearable devices 10, 20, and thereby an electric potential of the touched portion of the skin can be measured, at multiple locations, to measure the electric potential of the skin of the living being L as a first physiological parameter. Also, each wearable device 10, 20 can have a central palm electrode 42, 52, that is placed on an outer surface of wearable device 10, 20, that allows to provide for a large sensing surface of skin of user. Thereby, user U that is wearing wearable devices 10, 20 can place the palm of each hand to a skin of the living being, to measure a timely evolution of a potential between central palm electrodes 42, 52.

(00022) In addition, the first and second wearable devices 10, 20 can have sensor device

40, 50 that measure different types of physiological signals from living being L, thereby providing for different measurements from one hand, for example the left hand, to the other hand, for example the right hand, of the living being L. For example, wearable device 10 can have a sensor 40 that does not exist or is different than sensor 50 on wearable device 20, or vice versa. For example, depending on the handedness of user U, in other words whether the user U is right-handed or left-handed, one of the wearable devices 10, 20 that corresponds to the dominant hand can be equipped with a specific sensor 54, 56 that can be used by user U for specific measurements that will require to pointing or otherwise directing specific sensor 54, 56 to a skin area of the living being L, to measure a second physiological parameter that is different from the first one. For example, an optical sensor that requires pointing or directing to living being L can be placed on the finger coverings of the index finger of the glove that forms wearable device 20, in this case of a right-handed user U, for example an optical sensor 56 for detecting photoplethysmogram (PPG) that can be used to detect blood volume changes in the microvascular bed of tissue of living being L. This allows user U to hold exposed skin of living being L with both hands, by capturing first physiological signals or parameters from both first and second wearable device 10, 20 by at least some of sensors

41, 51, 42, 52, for example by holding the two hands of living being L, being the electric potential of a point of contact of sensors 41, 51, 42, 52, for example for electrocardiogram (ECG) measurement by sensors 41-52, and simultaneously, user U can point his index finger to point PPG sensor 56 to an exposed location on the skin of the living being L, for example at the back of the hand of L, to measure the PPG as a second, different physiological parameter. With this arrangement, and by user U holding both hands of living being L via first and second wearable devices 10, 20, two different types of measurements can be performed for capturing two different physiological parameters simultaneously (in this example ECG and PPG), and its data analyzed by data processing device 30. In this respect, user U can have a calming effect on living being L by holding hands, while simultaneously a variety of health and body status data can be captured, without the need of connection living being L to a variety of measurement devices and machines.

(00023) In another variant or in combination with the PPG sensor, a peripheral capillary oxygen saturation (Sp02) sensor can be arranged on either one of wearable devices 10, 20 to measure oxygen-saturated hemoglobin relative to total hemoglobin (unsaturated + saturated) in the blood of the living being L. For example, Sp02 sensor 46 could be arranged on the finger covering of middle finger of the first wearable device 10. This could allow user U to use his left hand with middle finger and thumb as a tweezer or clamp to hold a skin portion of the hand or other body part of living being L, for the Sp02 sensor 46 being able to make a measurement of another phycological parameter, for example the third physiological parameter being the oxygen-saturated hemoglobin. At the same time, with the left hand of user U, it is still possible to have at least some electrodes of sensors 41, 42 in contact with the skin for the ECG measurement.

(00024) Another sensor that can be part of wearable devices 10, 20 could be a camera device for capturing different types of images. For example, a camera 44 could be placed at a side of the finger covering of the index finger of first wearable device 10, that can be pointed by user U to record images from living being L. For example, the camera 44 could be based on the Omnivision OV2720 imaging chip which allows us to make a camera and light compartment having a small footprint or size, for example in a 6mm cube. In this respect, images captured of living being L can be considered another physiological parameter that can be captured from living being L for further data processing and analysis.

(00025) Another type of sensor that can measure an additional physiological parameter is a contactless or contact-based temperature sensor, to measure the body temperature, the skin temperature, or both. For example, the temperature sensor can include but is not limited to an infrared or near-infrared temperature sensor or thermistors. An example of a contactless temperature sensor is described in U.S. Patent No. 7,364,356 or in Int. Pat. Pub. No. W02002103306, this reference herewith incorporated by reference in its entirety.

(00026) In a variant, the communication link 75 may not be strictly speaking a wireless connection, but still does not require any additional cables or links between wearable devices 10, 20. For example, communication link 75 could be a cable or wire that links the wearable device 10, 20 together, or communication link 175 could be at least partially formed by the body of the wearer or user U of wearable devices 10, 20 itself, taking advantage of the conductivity of the body of a user U, as visualized in FIGs. 3 A and schematically shown in FIG. 3B. A closed loop electrical circuit is formed by at least one sensor 40, 50 of both wearable devices 10, 20. For example, electrodes can be in electrically conductive contact with both user U and the living being L, for example at least one of electrodes 41 or 42 and one of electrodes 51 or 52, such that these electrodes that can contact both the skin of the user U or wearer that is wearing wearable devices 10, 20 on her or his hands, and the skin of the living being L that is being touched by a respective two electrodes, as shown in FIG. 3A. (00027) Specifically, as schematically illustrated in FIG. 3B, two wearable devices 10, 20 are shown, for example a pair of gloves, with a living being L or patient P being held at the left side by the wearable device 10, 20, and at the right side, a user or operator U is wearing the wearable device 10, 20. For representation purposes, wearable devices 10, 20 are shown as H-shaped elements lying down, with an open area on the left symbolizing a space where the living being or patient P will be held, for example the outer surface 11 of a glove, while the open area on the right symbolizing a space where the user or operator U will put his hands into, for example the inside are of the glove, having an inner surface 12. An an electrode 42 of wearable device 10 can be exposed to an inner area or surface 11 of the wearable device 10, surface 11 being a surface that is in contact with living being L or patient P, and at the same time electrode 42 can also be exposed when user or operator U, for example a nurse, physician, or clinician that holds that holds the two hands of a patient L with her two hands. It is also possible that the entire wearable device 10 is conductive, for example by being woven of conductive material or filaments, or by having glove-traversing conductive filaments or conductors, for example gloves or types of gloves that are used to operate touch-screens, providing for an electric connection from touch screen to the finger of the user. In this respect, electrode 42 of wearable device 10, or wearable device 10 itself can form an electric contact or connection between the skin of user U and living being L.

(00028) In another variant of this embodiment, it is possible that system 100 only has one wearable device 20 and having the two electrodes 52.1 and 52.2, and instead of having a wearable device 10 with interconnecting electrode 42, a hand of user U and living being L or patient P are in touch with each other, such that an electrical connection is formed of each other’s bodies, via the human touch between user U and living being L. The electric interconnection of the hands of the user U and living being L could be further enhanced by using a conductive paste or gel applied on either one or both of hands of the user U and living being L.

(00029) At the other wearable device, two electrodes 52.1 and 52.2 are arranged, one 52.1 being exposed to the outer surface 21 of wearable device 20, for example an outer surface of glove that will be in touch with living being L, and the other one 52.2. being exposed to the inner surface 22 of wearable device 20, for example the inner surface of a lining of a glove. Unlike electrode 42, electrodes 52.1 and 52.2 will be electrically insulated from each other. Each electrode 52.1 and 52.2 are interconnected with an analog amplifier A1 and A2, respectively, for example provided at the wearable device 20 itself, or at an external device 300 with analog connection cables. Analog amplifiers A1 and A2 can lead to the inputs of an analog differential amplifier DA, to provide for an analog signal at its output that includes an ECG signal ECGSUP from both the user U and the patient P that is superposed. This signal can be digitally converted by an analog-to-digital converter AD, and then provided to a microprocessor MP for further digital data processing, so that the ECG signal of the patient P can be determined, as further discussed below. In FIG. 3B, this data processing environment is shown as a separate device 300, but it can also be part of the wearable device 10 or 20 itself, for example as an embedded computer with the signal processing elements Al, A2,

DA, AD, MP, and additional devices such as but not limited to filters, memory, communication interfaces. Also, in a variant, it is also possible that the signals from the outputs of the analog amplifiers Al, A2 are amplified and thereafter digitally converted, without the use of an analog differential amplifier DA.

(00030) Moreover, wearable device 10 is equipped with a PPG sensor 46 that can optically scan or measure the skin of living being L or patient P, and wearable device 20 is equipped with a PPG sensor 56 that can optically scan or measure the skin of user or operator U. It is also possible that the PPG sensors are both arranged on one wearable device 10, 20, for example on wearable device 20 together with the data processing device 30. The output signals PPGL and PPGu of the PPG sensors 46, 56 are provided to microprocessor MP for further data processing, with an operative connection by an analog signal, a digital signal, wired or wireless connection. Other sensors can be arranged at wearable devices 10, 20, as explained above. Moreover, with the electric connection between user U and living being L via electrode 42, a bodily communication link 175 is provided, that allows to capture two superposed signals from user U and living being L or patient P at electrodes 52.1 and 52.2 that are arranged at the other wearable device 20.

(00031) With the configuration of system 100 with two wearable device 10, 20, for an electrocardiogram (EKG or ECG) measurement, a dual diagram can be captured and further processed, showing the dual nurse/patient ECG potential. Because the body of the nurse forms part of the electrical loop or connection 175 the ECG will contain both the nurse’s ECG signal and the patient’s ECG signal. In this respect, communication link 175 between wearable devices 10, 20 can be said to be a purely analog signal communication, by measurement and transmitting the ECG signals of the patient and the user (e.g. nurse, physician, operator, wearer, or user U) in addition to other types of data communication that can exist between the two wearable devices 10, 20 and data processing device 30. By the electrical connection provided by electrode 42 and using the corpus of the user U and living being L or patient P as a conductor, two superposed ECG signals can be captured. In this respect, the voltage that is measured across electrodes 52.1 and 52.2 that is located on one wearable device 20 will include a superposition of the two ECG signals, with body of both wearer or user U and patient P and electrode 42 serving as an electric conductor or conductive line between wearable devices 10, 20 for communication link 175.

(00032) According to an aspect of the invention, medical system 100 including the two wearable devices 10, 20 can be configured to measure the superposed ECG of the wearer or user U and the patient P, and at least one of the processors 30, 200 can be configured to extract the data of the ECG of the patient. For example, the method that is performed on processor 30, 200, or both can include a step of measuring the PPG signal of living being L or patient P, and the PPG signal of the user or operator U, for example with a PPG sensor 46 on one of the gloves that forms wearable device 10 that is directed to living being L or patient P, and a PPG sensor 56 on the other one of the gloves that forms wearable device 20, or on the same glove that measures the PPG signal of the user or operator U, processing the data of the PPG signal of the operator or user U with a first matched filter, processing the data of the PPG signal of the patient P with a second matched filter, and using the filter data of the first and second matched filter to determine the PQRST period of the heart beats of both the wearer or user U and the patient P, PQRST being one complete heartbeat in the ECG of the person, with the P-wave, QRS complex, and the T-wave. In another data processing step, data of the first and second matched filter can also include true and false signals in real time, that can be used to identify when the data from the electrodes 52.1 and 52.2. needs to be processed to apply different types of algorithms, for example to make heart rate measurements and other types of ECG signal processing. A simplified schematic of this measurement method and the system is shown in FIG. 3B. The differential amplifier DA, analog to digital converter AD and microprocessor MP can be part of the processing device 30 of one or both wearable devices 10, 20, or can be part of an external data processing system 200, for example a local data processing device such as a PC, or a remote server. (00033) According to another aspect, other or additional data processing of the electrode voltage potentials of electrodes 52.1 and 52.2 can be done, for example to apply a pattern matching algorithm to recognize elements from the superposed ECG signals from both user U and patient P, and them removing the part, portion or elements of the ECG signal that belong to the user U, to perform cancellation of these parts from the superposed ECG signal. Thereby, the resulting signal is the ECG signal of patient P, and can be used for further processing by processors 30, 200. The method can be further optimized by performing a step of measuring the PPG of either patient P, user U, or both, to provide for additional date for the pattern matching algorithm. (00034) With the herein presented measuring and data processing method, the goal is to isolate the ECG signal of living being L or patient P, the ECGPAT from the superposed ECGSUP, that includes ECGPAT, ECGUSER, and noise. Generally, the ECGSUP will be a signal that is degraded by noise including slow and fast drifting offsets as well as 50/60Hz electrical interference and Gaussian noise. The ECG waveform of a human being is rather particular to the individual person, so much that it can actually be used as a biometric for identification alongside other biometrics. See e.g. Krasteva, Vessela, Irena Jekova, and Roger Abacherli. “Biometric verification by cross-correlation analysis of 12-lead ECG patterns: Ranking of the most reliable peripheral and chest leads,” Journal of Electrocardiology, Vol. 50, No. 6, 2017, pp. 847-854, this reference herewith incorporated by reference in its entirety. It is very unlikely that the ECG PQRST waveform of living being L or patient P looks identical to the clinicians and therefore methods of specific pattern recognition are proposed to identify and isolate the ECGPAT. When two signals are superimposed, for example when music is played with two different instruments, it is possible to use deep learning, for example by Deep Neural Networks DNN, based methods to extract each signal individually. This is done by effectively training the DNN to isolate power spectra when the input signal is subject to Fourier transformation, for example by Fast Fourier Transform (FFT) or Discrete Fourier Transform (DFT), and to inverse Fourier transformation of the data to reproduce a signal that excludes one of the two musical instrument. These principles can be applied to the data processing of ECG signals, also the superposed ones with ECGSUP. Because FFT and DFT are in effect networks of time delayed samples, the method lends itself well to DNNs. See for example Uhlich, Stefan, Franck Giron, and Yuki Mitsufuji, “Deep neural network based instrument extraction from music,” IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 2135-2139, IEEE, 2015, this reference herewith incorporated by reference in its entirety. (00035) However, the ECG signals are not like music and the ECG signals occur in repeating short moments like a chirp of a bird, occurring as a short burst between periods of silence. Detecting the occurrence of specific looking chirps has been achieved for many years using matched filters which correlate a template with the input signal and has been successful to extract signals buried in high noise levels, which is the case of the ECG here. Accordingly, the proposed method is similar to the ones used for radar reflections, sonar reflections and so on. The matched filter is simply a network of nodes and coefficients whose output rises if and when the signal enters the filter. DNNs therefore have the potential structure to be trained for detecting specific shaped signals when arriving as chirps. DNNs can be trained in supervised learning by comparing the output of the network (that is ideally a reconstruction of the ECGPAT) to a training marker and if in error, the error is used to drive changes in the network’s coefficients until the DNN reliably generates an output in sync with the marker. An example of such DNN training approach can be found in Gabbard, Hunter, Michael Williams, Fergus Hayes, and Chris Messenger, “Matching matched filtering with deep networks for gravitational-wave astronomy,” Physical Review Letters, Vol. 120, No. 14, year 2018, p. 141103, this reference herewith incorporated by reference in its entirety. In the method for determining the ECGPAT from the ECGSUP, DNNS can be applied to the superposed signal ECGSUP (that also includes noise) to detect the occurrence of the ECG PQRST waveform or chirp of the user or operator U, and the ECG PQRST waveform or chirp of the patient P or living being L.

(00036) The DNN can be trained using different methods. In a first method, one presents the ECGSUP waveform to the DNN off line, where the ECGSUP waveform data is manually marked up with a signal demarking the beginning and ending of the different PQRST waveforms. The error between the output of the DNN and the demarker is used to drive and modify the coefficients of the DNN so as to reduce the error to a minimal value, for example a predefined threshold. A second method is to perform the same supervised learning method but using the measured blood flow signals, for example the PPG signal as an automatic marker from both the user or operator U, and the patient P or living being L, instead of the manual defined demarcation. Because the PPG and ECG occur in real time, this training can be done in real time saving any manual effort. Although the PPG signal is delayed slightly this delay can be simply handled by the DNN as a systematic delay. A third method takes into account more factors that can influence the automatic PPG marking method. In order to take account of the movement of the hand of user or operator U, and the patient P or living being L, Inertial Measurement Units (“IMU”) providing data is integrated into the DNN inputs, as well as the persons age, gender and temperature. For this purpose, the wearable devices 10, 20 can be equipped with one or more IMUs, or the IMU can be part of processor MP, and are in operative connection with the processor MP or other data processing device. (00037) If two DNNs are trained, one using data of the user U or operator, and one using data of the patient P or living being U, where data can include but is not limited to PPG data, IMU data) then we can use the trained DNNs to indicate when the PQRST waveform of one or the other is occurring. Using these outputs we can segment the data so as to arrange non overlapping patient signals P and user signals U, for ECGPAT et ECGUSER. However, in the case that the heartbeats of the two people (user U and living being L) are precisely in phase, overlapping and the same heart rate, this method may be insufficient. However it is possible to detect this situation, and then the system can indicate to the user or operator U that it is occurring and the user or operator U could simply take a deep breath of air and this would undoubtedly modify the phase relationship immediately. This could be done by a screen, indicator light, audio buzzing, or other type of signal in operative connection with the processor MP that could be generated by the processor MP, such that the user or operator U is alerted. (00038) Once we have the separated ECG waveforms, further processing can be performed using different types of ECG waveform processing methods, and data can be send to a remote data processing device 200 for further processing and storage. In some cases, the time and effort taken to train the DNN to detect the ECGPAT of the patient takes too long or is inconvenient. In such situation, the following method can be proposed. A DNN is trained as above but for a large number of different people instead of just one patient. In this case the DNN can generalize the result and can show a matched response across a large number of people, effectively it is detecting using a broad template and detects the PQRST period for any human. This generalized DNN can now be used as before and applied to the patient's signals. In this case the generalized DNN will generate an output for both the patient P and the user U, for example a clinician. However, since the clinician's DNN is precisely trained to only the clinician's ECG, we can logically separate out the signals as before. Because the generalized DNN has a broad recognition, the accuracy in noise may be less than the personalized trained DNN.

(00039) In sum, with the herein presented system and method, we will be integrating sensors and actuators into body conforming wearable devices 10, 20, for example but not limited to a pair of gloves, specifically with the aim of augmenting human capability required for carrying out a task or maintaining safety, and this can be achieved by proposing two gloves that are worn on the hands of a user, operator, doctor, physician, so that the patient can be touched by two hands simultaneously. The collaboration of different sensors on two different gloves that are worn by a single user presents specific advantages, with respect to the capturing of a large diversity of data from the patient, by the operator or user touching the patient on his or her skin with both hands.

(00040) While the invention has been disclosed with reference to certain preferred embodiments, numerous modifications, alterations, and changes to the described embodiments, and equivalents thereof, are possible without departing from the sphere and scope of the invention. Accordingly, it is intended that the invention not be limited to the described embodiments, and be given the broadest reasonable interpretation in accordance with the language of the appended claims.

Claims

1. A medical system comprising: a first wearable device configured to be worn by a first hand of a user; a second wearable device configured to be worn by a second hand of the user ; a first sensor device arranged on the first wearable device, the first sensor device configured to measure a first physiological parameter from a living being; a second sensor device arranged on the second wearable device, the second sensor device configured to measure a second physiological parameter from a living being; and a data processing device configured to obtain data on the first and second physiological parameters from the first sensor device and the second sensor device, and configured to process the data to obtain information characterizing a health and body status of the living being.

2. The medical system according to claim 1, wherein the data processing device is arranged on the first wearable device and/or on the second wearable device.

3. The medical system according to claim 2, further comprising: a wireless communication link between the first wearable device and the second wearable device, for sending the data from the first wearable device to the second wearable device and vice versa.

4. The medical system according to claim 1, wherein the first sensor device is a first electrode, and the second sensor device is a second electrode, the first and the second electrodes configured to be both in contact with (i) a skin of the user that is wearing the first and second wearable devices, and in contact with (ii) a skin of the living being when touched by the first and second wearable devices by the user.

5. The medical system according to claim 4, wherein a skin of the body of at least one of the user and the living being is in contact with both the first electrode and the second electrode, the body of at least one of the user and the living being forming a communication link between the first wearable device and the second wearable device.

6. The medical system according to claim 4, wherein the data processing device is configured to measure a voltage signal across the first and the second electrode, the voltage signal including the ECG signal of both the user and the living being, and is further configured to extract the ECG signal of the living being from the voltage signal measured across the first electrode and the second electrode.

7. The medical system according to claim 4, wherein the contact (i) and the contact (ii) form a communication link between the first wearable device and the second wearable device, for sending the data from the first wearable device to the second wearable device and vice versa.

8. A method for determining an electrocardiogram from a living being, comprising the steps of: measuring a superposed electrocardiogram signal, the superposed electrocardiogram signal including an electrocardiogram signal of the living being and an electrocardiogram signal from an operator; subjecting the superposed electrocardiogram signal to a neural network to extract the electrocardiogram signal of the living being, wherein the neural network has been trained by using an error function that minimizes the error of the electrocardiogram signal of the living being.

9. The method of claim 8, wherein the step of subjecting further comprises: using a matched fdter to detect signal bursts of the superposed electrocardiogram signal.

10. The method of claim 8, wherein the step of subjecting further comprises: apply a pattern matching algorithm to recognize signal elements from the superposed electrocardiogram signal,