Respiratory system symptom sensing system and device

Abstract

A respiratory system condition sensing system includes a patch and a processor. The patch comprises at least one vibrator and a plurality of receivers. Each of the plurality of receivers generates a reference vibration signal corresponding to the at least one vibrator and a vibration signal corresponding to the respiratory system of the living body. The processor determines the relative position of each receiver relative to the respiratory system of the living body according to the plurality of reference vibration signals, the plurality of vibration signals and a plurality of position relations of the at least one vibrator corresponding to the plurality of receivers. The processor generates a spatial position corresponding to the respiratory system and a corresponding symptom type according to the plurality of vibration signals and the plurality of relative positions corresponding to the plurality of receivers by using a classification model. According to the respiratory system symptom sensing technology provided by the invention, the physiological state in the living body is judged by sensing the vibration of the living body, and the technology can be applied to the judgment of the type and the position of the symptom.

Description

Technical Field

[0001] The present disclosure relates to a disease symptom sensing system, device and method, and more particularly to a disease symptom sensing system, device and method for a biological respiratory system. Background Art

[0002] The current clinical diagnosis of a patient's respiratory system is mostly done by doctors using a stethoscope to listen to the patient's lung sounds, and then further determine the type and location of the symptoms based on the lung sounds. However, the range of amplitude and frequency of the sounds or vibrations that the human ear can hear is limited, and sounds that are too high or too low in frequency and subtle sounds cannot be heard. In addition, doctors cannot continuously monitor the patient's lung sounds and cannot monitor the patient's lung symptoms for a long time.

[0003] On the other hand, when conducting animal experiments in current drug experiments, in order to confirm the changes in drug concentration in the blood of animals after taking the medicine, it is necessary to draw blood from the animals at different stages. However, the blood of animals (such as mice) is limited, which limits the number of blood draws. In addition, if it is necessary to determine the changes in symptoms in the animal's body, only slices can be observed, and it is impossible to continuously track the symptoms in a non-invasive way.

[0004] In view of this, how to improve the accuracy of detecting lung sounds and conduct long-term monitoring to diagnose the type, location, severity, etc. of the patient's symptoms is a possible feasible goal. Summary of the invention

[0005] In order to solve the above problems, the present disclosure provides a respiratory system symptom sensing system, comprising a patch and a processor. The patch comprises at least one vibrator and a plurality of receivers, and the patch is used to detect vibrations generated by the respiratory system of the corresponding organism, wherein the at least one vibrator is used to generate reference vibrations. The processor is communicatively connected to the patch. The respiratory system symptom sensing system is used to perform the following operations: each of the plurality of receivers senses the reference vibration corresponding to the at least one vibrator and generates a reference vibration signal respectively; each of the plurality of receivers senses the vibration corresponding to the respiratory system of the organism and generates a vibration signal respectively; the processor determines the relative position of each of the plurality of receivers relative to the respiratory system of the organism according to the plurality of reference vibration signals, the plurality of vibration signals and the plurality of positional relationships of the at least one vibrator corresponding to the plurality of receivers respectively; the processor extracts the corresponding abnormal signal from each of the plurality of vibration signals; and the processor uses a classification model to generate a spatial position in the respiratory system of the organism and a symptom type corresponding to the spatial position according to the plurality of abnormal signals corresponding to the plurality of vibration signals and the plurality of relative positions corresponding to the plurality of receivers.

[0006] The present disclosure also proposes a respiratory system symptom sensing device, comprising a transceiver interface and a processor. The transceiver interface is communicatively connected to a patch, wherein the patch comprises at least one vibrator and a plurality of receivers. The respiratory system symptom sensing device is used to perform the following operations: the transceiver interface receives a plurality of reference vibration signals corresponding to the at least one vibrator and a plurality of vibration signals corresponding to the respiratory system of the organism from the patch; the processor determines the relative position of each of the plurality of receivers relative to the respiratory system of the organism according to the plurality of reference vibration signals, the plurality of vibration signals and the plurality of positional relationships of the at least one vibrator of the patch corresponding to the plurality of receivers of the patch; the processor extracts a corresponding abnormal signal from each of the plurality of vibration signals; and the processor uses a classification model to generate a spatial position in the respiratory system of the organism and a symptom type corresponding to the spatial position according to the plurality of abnormal signals corresponding to the plurality of vibration signals and the plurality of relative positions corresponding to the plurality of receivers.

[0007] In one embodiment of the present disclosure, the operations performed by the processor further include: the processor generates a lung sound signal corresponding to the respiratory system according to each of the multiple vibration signals; and the processor generates the spatial position in the respiratory system corresponding to the organism and the symptom type corresponding to the spatial position according to the multiple lung sound signals and the multiple relative positions corresponding to the multiple receivers.

[0008] In one embodiment of the present disclosure, the operation of the processor generating the lung sound signal further includes the processor filtering out noise and heart sound signals from the plurality of vibration signals.

[0009] In one embodiment of the present disclosure, the operation of the processor capturing the abnormal signal further includes: the processor converting the vibration signal into a spectrum signal based on the frequency domain; and the processor capturing the abnormal signal according to the spectrum signal.

[0010] In one embodiment of the present disclosure, the classification model is used to calculate the disease type corresponding to the vibration signal and the probability that the organism has the corresponding disease type.

[0011] In one embodiment of the present disclosure, the classification model is generated by the following operations: training a machine learning model based on a plurality of historical abnormal signals and the symptom diagnosis results corresponding to each of the plurality of historical abnormal signals; and using the trained machine learning model as the classification model.

[0012] In one embodiment of the present disclosure, the respiratory system symptom sensing system is further configured to perform the following operation: the processor generates a concentration diagnosis result based on the multiple vibration signals using a concentration model, wherein the concentration diagnosis result is used to represent the change in drug concentration in the blood of the organism.

[0013] In one embodiment of the present disclosure, the concentration model is generated by the following operations: training a machine learning model based on a plurality of historical vibration signal groups and historical concentration diagnosis results corresponding to each of the plurality of historical vibration signal groups; and using the trained machine learning model as the concentration model.

[0014] In one embodiment of the present disclosure, the processor calculates the relative position of each of the plurality of receivers relative to the respiratory system of the organism and the corresponding spatial position in the respiratory system of the organism by means of time-of-flight ranging and triangulation positioning.

[0015] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are intended to provide further explanation of the disclosure as claimed. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to make the above and other objects, features, advantages and embodiments of the present disclosure more clearly understood, the accompanying drawings are described as follows:

[0017] Figure 1 is a schematic diagram of a respiratory system disease symptom sensing system in a first embodiment of the present disclosure;

[0018] Figure 2 A flowchart of a respiratory system disease symptom sensing method according to a second embodiment of the present disclosure; and

[0019] Figure 3 FIG. 1 is a schematic diagram of a respiratory system disease symptom sensing device according to a third embodiment of the present disclosure. DETAILED DESCRIPTION

[0020] In order to make the description of the present disclosure more detailed and complete, reference may be made to the attached drawings and various embodiments described below, in which the same numbers in the drawings represent the same or similar elements.

[0021] Please refer to Figure 1 , which is a schematic diagram of a respiratory system symptom sensing system 1 in some embodiments of the present disclosure. The respiratory system symptom sensing system 1 includes a processor 12 and a patch 14, wherein the processor 12 is communicatively connected to the patch 14. The respiratory system symptom sensing system 1 is used to determine the physiological state of the organism according to the vibration or sound of the organism.

[0022] In some embodiments, the processor 12 may include a central processing unit (CPU), a multi-processor, a distributed processing system, an application specific integrated circuit (ASIC), and / or a suitable computing unit.

[0023] The patch 14 includes a vibrator V and receivers R1, R2 and R3. The patch 14 is used to attach to the surface of a living body and detect the vibrations (which can also be understood as the sounds produced by the organs in the respiratory system) generated by the organs in the living body's respiratory system (e.g., bronchi, lungs, trachea) through the receivers R1, R2 and R3. The vibrator V is used to generate reference vibrations, and the receivers R1, R2 and R3 can confirm their relative positions relative to the living body based on the reference vibrations and the vibrations generated by the living body (e.g., heart sounds, lung sounds).

[0024] It should be noted that the patch may include one or more vibrators, and the present invention does not limit the number of vibrators. It should be understood that the determination result can be more accurate based on multiple vibrators. For ease of description, the following embodiments will be described using the vibrator V as an example.

[0025] In some embodiments, the patch 14 is a soft film provided with a vibrator V and receivers R1, R2 and R3. In addition, the vibrator V and receivers R1, R2 and R3 in the patch 14 can exchange control signals and data with the processor 12 in a wired or wireless manner.

[0026] In some embodiments, the vibrator V is an electronic component that can generate vibration or sound, such as a vibration element or a sound unit. The vibrator V is used to provide a reference vibration source, and the reference vibration can be used to correct the signal of the receiver. The reference vibration can also be used to provide the receiver with a basis for calculating the relative position with other vibration sources (for example, the heart, lungs). In addition, the reference vibration can also be used as a basis for adjusting parameters for different organisms. Since the efficiency of vibration transmission in different body tissues is not the same, for example, the transmission efficiency of fat is lower than that of muscle, the body tissue composition of the organism can be inferred through the known frequency and power of the reference vibration.

[0027] In some embodiments, the receivers R1 , R2 , and R3 are electronic components that can sense vibration or sound and convert it into a signal (eg, a digital signal), such as a microphone.

[0028] In some embodiments, the patch 14 can be attached to a position on the surface of a biological body close to the heart and lungs, such as the human chest, so that the receivers R1, R2 and R3 can sense clearer heart and lung sounds.

[0029] It should be noted that, since the organism is a three-dimensional object, the surface of the organism to which the patch 14 is attached may not be a plane but a curved surface or other shapes. Therefore, in order to locate the positions of the receivers R1, R2 and R3 relative to the internal organs of the organism in three-dimensional space, the respiratory system disease symptom sensing system 1 needs to include 3 or more receivers. However, the number of vibrators V and receivers R1, R2 and R3 in this embodiment is only for example. In other embodiments, the patch of the respiratory system disease symptom sensing system may be provided with more receivers and vibrators as reference vibration sources.

[0030] Furthermore, the positional relationship between the vibrator V and the receivers R1, R2 and R3 provided on the patch 14 is known. Figure 1 In the illustrated embodiment, the receiver R1 may be located 15 cm from the upper left corner of the vibrator V, the receiver R2 may be located 15 cm from the upper right corner of the vibrator V, and the receiver R3 may be located 15 cm directly below the vibrator V. However, the above embodiment is only an example, and in fact the size of the patch 14 and the positional relationship between the vibrator V and the receivers R1, R2 and R3 may be adjusted according to the actual application scenario. For example, the size of the patch used to measure the human body may be roughly the area of ​​the human chest cavity, and the vibrators and receivers therein may be evenly distributed in the patch to ensure that the vibrations of various areas of the lungs can be sensed; on the other hand, the size of the patch used to measure mice needs to be adjusted to an area corresponding to the size of the mouse lungs, and the distance and position distribution between the vibrators and receivers therein may also be adjusted accordingly to ensure that the vibrations of various areas of the lungs can be sensed.

[0031] After the patch 14 is attached to the surface of the organism, since the body shape and internal organs of the organism vary, it is necessary to first confirm the relative positions of the receivers R1, R2 and R3 and the heart, lungs and / or other organs of the organism.

[0032] First, each of the receivers R1 , R2 and R3 senses the reference vibration emitted by the corresponding vibrator V and generates a reference vibration signal respectively.

[0033] After the vibrator V emits a reference vibration on the surface of the biological body, the receivers R1, R2 and R3 can sense the vibration of the reference vibration and generate reference vibration signals respectively, wherein the reference vibration signal can include the frequency, amplitude and / or sensing time of the vibration.

[0034] Next, each of the receivers R1 , R2 and R3 senses the vibration of the respiratory system of the corresponding organism and generates a vibration signal respectively.

[0035] The receivers R1 , R2 and R3 sense vibrations emitted by the biological body, such as heart sounds and / or lung sounds, and generate vibration signals respectively, wherein the vibration signals may include the frequency, amplitude and / or sensing time of the vibrations.

[0036] It should be noted that there is no timing restriction on the operation of the receivers R1, R2 and R3 sensing the reference vibration and the vibration. In fact, they can be performed in different orders or simultaneously, and the reference vibration from the vibrator V and the vibration from the biological body can be separated by spectrum conversion and the like.

[0037] Next, the processor 12 determines the relative position of each of the receivers R1, R2 and R3 with respect to the respiratory system of the biological body according to the multiple reference vibration signals, the multiple vibration signals and multiple position relationships of the vibrator V corresponding to the receivers R1, R2 and R3 respectively.

[0038] After the processor 12 receives the corresponding reference vibration signal and vibration signal from the receivers R1, R2 and R3, respectively, since the relative positions and distances of the vibrator V on the patch 14 relative to the receivers R1, R2 and R3 are known, the processor 12 can use Time of Flight (ToF) and triangulation to calculate the relative positions of the receivers R1, R2 and R3 relative to the vibration source of the organism (for example, the heart, lungs), and further confirm the relative positions of the receivers R1, R2 and R3 relative to the respiratory system of the organism.

[0039] In some embodiments, the processor 12 can also establish a three-dimensional model based on the obtained relative position, where the three-dimensional model includes the patch 14 and the heart, lungs and / or various organs in the respiratory system of the organism, and is used to represent the positional relationship between the patch 14 and each organ in the three-dimensional space, as well as the volume size of each organ.

[0040] Specifically, when the heart and respiratory organs of a living organism are functioning, various parts (e.g., heart valves, bronchi, and lungs) will vibrate regularly. Therefore, the positions of the heart and respiratory organs can be calculated through the vibrations generated in different parts, and the volumes and relative positions of the heart and respiratory organs can be further obtained.

[0041] Next, the processor 12 extracts a corresponding abnormal signal from each of the plurality of vibration signals.

[0042] Since the vibrations generated by various parts of the heart and respiratory organs of a biological body in a healthy state are regular, the processor 12 can remove the vibration signals of the heart and respiratory organs in a normal state from the vibration signals to obtain the vibration signals of the heart or respiratory organs with abnormal symptoms (i.e., the abnormal signals).

[0043] In some embodiments, the processor 12 can use spectrum transformation (for example, Fourier transform, wavelet transform, and Hilbert-Huang transform) to convert the vibration signal into a spectrum signal based on the frequency domain, and then, based on the spectrum signal, remove the vibration signals of the heart and respiratory system organs under normal conditions and capture the abnormal signal.

[0044] Finally, the processor 12 utilizes the classification model to generate a spatial position in the respiratory system corresponding to the organism and a symptom type corresponding to the spatial position according to the multiple abnormal signals corresponding to the multiple vibration signals and the multiple relative positions of the corresponding receivers R1, R2 and R3.

[0045] Specifically, after the processor 12 confirms the relative positions of the receivers R1, R2 and R3 relative to the organism, it uses the trained machine learning model (i.e., the classification model) to determine the type of symptom corresponding to the abnormal signal based on the abnormal signal, so as to determine the symptom in the organism. For example, based on the vibration generated by the lungs, it can be determined whether there is phlegm in a specific position of the lungs or that the lungs are healthy (i.e., no symptoms). The processor 12 can locate the position of the corresponding physiological state based on data such as the amplitude, frequency, and reception time in the abnormal signal.

[0046] In some embodiments, the processor 12 may also use the classification model to determine the severity of the disease symptom corresponding to the abnormal signal according to the abnormal signal, for example, to determine the stage of lung cancer of the organism.

[0047] In some embodiments, the processor 12 may utilize time of flight (ToF) and triangulation to calculate the corresponding spatial position according to the abnormal signal, and further confirm the location of the disease symptom.

[0048] It should be noted that the spatial position described in the present disclosure is the spatial position of the abnormal signal in the three-dimensional space formed by the respiratory system of the organism. However, the spatial position is not limited to a point, but may also be a planar area formed by a two-dimensional plane or a spatial area formed by a three-dimensional space. For example, if the respiratory system symptom sensing system 1 determines that an area in the left lung has fibrosis symptoms based on the abnormal signal, the spatial position corresponds to the range of the area.

[0049] In some embodiments, the processor 12 generates a lung sound signal corresponding to the respiratory system of the organism according to each of the multiple vibration signals; and the processor 12 generates the spatial position in the respiratory system corresponding to the organism and the symptom type corresponding to the spatial position according to the multiple lung sound signals and the multiple relative positions corresponding to the multiple receivers.

[0050] Since the vibrations received by the receivers R1, R2 and R3 corresponding to the organism may be the superposition of heart sounds and lung sounds, before determining the physiological state of the organism, the processor 12 may first divide the vibration signals into lung sound signals, heart sound signals and noise, and further generate symptom diagnosis results corresponding to specific locations in the respiratory system of the organism based on the lung sound signals.

[0051] It should be noted that the processor 12 can use a variety of different technical means to filter out noise and heart sound signals from the vibration signal to capture the lung sound signal. For example, the processor 12 can use a trained neural network (e.g., a long short term memory network (LSTM)) to separate the heart sound signal from the lung sound signal; or the processor 12 can also use a high-pass filter to filter out low-frequency noise, and then use a wavelet threshold to separate the heart sound signal from the lung sound signal. The means of separating the heart sound signal and the lung sound signal in the aforementioned embodiment are only used as examples, and the present disclosure is not limited thereto.

[0052] In some embodiments, the classification model is based on a plurality of historical abnormal signals and the disease symptom diagnosis results corresponding to each of the plurality of historical abnormal signals to train a machine learning model; and the trained machine learning model is used as the classification model.

[0053] The classification model can be a machine learning model trained with training data. For example, the respiratory system symptom sensing system 1 is used clinically to obtain the patient's vibration signal, and the corresponding abnormal signal (i.e., historical abnormal signal) is captured by the aforementioned method, and then the doctor confirms the type, location, range and severity of the patient's symptom (i.e., symptom diagnosis result) through X-ray, auscultation, photography or other methods. After obtaining the training data and the corresponding labels, the machine learning model is trained to obtain the classification model, wherein after the vibration signal is input into the classification model, the classification model can calculate the symptom type corresponding to the vibration signal and the probability that the organism has the corresponding symptom type.

[0054] In some embodiments, the processor 12 generates a concentration diagnosis result according to the plurality of vibration signals using a concentration model, wherein the concentration diagnosis result is used to represent the change in drug concentration in the blood of the biological subject.

[0055] In addition to confirming whether there are symptoms in the body, the respiratory system disease symptom sensing system 1 can also be used to confirm the change in blood drug concentration in the body after the organism takes the drug, wherein the concentration model can be constructed by obtaining the corresponding relationship between the blood drug concentration change in the body and the vibration signal after multiple experiments.

[0056] In some embodiments, the concentration model trains a machine learning model based on a plurality of historical vibration signal groups and historical concentration diagnosis results corresponding to each of the plurality of historical vibration signal groups; and the trained machine learning model is used as the concentration model.

[0057] For example, for inhaled lung disease medication, after medication is taken, the respiratory symptom sensing system 1 continuously records the vibration signal of the organism, and at the same time, blood is drawn from the organism at different time points to record the concentration change of the drug in the blood. Furthermore, the drug concentration change (i.e., the historical concentration diagnosis result) and multiple vibration signals corresponding to the same time point and sensed from multiple receivers (i.e., the historical vibration signal group) are used as training data and provided to a machine learning model for training to obtain a trained model (i.e., the concentration model). In this way, a corresponding relationship between the drug concentration change and the vibration signal can be established, and the processor 12 can use the concentration model to generate the corresponding drug concentration change in the blood of the organism after inputting the vibration signal.

[0058] It should be noted that the present disclosure does not limit the model types of the classification model and the concentration model. In practice, a deep learning or neural network model can be selected, or a machine learning algorithm such as a decision tree, K-means clustering, Naive Bayes classifier, eXtreme Gradient Boosting (XGBoost) can be selected. However, there are large differences between individuals of biological organisms, and there are many variables that affect diagnosis. Therefore, in some embodiments, the respiratory symptom sensing system 1 selects a supervised learning or semi-supervised learning machine learning model to form the classification model and the concentration model.

[0059] In some embodiments, the respiratory system disease symptom sensing system 1 may also be trained with a plurality of different models respectively, and a model with a higher judgment accuracy is selected (ie, competing with each other) as the classification model and the concentration model according to the training results.

[0060] Please refer to Figure 2, which is a flow chart of a respiratory system symptom sensing method 2 in the second embodiment of the present disclosure. The respiratory system symptom sensing method 2 includes steps S21, S22, S23 and S24. The respiratory system symptom sensing method 2 is applicable to a respiratory system symptom sensing system (e.g., respiratory system symptom sensing system 1). The respiratory system symptom sensing system includes a processor (e.g., processor 12) and a patch (e.g., patch 14), wherein the processor is communicatively connected to the patch. The patch includes a vibrator (e.g., vibrator V) and a plurality of receivers (e.g., receivers R1, R2 and R3). The patch is used to detect vibrations generated by the respiratory system of the corresponding organism, and the vibrator is used to generate reference vibrations.

[0061] In step S21 , each of the plurality of receivers senses the reference vibration corresponding to the vibrator and generates a reference vibration signal respectively.

[0062] In step S22, each of the plurality of receivers senses the vibration of the respiratory system of the organism and generates a vibration signal respectively.

[0063] In step S23, the processor determines the relative position of each of the plurality of receivers with respect to the respiratory system of the organism according to the plurality of reference vibration signals, the plurality of vibration signals and the plurality of positional relationships of the vibrator corresponding to the plurality of receivers.

[0064] In step S24 , the processor extracts a corresponding abnormal signal from each of the plurality of vibration signals.

[0065] In step S25, the processor uses a classification model to generate a spatial position in the respiratory system corresponding to the organism and a symptom type corresponding to the spatial position based on the multiple abnormal signals corresponding to the multiple vibration signals and the multiple relative positions corresponding to the multiple receivers.

[0066] In some embodiments, the respiratory system symptom sensing method 2 further includes the processor generating a lung sound signal corresponding to the respiratory system according to each of the multiple vibration signals; and the processor generating the spatial position in the respiratory system corresponding to the organism and the symptom type corresponding to the spatial position according to the multiple lung sound signals and the multiple relative positions corresponding to the multiple receivers.

[0067] In some embodiments, the step of the processor generating the lung sound signal further includes the processor filtering out noise and heart sound signals from the plurality of vibration signals.

[0068] In some embodiments, the step of the processor capturing the abnormal signal further comprises the processor converting the vibration signal into a frequency domain-based spectrum signal; and the processor capturing the abnormal signal according to the spectrum signal.

[0069] In some embodiments, the classification model is used to calculate the disease type corresponding to the vibration signal and the probability that the organism has the corresponding disease type.

[0070] In some embodiments, the classification model is generated by the following steps: training a machine learning model based on multiple historical abnormal signals and the disease symptom diagnosis results corresponding to each of the multiple historical abnormal signals; and using the trained machine learning model as the classification model.

[0071] In some embodiments, the respiratory system symptom sensing method 2 further includes the processor generating a concentration diagnosis result according to the plurality of vibration signals using a concentration model, wherein the concentration diagnosis result is used to represent a change in drug concentration in the blood of the biological subject.

[0072] In some embodiments, the concentration model is generated by the following steps: training a machine learning model based on a plurality of historical vibration signal groups and historical concentration diagnosis results corresponding to each of the plurality of historical vibration signal groups; and using the trained machine learning model as the concentration model.

[0073] In some embodiments, the processor calculates the relative position of each of the plurality of receivers relative to the respiratory system of the organism and the corresponding spatial position in the respiratory system of the organism by means of time-of-flight ranging and triangulation positioning.

[0074] Please refer to Figure 3 , is a schematic diagram of a respiratory system symptom sensing device 3 in the third embodiment of the present disclosure. The respiratory system symptom sensing device 3 includes a processor 32 and a transceiver interface 34. The respiratory system symptom sensing device 3 is used to determine the physiological state of the body according to the signal obtained from the patch, wherein the patch can be the patch 14 of the respiratory system symptom sensing system 1 in the first embodiment.

[0075] In some embodiments, the processor 32 may include a central processing unit (CPU), a multi-processor, a distributed processing system, an application specific integrated circuit (ASIC), and / or a suitable computing unit.

[0076] The transceiver interface 34 is used to communicate with the patch and transmit and receive signals to the patch, wherein the patch includes a vibrator and multiple receivers. The transceiver interface 34 may include one or more interfaces of different communication protocols, such as a near-field communication (NFC) interface, a Bluetooth interface, an Ethernet interface, a WiFi network interface, and / or other communication interfaces.

[0077] The processor 32 has the same functions as the processor 12 of the respiratory symptom sensing system 1 in the first embodiment and is used to perform the same operations, so the relevant details are not repeated for the sake of convenience.

[0078] First, the transceiver interface 34 receives a plurality of reference vibration signals corresponding to the vibrator and a plurality of vibration signals corresponding to the respiratory system of the organism from the patch.

[0079] Next, the processor 32 determines the relative position of each of the multiple receivers relative to the respiratory system of the organism based on the multiple reference vibration signals, the multiple vibration signals and the multiple positional relationships of the vibrator of the patch corresponding to the multiple receivers of the patch.

[0080] Next, the processor 32 extracts a corresponding abnormal signal from each of the plurality of vibration signals.

[0081] Finally, the processor 32 utilizes the classification model to generate a spatial position in the respiratory system corresponding to the organism and a disease type corresponding to the spatial position according to the multiple abnormal signals corresponding to the multiple vibration signals and the multiple relative positions corresponding to the multiple receivers.

[0082] In some embodiments, the processor 32 also generates a lung sound signal corresponding to the respiratory system based on each of the multiple vibration signals; and the processor 32 generates the spatial position in the respiratory system corresponding to the organism and the symptom type corresponding to the spatial position based on the multiple lung sound signals and the multiple relative positions corresponding to the multiple receivers.

[0083] In some embodiments, the operation of the processor 32 generating the lung sound signal further includes the processor 32 filtering out noise and heart sound signals from the plurality of vibration signals.

[0084] In some embodiments, the operation of the processor 32 capturing the abnormal signal further includes the processor 32 converting the vibration signal into a spectrum signal based on the frequency domain; and the processor 32 capturing the abnormal signal according to the spectrum signal.

[0085] In some embodiments, the classification model is used to calculate the disease type corresponding to the vibration signal and the probability that the organism has the corresponding disease type.

[0086] In some embodiments, the classification model is generated by the following operations: training a machine learning model based on a plurality of historical abnormal signals and the disease symptom diagnosis results corresponding to each of the plurality of historical abnormal signals; and using the trained machine learning model as the classification model.

[0087] In some embodiments, the processor 32 further generates a concentration diagnosis result according to the plurality of vibration signals using a concentration model, wherein the concentration diagnosis result is used to represent the change in drug concentration in the blood of the biological subject.

[0088] In some embodiments, the concentration model is generated by the following operations: training a machine learning model based on a plurality of historical vibration signal groups and historical concentration diagnosis results corresponding to each of the plurality of historical vibration signal groups; and using the trained machine learning model as the concentration model.

[0089] In some embodiments, the processor 32 calculates the relative position of each of the plurality of receivers relative to the respiratory system of the organism and the corresponding spatial position in the respiratory system of the organism by means of time-of-flight ranging and triangulation positioning.

[0090] According to the above-mentioned embodiment, the respiratory system symptom sensing device 3 can receive a reference vibration signal and a vibration signal from the patch, and generate the aforementioned operation output results such as the spatial position in the respiratory system of the corresponding organism, the type of symptom corresponding to the spatial position, the concentration diagnosis result, etc. according to the reference vibration signal and the vibration signal through the operation as described in the first embodiment.

[0091] In this way, doctors can use the respiratory system disease symptom sensing device 3 to provide remote medical services to patients. Patients can attach the patch to their bodies, and doctors receive signals from the patch through the respiratory system disease symptom sensing device 3 and generate corresponding patient diagnosis information based on the signals, thus breaking the limitations of time and space and realizing the remote medical operation mode.

[0092] In summary, the respiratory system symptom sensing system and method provided by the present disclosure can use multiple receivers to sense the sound or vibration of the organism, confirm the location of the respiratory system organs of the organism, and then judge the physiological state of the organism. The respiratory system symptom sensing system and method can be used to determine the type and location of symptoms, as well as changes in the organism after taking medicine and infer changes in drug concentration in the blood. Since the respiratory system symptom sensing system and method use the technical means of generating reference vibrations by a vibrator for positioning, it is not limited by the body shape of the organism and can be applied to different organisms. In this way, the scope of application is not only for human diagnosis, but also for animal diagnosis or animal experiments. Furthermore, through the remote operation of the patch and the respiratory system symptom sensing device, remote medical services can also be realized.

[0093] Although several embodiments are described above as examples, the respiratory system symptom sensing system, device and method proposed in the present disclosure may also be implemented by other systems, hardware, software, storage media or a combination thereof. Therefore, the protection scope of the present disclosure should not be limited to the specific implementation methods described in the embodiments of the present disclosure, but should be determined by the appended claims.

[0094] It is obvious to those skilled in the art that various modifications and changes can be made to the structure of the present disclosure without departing from the scope or spirit of the present disclosure. In view of the foregoing, the scope of protection of the present disclosure also covers the modifications and changes made within the appended claims.

[0095]

Explanation of symbols

[0096] 1: Respiratory disease symptom sensing system

[0097] 12: Processor

[0098] 14: Patch

[0099] R1~R3: Receiver

[0100] V: Vibrator

[0101] 2: Respiratory system symptom sensing method

[0102] S21~S24: Steps

[0103] 3: Respiratory symptom sensing device

[0104] 32: Processor

[0105] 34: Transceiver interface.

Claims

1. A respiratory system symptom sensing system, characterized in that: Include: The patch is used to detect the vibrations generated by the respiratory system of the corresponding organism, including: at least one vibrator for generating a reference vibration; and Multiple receivers; as well as A processor, communicatively connected to the patch; The respiratory system disease symptom sensing system is used to perform the following operations: Each of the plurality of receivers senses the reference vibration corresponding to the at least one vibrator and generates a reference vibration signal respectively; Each of the plurality of receivers senses the vibration of the respiratory system of the organism and generates a vibration signal respectively; The processor determines the relative position of each of the plurality of receivers relative to the respiratory system of the biological body according to the plurality of reference vibration signals, the plurality of vibration signals and the plurality of positional relationships of the at least one vibrator corresponding to the plurality of receivers respectively; The processor extracts a corresponding abnormal signal from each of the plurality of vibration signals; as well as The processor utilizes a classification model to generate a spatial position in the respiratory system corresponding to the organism and a disease symptom type corresponding to the spatial position according to the multiple abnormal signals corresponding to the multiple vibration signals and the multiple relative positions corresponding to the multiple receivers.

2. The respiratory system symptom sensing system according to claim 1, characterized in that: The operations performed by the processor further include: The processor generates a lung sound signal corresponding to the respiratory system according to each of the plurality of vibration signals; and The processor generates the spatial position in the respiratory system corresponding to the biological body and the disease symptom type corresponding to the spatial position according to the plurality of lung sound signals and the plurality of relative positions corresponding to the plurality of receivers.

3. The respiratory system symptom sensing system according to claim 2, characterized in that: The operation of the processor generating the lung sound signal further includes filtering out noise and heart sound signals from the plurality of vibration signals.

4. The respiratory system symptom sensing system according to claim 1, characterized in that: The operation of the processor capturing the abnormal signal further includes: The processor converts the vibration signal into a frequency spectrum signal based on the frequency domain; and The processor captures the abnormal signal according to the frequency spectrum signal.

5. The respiratory system disease symptom sensing system according to claim 1, characterized in that: The classification model is used to calculate the disease type corresponding to the vibration signal and the probability that the organism has the corresponding disease type.

6. The respiratory system disease symptom sensing system according to claim 1, characterized in that: The classification model is generated by the following operations: Training a machine learning model based on a plurality of historical abnormal signals and a disease symptom diagnosis result corresponding to each of the plurality of historical abnormal signals; and The trained machine learning model is used as the classification model.

7. The respiratory system disease symptom sensing system according to claim 1, characterized in that: The respiratory system disease symptom sensing system is further configured to perform the following operations: The processor generates a concentration diagnosis result according to the plurality of vibration signals by using a concentration model, wherein the concentration diagnosis result is used to represent the change of the drug concentration in the blood of the biological subject.

8. The respiratory system disease symptom sensing system according to claim 7, characterized in that: The concentration model is generated by the following operation: training a machine learning model based on a plurality of historical vibration signal groups and historical concentration diagnosis results corresponding to each of the plurality of historical vibration signal groups; and The trained machine learning model is used as the concentration model.

9. The respiratory system disease symptom sensing system according to claim 1, characterized in that: The processor calculates the relative position of each of the plurality of receivers relative to the respiratory system of the organism and the corresponding spatial position in the respiratory system of the organism by means of time-of-flight ranging and triangulation positioning.

10. A respiratory system disease symptom sensing device, characterized in that: Include: A transceiver interface, communicatively connected to the patch, wherein the patch includes at least one vibrator and a plurality of receivers; and processor; The respiratory system symptom sensing device is used to perform the following operations: The transceiver interface receives a plurality of reference vibration signals corresponding to the at least one vibrator and a plurality of vibration signals corresponding to the respiratory system of the organism from the patch; The processor determines the relative position of each of the plurality of receivers relative to the respiratory system of the organism according to the plurality of reference vibration signals, the plurality of vibration signals and the plurality of positional relationships between the at least one vibrator of the patch and the plurality of receivers of the patch; The processor extracts a corresponding abnormal signal from each of the plurality of vibration signals; as well as The processor utilizes a classification model to generate a spatial position in the respiratory system corresponding to the organism and a disease symptom type corresponding to the spatial position according to the multiple abnormal signals corresponding to the multiple vibration signals and the multiple relative positions corresponding to the multiple receivers.

Images