System and method for health monitoring controlled by speech

Abstract

A system and method for speech-controlled or speech-enabled health monitoring of a subject is disclosed. The device includes a substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic and force signals related to the subject, an acoustic interface configured to communicate with the subject, and a processor connected to the plurality of sensors and the acoustic interface. The processor is configured to determine a biosignal from one or more of the acoustic and force signals and detect the presence of speech in the acoustic signals to monitor the subject's health status. The acoustic interface is configured to interact with the subject or at least one entity associated with the subject based on at least one of an action required due to the subject's health status and a verbal command in the detected speech.

Description

Technical Field

[0001] The present disclosure relates to systems and methods for monitoring the health of a subject.

Background Art

[0002] Speech-enabled technology (technology enabled by speech) has become a standard way of interacting with consumer electronic devices due to its convenience and easy accessibility, enabling more efficient and faster operation. The medical applications of speech technology have been almost limited to care checklists, panic calls, and prescription refills. This is mainly due to the fact that these voice-enabled devices do not have the ability to directly measure and monitor the physiological parameters of the subject. Unlike a permanent state, an episodic state that occurs suddenly or intermittently requires an in-home screening solution that can be used immediately and continuously, and a simple method such as speech to initiate a health check. Also, many people are bedridden or living in poor health. These people are at risk of experiencing sudden health episodes such as falls, apnea, pressure, ulcers, atrial fibrillation, and heart attacks. In the case of living alone, there is no one to notice an early warning, observe the situation, and seek help.

Summary of the Invention

[0003] Disclosed herein is an implementation of a system and method for monitoring the health of a subject that is controlled by speech or enabled by speech.

[0004] In some implementations, the apparatus includes a substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic signals and force signals regarding the subject, an acoustic interface configured to communicate with the subject, and a processor connected to the plurality of sensors and the acoustic interface. The processor is configured to determine a biosignal from one or more of the acoustic signals and the force signals to monitor the health state of the subject, and to detect the presence of speech in the acoustic signals. The acoustic interface is configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and an oral command within the detected speech.

[0005] The present disclosure is best understood from the following detailed description when read in conjunction with the accompanying drawings. It is emphasized that, in accordance with common practice, the various features of the drawings are not to scale. On the contrary, the dimensions of the various features are arbitrarily enlarged or reduced for clarity.

Brief Description of the Drawings

[0006]

Figure 1

[0007]

Figure 2A

Figure 2B

Figure 2C

Figure 2D

Figure 2E

Figure 2F

Figure 2G

Figure 2H

Figure 2I

Figure 2J

[0008]

Figure 3

[0009]

Figure 4

[0010]

Figure 5

[0011]

Figure 6

[0012]

Figure 7

[0013]

Figure 8

Mode for Carrying Out the Invention

[0014] Disclosed herein is an implementation of a system and method for monitoring a subject's health that is controlled or enabled by speech. The system and method can passively and continuously monitor the subject's health and be used to communicate verbally (in words) with the subject, initiate a health check, provide information regarding the subject's health status, record health-related episodes, or perform actions such as calling emergency services. The subject's health and well-being can be monitored using a system that communicates verbally with the subject. Monitoring of sleep, heart, respiration, movement, and sleep apnea are examples. The subject can interact (communicate) with the system using their own speech (voice), request actions to be performed by the system, or obtain information regarding the subject's health status. The system can be used to respond to instructions from the subject's partner if the subject is unable or incapacitated.

[0015] The system and method use one or more non-contact sensors, such as acoustic sensors, accelerometers, pressure sensors, load sensors, weight sensors, force sensors, motion sensors, or vibration sensors, to capture sound (speech and breathing disorders) and mechanical vibrations of the body (movement and physiological movement of the heart and lungs) and convert them into biosignal information used for screening and identifying health and disease states.

[0016] In some implementations, the system includes one or more microphones or acoustic sensors placed near the subject for recording acoustic signals, one or more speakers placed near the subject for playing audio, a physiological measurement system using one or more non-contact sensors for recording mechanical vibrations of the body, such as an accelerometer, a pressure sensor, a load sensor, a weight sensor, a force sensor, a motion sensor, or a vibration sensor, a speech recognition system, a speech synthesis device, and a processor configured to record the subject's acoustic and biometric signals, process them, detect the subject's speech, process the subject's speech, and initiate a response to the subject's speech. In some implementations, the one or more microphones or acoustic sensors, and the one or more non-contact sensors can be placed below the substrate, such as a bed, a sofa, a chair, an examination table, a floor, or can be incorporated into the substrate. For example, the one or more microphones or acoustic sensors, and the one or more non-contact sensors can be placed or positioned inside or below and attached to a control box, legs, a bed frame, a headboard, or a wall. In some implementations, the processor can be present within the device (control box) or within a computing platform (cloud).

[0017] In some implementations, the processor is configured to record mechanical forces and vibrations of the body, including movement and physiological movements of the heart and lungs, using one or more non-contact sensors such as an accelerometer, pressure sensor, load sensor, weight sensor, force sensor, motion sensor, or vibration sensor. The processor is further configured to enhance such data to perform cardiac analysis (including determination of heart rate, heart beat timing, variability, heart beat morphology, corresponding changes from their baselines or ranges), respiratory analysis (including determination of respiratory rate, respiratory phase, depth, timing and variability, respiratory morphology, corresponding changes from their baselines or ranges), and motion analysis (including determination of movement amplitude, time, periodicity, pattern, corresponding changes from their baselines or ranges). The processor is configured to record acoustic information, filter out unwanted interference, and enhance it for analysis and determination.

[0018] For example, the processor may use the enhanced acoustic information to identify sleep apnea. The processor may then determine an appropriate response to the detected sleep apnea, for example, by changing adjustable features of the bed (e.g., firmness) or the bedroom (e.g., lighting), or may sound a noise to cause the sleeper to change position or transition to a lighter sleep state to assist in stopping, reducing, or altering the sleep apnea. For example, the processor may correlate irregular lung or body movements with lung or body sounds using the enhanced acoustic information. Wheezing and other abnormal sounds are examples. For example, the processor may use the enhanced acoustic information to detect whether speech has started. The processor compares the acoustic stream against a dictionary of electronic commands to discard irrelevant conversations and determine whether an oral command to interact with the system has been initiated.

[0019] In some implementations, the processor is configured to process speech recognition. For example, the processor may perform speech recognition. This may include detecting a trigger (e.g., a pre-set keyword or phrase) and discerning the context (content). The keyword may be, for example, "Afib (atrial cells)" which triggers annotation (marking) of a heart record or generation of an alert. For example, the processor may communicate via an API with other speech function devices (e.g., Alexa (registered trademark), Siri (registered trademark), Google (registered trademark), etc.) responsible for speech recognition and synthesis.

[0020] In some implementations, the processor is configured to classify and initiate a response to the recognized speech. The response may be the initiation of an interactive session with the subject (e.g., playing a tone (audible sound) or a synthesized speech (voice)), or the execution of a response action (e.g., turning on / off a home automation function, labeling data at a health marker for future access by the subject or the subject's physician, calling an emergency service). The response may also include communication with other speech function devices connected to a home automation system or a notification system. The system may also be used to create events based on the analysis, and the event may be an audible tone or message sent to the cloud for a critical situation.

[0021] The sensor is connected to the processor, either wired, wirelessly, or optically. The processor may exist on the Internet and may execute artificial intelligence software. Signals from the sensor can be analyzed locally using a locally existing processor, or alternatively, the data can be networked with other computers and remote storage that can process and analyze real-time data and / or historical data, either wired or by other means. The processor can be a single processor that corresponds to both a mechanical force sensor and an acoustic sensor, or alternatively, a series of processors that process mechanical forces and interact (communicate) with other speech function devices. Other sensors such as a blood pressure sensor, a body temperature sensor, a blood oxygen sensor, a pulse oximetry sensor, etc. can be added to enhance monitoring and health status assessment. The system can use artificial intelligence and / or machine learning to train a classifier used to process force sensor signals, acoustic sensor signals, and other sensor signals.

[0022] In some implementations, the speech function device can function as a speech recognition device or a speech synthesis device and can assist in one-way and two-way communication with the subject. The speech recognition device converts speech (voice) into text, and the speech synthesis device converts text into speech (voice). Both are based on a dictionary of predefined keywords or phrases. The system includes two-way audio (microphone and speaker) that enables two-way communication with the patient (the subject's speech functions as a command, and the device responds when the command is received). The system can additionally include an interface to other voice assistant devices (e.g., Alexa (registered trademark), Siri (registered trademark), Google (registered trademark), etc.) and can process the subject's speech, play the synthesized response, or do both.

[0023] The systems and methods described herein can be used by a subject when experiencing a complication or exhibiting early warning signs of a health - related condition, or can be used when directed by a physician in a telemedicine application. For example, the system can be utilized for a stress test at home, in which case sensor data can be used to monitor indicators of heart rate variability to quantify dynamic autonomic regulation or heart rate recovery.

[0024] The system can be programmed to limit the number of individuals with whom it can verbally interact. For example, the system can accept and respond only to verbal commands from one person (the subject) or the subject's partner. In such a case, speech recognition will include voice recognition that responds only to a specific individual. Verbal commands can include requests to have the subject perform a specific health assessment (e.g., a heart examination or stress test), provide an update on the subject's health status, mark data when the subject is experiencing a health episode or condition, send a health report to the subject's physician, call (phone) emergency services, order a product via API integration with a third party (e.g., purchase something from an internet vendor), and / or interact with adjustable features of home automation, but are not limited to these. The system can be integrated with other communication means such as a tablet or smartphone and can provide video communication.

[0025] FIG. 1 is a system architecture of a speech-controlled or speech-enabled health monitoring system (SHMS) 100. SHMS 100 includes one or more devices 110 that are connected to or communicating with (collectively "connected to") a computing platform 120. In some implementations, a machine learning training platform 130 may be connected to the computing platform 120. In some implementations, a speech function device 150 may be connected to the computing platform 120 and the one or more devices 110. In some implementations, a user may access data via a connection device 140, and the connection device 140 may receive data from the computing platform 120, the device 110, the speech function device 150, or a combination thereof. The connections between the one or more devices 110, the computing platform 120, the machine learning training platform 130, the speech function device 150, and the connection device 140 may be wired, wireless, optical, a combination thereof, etc. The system architecture of SHMS 100 is exemplary and may include more (additional), fewer, or different devices, entities, etc., designed the same or differently without departing from the scope of this specification and the claims. Further, the illustrated devices may perform other functions without departing from the scope of this specification and the claims.

[0026] In one implementation, device 110 may include an acoustic interface 111, one or more sensors 112, a controller 114, a database 116, and a communication interface 118. In one implementation, device 110 may include a classifier 119 for applicable and appropriate machine learning techniques described herein. One or more sensors 112 may detect sound, waveform patterns, and / or a combination of sound and waveform patterns. A waveform pattern is a waveform pattern of vibration, pressure, force, weight, presence, and movement due to the activities and / or forms (postures) of a subject with respect to the one or more sensors 112. In some implementations, one or more sensors 112 may generate two or more data streams. In some implementations, one or more sensors 112 may be of the same type. In some implementations, one or more sensors 112 may be time-synchronized. In some implementations, one or more sensors 112 may measure partial forces of gravity in a substrate, furniture, or other object. In some implementations, one or more sensors 112 may independently capture multiple external sources of data, such as body weight, heart rate, respiratory rate, vibration, and movement, from one or more subjects or objects, within one stream (i.e., a multivariate signal). In one implementation, the data captured by each sensor 112 is correlated with the data captured by at least one, some, all, or a certain combination of the other sensors 112. In some implementations, changes in amplitude are correlated. In some implementations, the rate (frequency) and magnitude of the changes are correlated. In some implementations, the phase and direction of the changes are correlated. In some implementations, the arrangement of one or more sensors 112 triangulates the position of the center of gravity. In some implementations, one or more sensors 112 may be placed below the legs of a bed, chair, sofa, etc., or embedded within the legs. In some implementations, one or more sensors 112 may be placed below the edge of an infant bed, or embedded within the edge. In some implementations, one or more sensors 112 may be placed below the floor, or embedded within the floor.In some implementations, one or more sensors 112 may be disposed below the surface area or may be embedded within the surface area. In some implementations, the position of one or more sensors 112 is used to generate a surface map that covers the entire area surrounded by the sensors. In some implementations, one or more sensors 112 may measure data from a source that can be anywhere within the area surrounded by the one or more sensors 112. The source may be present directly on top of the one or more sensors 112, in the vicinity of the one or more sensors 112, or at a distance from the one or more sensors 112. The one or more sensors 112 do not interfere with the subject.

[0027] The one or more sensors 112 may include one or more non-contact sensors such as acoustic sensors, microphones, or acoustic sensors for capturing sound (speech and sleep apnea), and sensors including accelerometers, pressure sensors, load sensors, weight sensors, force sensors, motion sensors, or vibration sensors for measuring partial forces of gravity in a substrate, furniture, or other object, and sensors for mechanical vibrations of the body (movement and physiological movements of the heart and lungs).

[0028] The acoustic interface 111 provides a two-way acoustic interface (microphone and speaker) to enable two-way communication with the patient (the subject's speech functions as a command and the device responds when it receives the command).

[0029] Controller 114 may apply the processes and algorithms described herein in connection with FIGS. 3 through 8 to sensor data to determine biometric parameters and other personally identifiable information of one or more subjects at rest and during movement. Classifier 119 may apply the processes and algorithms described herein in connection with FIGS. 3 through 8 to sensor data to determine biometric parameters and other personally identifiable information of one or more subjects at rest and during movement. Classifier 119 may apply the classifier to sensor data to determine biometric parameters and other personally identifiable information via machine learning. In some implementations, classifier 119 may be implemented by controller 114. In some implementations, the sensor data and the biometric parameters and other personally identifiable information may be stored in database 116. In some implementations, the sensor data, the biometric parameters and other personally identifiable information, and / or combinations thereof may be transferred or transmitted to computing platform 120 via communication interface 118 for processing, storage, and / or combinations thereof. Communication interface 118 may be any interface and may use any communication protocol to communicate or transfer data between a source endpoint and a destination endpoint. In one implementation, device 110 may be any platform or structure that uses one or more sensors 112 to collect data from a subject for use by controller 114 and / or computing platform 120 as described herein. For example, device 110 may be a combination of a substrate, a frame, legs, and a plurality of load sensors or other sensors 112, as shown in FIG. 2. Device 110 and its internal elements may include other elements that may be desirable or necessary to implement the devices, systems, and methods described herein. However, such elements and processes are well known in the art and do not promote a better understanding of the disclosed embodiments, so descriptions of such elements and processes may not be provided herein.

[0030] In some implementations, computing platform 120 may include a processor 122, a database 124, and a communication interface 126. In some implementations, computing platform 120 may include a classifier 129 for applicable and appropriate machine learning techniques as described herein. Processor 122 may obtain sensor data from sensor 112 or controller 114 and may apply the processes and algorithms described herein with reference to FIGS. 3-8 to the sensor data to determine biometric parameters and other personally identifiable information of one or more subjects at rest and during movement. In some implementations, processor 122 may obtain biometric parameters and other personally identifiable information from controller 114 and store them in database 124 for temporary and other types of analysis. In some implementations, classifier 129 may apply the processes and algorithms described herein with reference to FIGS. 3-8 to the sensor data to determine biometric parameters and other personally identifiable information of one or more subjects at rest and during movement. Classifier 129 may apply the classifier to the sensor data and may determine biometric parameters and other personally identifiable information via machine learning. In some implementations, classifier 129 may be implemented by processor 122. In some implementations, the sensor data, and the biometric parameters and other personally identifiable information, may be stored in database 124. Communication interface 126 may be any interface and may use any communication protocol and may communicate or transfer data between a source endpoint and a destination endpoint. In some implementations, computing platform 120 may be a cloud-based platform. In some implementations, processor 122 may be a cloud-based computer or an off-site controller. In some implementations, processor 122 may be a single processor corresponding to both a mechanical force sensor and an acoustic sensor, or may be a series of processors that process mechanical forces and interact (communicate) with speech function device 150.Computing platform 120 and the elements within it may include other elements that may be desirable or necessary to implement the devices, systems, and methods described herein. However, since such elements and processes are well known in the art and do not contribute to a better understanding of the disclosed embodiments, descriptions of such elements and processes may not be provided herein.

[0031] In some implementations, the machine learning training platform 430 may access and process sensor data and train and generate multiple classifiers. The multiple classifiers may be transferred or sent to classifier 129 or classifier 119.

[0032] In some implementations, SHMS 100 may optionally include a speech function device 150 as a two-way speech interface. In some implementations, the speech function device 150 may replace the acoustic interface 111 or cooperate with the acoustic interface 111. The speech function device 150 may communicate with the device 110 and / or the computing platform 120. In one implementation, the speech function device 150 may be a voice assistant device (e.g., Alexa®, Siri®, Google®) that communicates with the device 110 or the computing platform 120 via an API. The speech function device 150 may function as a speech recognition device or a speech synthesis device that supports one-way and two-way communication with the subject.

[0033] Figures 2A through 2J are diagrams of the sensor arrangements and configurations. As described herein, the SHMS100 may include one or more acoustic input sensors 200 such as microphones or acoustic sensors. The sensor arrangements and configurations shown in Figures 2A through 2J relate to the bed 230 and the surrounding environment. For example, U.S. Patent Application No. 16 / 595,848, filed on October 8, 2019, which is hereby incorporated by reference in its entirety, describes exemplary beds and environments applicable to the sensor arrangements and configurations described herein.

[0034] Figure 2A shows an example of one or more acoustic input sensors 200 within the control box (controller) 240. Figure 2B shows an example of one or more acoustic input sensors 200 attached to the headboard 250 proximate to the bed 230. Figure 2C shows an example of one or more acoustic input sensors 200 attached to the wall 260 proximate to the bed 230. Figure 2D shows an example of one or more acoustic input sensors 200 inside or attached to the legs 270 of the bed 230. Figure 2E shows an example of one or more acoustic input sensors 200 integrated within the force sensor box 280 below the legs 270 of the bed 230. Figure 2F shows an example of one or more acoustic input sensors 200 disposed within or attached to the bed frame 290 of the bed 230.

[0035] In some implementations, the SHMS 100 may include one or more speakers 210. FIG. 2G shows an example of one or more speakers 210 within the control box (controller) 240. FIG. 2F shows an example of one or more speakers 210 disposed within or attached to the bed frame 290 of the bed 230. FIG. 2H shows an example of one or more speakers 210 integrated within the force sensor box 280 below the legs 270 of the bed 230. FIG. 2I shows an example of one or more speakers 210 attached to the wall 260 adjacent to the bed 230. FIG. 2J shows an example of one or more speakers 210 attached to the headboard 250 adjacent to the bed 230.

[0036] FIGS. 2A through 2E and 2G are examples of systems with one-way audio communication, and FIGS. 2F and 2H through 2J are examples of systems with two-way audio communication.

[0037] FIG. 3 is a processing pipeline 300 for acquiring sensor data such as, but not limited to, force sensor data, acoustic sensor data, and other sensor data, and for processing the force sensor data, the acoustic sensor data, and the other sensor data.

[0038] An analog sensor data stream 320 is received from the sensor 310. The sensor 310 may record mechanical forces and vibrations of the body (including movement and physiological movements of the heart and lungs) using one or more non-contact sensors such as an accelerometer, a pressure sensor, a load sensor, a weight sensor, a force sensor, a motion sensor, or a vibration sensor. A digitizer 330 digitizes the analog sensor data stream into a digital sensor data stream 340. A framer 350 generates a digital sensor data frame 360 from the digital sensor data stream 340. It includes all digital sensor data stream values within a fixed or adaptive time window. An encryption engine 370 encrypts the digital sensor data frame 360 so that the data is protected from unauthorized access. A compression engine 380 compresses the encrypted data to reduce the data size so that it can be stored in a database 390. This reduces costs and provides faster access during read time. The database 390 can be local storage, off-site storage, cloud-based storage, or a combination thereof.

[0039] The analog sensor data stream 321 is received from the sensor 311. The sensor 311 may record acoustic information including the subject's breathing and speech. The digitizer 331 digitizes the analog sensor data stream into a digital sensor data stream 341. The framer 351 generates a digital sensor data frame 361 from the digital sensor data stream 341. It includes all digital sensor data stream values within a fixed or adaptive time window. The encryption engine 371 encrypts the digital sensor data frame 361 so that the data is protected from unauthorized access. In some implementations, the encryption engine 371 may filter the digital acoustic sensor data frame 361 into a lower and narrower frequency band. In some implementations, the encryption engine 371 may mask the digital acoustic sensor data frame 361 using a mask template. In some implementations, the encryption engine 371 may transform the digital acoustic sensor data frame 361 using a mathematical formula. The compression engine 381 compresses the encrypted data to reduce the data size so that it can be stored in the database 390. This reduces costs and provides faster access during read times. The database 390 may be local storage, off-site storage, cloud-based storage, or a combination thereof.

[0040] The processing pipeline 300 shown in FIG. 3 is exemplary and may include any, all, or a combination of the blocks or modules shown in FIG. 3, may include none of them, or may include combinations thereof. The order of processing shown in FIG. 3 is exemplary and the order of processing may be changed without departing from the scope of this specification or the claims.

[0041] FIG. 4 is a preprocessing pipeline 400 for processing force sensor data. The preprocessing pipeline 400 processes digital force sensor data frames 410. A noise reduction unit 420 removes or attenuates noise sources that can affect each sensor at the same level or different levels. The noise reduction unit 420 can utilize various techniques including, but not limited to, subtraction, combination of input data frames, adaptive filtering, wavelet transform, independent component analysis, principal component analysis, and / or other linear or non-linear transforms. A signal enhancement unit 430 can improve the signal-to-noise ratio of the input data. The signal enhancement unit 430 can be implemented as a linear or non-linear combination of the input data frames. For example, the signal enhancement unit 430 can combine signal deltas and increase the signal strength for higher resolution algorithm analysis. Subsampling units 440, 441, and 442 sample the digital enhanced sensor data and can include downsampling, upsampling, or resampling. The subsampling can be implemented as multistage sampling or polyphase sampling and can use the same or different sampling rates for heart analysis, respiratory analysis, and cough analysis.

[0042] The heart analysis 450 determines (judges) the heart rate, heart beat timing, variability, and heart beat morphology, as well as the corresponding changes from their baselines or predetermined ranges. An exemplary process of heart analysis is shown in U.S. Provisional Patent Application No. 63 / 003,551 filed on April 1, 2020. The entire disclosure of the provisional application is incorporated herein by reference. The respiration analysis 460 determines (judges) the respiration rate, respiration phase, depth, timing, variability, and respiration morphology, as well as the corresponding changes from their baselines or desired ranges. An exemplary process of respiration analysis is shown in U.S. Provisional Patent Application No. 63 / 003,551 filed on April 1, 2020. The entire disclosure of the provisional application is incorporated herein by reference. The movement analysis 470 determines (judges) the amplitude, time, periodicity, and pattern of movement, as well as the corresponding changes from their baselines or desired ranges. The health and sleep state analysis 480 combines the data from the heart analysis 450, respiration analysis 460, and movement analysis 470 to determine (judge) the health state, sleep quality, abnormal events, diseases, and conditions of the subject.

[0043] The processing pipeline 400 shown in FIG. 4 is exemplary and may include any one or all of the blocks or modules shown in FIG. 4, may include none of them, or may include combinations thereof. The order of processing shown in FIG. 4 is exemplary and the order of processing may be changed without departing from the scope of this specification or the claims.

[0044] FIG. 5 is an exemplary process 500 for analyzing acoustic sensor data. Pipeline 500 processes digital acoustic sensor data frame 510. Noise reduction unit 520 removes or attenuates environmental or other noise sources that may affect each sensor at the same level or different levels. Noise reduction unit 520 may utilize various techniques including, but not limited to, subtraction, combination of input data frames, adaptive filtering, wavelet transform, independent component analysis, principal component analysis, and / or other linear or non-linear transforms. Signal enhancement unit 530 may improve the signal-to-noise ratio of the input data. Speech start detector 540 determines whether the subject is communicating verbally with the system. The detector 540 compares the acoustic stream to a dictionary of electronic commands to discard irrelevant conversations and determines whether a verbal command for interaction has been initiated (545).

[0045] If no verbal command has been initiated, the enhanced digital acoustic sensor data frame will be analyzed using sleep apnea analyzer 550 to detect a breathing disorder. Sleep apnea analyzer 550 uses digital acoustic sensor data frame 510, digital force sensor data frame 410, or both, to determine a breathing disorder. Sleep apnea analyzer 550 uses an envelope detection algorithm, time domain analysis, spectral domain analysis, or time-frequency domain analysis, to identify the presence, intensity, magnitude, duration, and type of sleep apnea.

[0046] If it is determined that a verbal command has been initiated, the speech recognition device 560 processes the enhanced digital acoustic sensor data frame to identify the content of the speech. In some implementations, the speech recognition device 560 includes an electronic command recognition device that compares the subject's speech to a dictionary of electronic commands. In some implementations, the speech recognition device 560 uses an artificial intelligence algorithm to identify the speech. In some implementations, the speech recognition device 560 uses a speech-to-text engine to convert (translate) the subject's verbal command into a text string. The response categorizer 570 processes the output from the speech recognition device 560 to determine whether a dialog session should be initiated (580) or a response action should be performed (590). Examples of dialog sessions include the playback of tones or the playback of synthetic speech. Examples of response actions include turning on / off a home automation function, labeling data with health status markers for future access by the subject or the subject's physician, calling an emergency service, and interacting with another speech-enabled device.

[0047] The processing pipeline 500 shown in FIG. 5 is exemplary and may include any, all, or combinations of the components, blocks, or modules shown in FIG. 5, may include none of them, or may include combinations thereof. The order of processing shown in FIG. 5 is exemplary, and the order of processing may be changed without departing from the scope of this specification or the claims.

[0048] FIG. 6 is an exemplary process 600 for analyzing acoustic sensor data by interacting with a speech-enabled device. In some implementations, the speech-enabled device can be a voice assistant device (e.g., Alexa®, Siri®, Google®) that acts as a speech recognition device communicating via an API.

[0049] Pipeline 500 receives speech data from the speech function device (610). A noise reduction unit 620 removes or attenuates an environment or other noise source that can affect the speech data at the same level or at a different level. The noise reduction unit 620 can utilize various techniques including, but not limited to, subtraction, combination (combination) of input data frames, adaptive filtering, wavelet transform, independent component analysis, principal component analysis, and / or other linear or non-linear transforms. A signal enhancement unit 630 can improve the signal-to-noise ratio of the speech data. A speech start detector 640 determines whether the subject is communicating verbally with the system. The detector 640 compares the speech data with a dictionary of electronic commands to discard (discard) irrelevant conversations and determines whether a verbal command for dialogue has been started (645).

[0050] If no verbal command has been started, the enhanced digital speech data frame will be analyzed using the sleep apnea analyzer 650 to detect apnea. The sleep apnea analyzer 650 uses the speech data 610, the digital force sensor data frame 410, or both, to determine apnea. The sleep apnea analyzer 650 uses an envelope detection algorithm, time domain analysis, spectral domain analysis, or time-frequency domain analysis, to identify the presence, intensity, magnitude, duration, and type of sleep apnea.

[0051] When it is determined that the verbal command has started, the speech recognition device 660 processes the speech data frame to identify the content of the speech. In some implementations, the speech recognition device 660 includes an electronic command recognition device that compares the subject's speech with a dictionary of electronic commands. In some implementations, the speech recognition device 660 uses an artificial intelligence algorithm to identify the speech. In some implementations, the speech recognition device 660 uses a speech-to-text engine to convert (translate) the subject's verbal command into a text string. The response categorizer 670 processes the output from the speech recognition device 660 to determine whether the interactive session 680 should be started or whether the response operation 690 should be executed. The command corresponding to the categorized response is sent to the speech function device via the API (675). In some implementations, the speech function device may function as a speech synthesis device and may start an interactive session (680). In some implementations, the speech function device may be connected to a home automation system or a notification system and may execute a response operation (690). An example of an interactive session is the playback of a tone or the playback of synthesized speech (voice). Examples of response operations are turning on / off a home automation function, labeling data at a health status marker for future access by the subject or the subject's physician, calling an emergency service, and interacting with another speech function device.

[0052] The processing pipeline 600 shown in FIG. 6 is illustrative and may include any, all, or a combination of the components, blocks, or modules shown in FIG. 6, may include none of them, or may include a combination thereof. The order of the processing shown in FIG. 6 is illustrative, and the order of the processing may be changed without departing from the scope of this specification or the claims.

[0053] FIG. 7 is an exemplary process 700 for recognizing speech (voice) by a speech recognition device. After it is determined that speech has started as described in FIG. 5, the speech recognition device receives an enhanced acoustic data stream (710). The speech recognition device detects a portion of an electronic command that matches a specific request by speech processing, i.e., detects a trigger (720). The speech recognition device converts the speech (voice) into text (730). The speech recognition device collates the text string against a dictionary of electronic commands (750) (740). The speech recognition device determines the content (context) of the speech (voice) (760). The content is a general category of the subject's oral request. Examples thereof are performing a health check, labeling or annotating data for a health-related episode, communicating with the subject's doctor, communicating with an emergency service, placing an order for a product, interacting with home automation, etc. The speech recognition device encrypts the content (770) and prepares it for the response categorizer 570.

[0054] The processing pipeline 700 shown in FIG. 7 is exemplary and may include any, all, or a combination of the components, blocks, or modules shown in FIG. 7, may include none of them, or may include combinations thereof. The order of the processing shown in FIG. 7 is exemplary and the order of the processing may be changed without departing from the scope of this specification or the claims.

[0055] FIG. 8 is an exemplary process 800 for the detection and response of sleep disordered breathing (SDB). A digital force sensor frame 810 is received as processed in FIGS. 3 and 4. A respiration analysis 830 is performed on the digital force sensor frame 810. The respiration analysis 830 may include filtering, combining, envelope detection, and other algorithms. A spectrum or time-frequency spectrum is calculated for the output of the respiration analysis 830 (850). A digital acoustic sensor frame 820 is received as processed in FIGS. 3 and 5. An envelope detection 840 is performed on the digital acoustic sensor frame 820. A spectrum or time-frequency spectrum is calculated for the output of the envelope detection 840 (860). For the digital force sensor frame 810 and the digital acoustic sensor frame 820, a fusion sensor processing 870 such as a normalized amplitude or frequency parameter, cross-correlation, or a metric (measurement criterion) of coherence or similar similarity is performed to create a combined (assembled) set of signals or features.

[0056] Sleep disordered breathing (SDB) is determined (880) using envelopes, time domain, frequency domain, time-frequency, and parameters from the fusion of a force sensor and an acoustic sensor. Some implementations include the use of threshold-based techniques, template matching methods, or classifiers to detect sleep disordered breathing. When sleep disordered breathing is detected, process 880 determines the intensity (e.g., mild, moderately mild, moderate, severe), magnitude, duration, and type of the sleep disordered breathing. If sleep disordered breathing is detected (885), an appropriate response 890 is determined for the detected SDB, for example, by changing adjustable functions of the bed (e.g., firmness) or the bedroom (e.g., lighting), or by making a sound to cause the sleeper to change position (posture) or transition to a lighter sleep state to assist in stopping, reducing, or changing the sleep disordered breathing.

[0057] The processing pipeline 800 shown in FIG. 8 is exemplary and may include any, all, or a combination of the components, blocks, or modules shown in FIG. 8, or may include none of them. The order of processing shown in FIG. 8 is exemplary, and the order of processing may be changed without departing from the present specification or the scope of the claims.

[0058] FIG. 7 is a flowchart of a method 700 for determining (judging) weight from MSMDA data. The method 700 includes a step 710 of acquiring MSMDA data, a step 720 of calibrating the MSMDA data, a step 730 of performing an overlay analysis on the calibrated MSMDA data, a step 740 of converting the MSMDA data into weight, a step 750 of finalizing (determining) the weight, and a step 760 of outputting the weight.

[0059] The method 700 includes a step 710 of acquiring MSMDA data. The MSMDA data is generated from the preprocessing pipeline 600 as described.

[0060] The method 700 includes a step 720 of calibrating the MSMDA data. The calibration process compares a plurality of sensor readings with an expected value or an expected range. If the values are different, the MSMDA data is adjusted to be calibrated to the expected value or the expected range. Calibration is implemented by turning off (i.e., setting them to zero) all other sources in order to determine (judge) the weight of a new object. For example, the weights of the bed, bedding, and pillow are determined before the new object. For example, a baseline is established for the device before use. In one implementation, when a subject or object (collectively "item") is present on the device, an item baseline is determined and saved. This is done so that data from a device having multiple items can be correctly processed using the method described herein.

[0061] Method 700 includes a step 730 of performing an overlay analysis on the calibrated MSMDA data. The overlay analysis provides the sum of the readings that occur when each independent sensor operates alone. The overlay analysis can be implemented as an algebraic sum, weighted sum, or non-linear sum of the responses from all sensors.

[0062] Method 700 includes a step 740 of converting the MSMDA data to weight. Various known or to-be-known techniques can be used to convert the sensor data, i.e., the MSMDA data, to weight.

[0063] Method 700 includes a step 750 of finalizing the weight. In one implementation, the finalization of the weight can include smoothing, checking against a range, checking against a dictionary, or past values. In one implementation, the finalization of the weight can include adjustments due to other factors such as the type of the bed, the size of the bed, the location of the sleeper, the position (posture) of the sleeper, the orientation of the sleeper, etc.

[0064] Method 700 includes a step 760 of outputting the weight. The weight is saved for use in the methods described herein.

[0065] The implementation of controller 200, controller 214, processor 422, and / or controller 414 (and algorithms, methods, instructions, etc. stored in and / or executed by them) can be realized in hardware, software, or any combination thereof. The hardware can include, for example, a computer, an IP core, an application-specific integrated circuit (ASIC), a programmable logic array, an optical processor, a programmable logic controller, microcode, a microcontroller, a server, a microprocessor, a digital signal processor, or any other suitable circuit. In the claims, the term "controller" should be understood to encompass any of the aforementioned hardware, either alone or in combination.

[0066] Further, in one aspect, for example, controller 200, controller 214, processor 422, and / or controller 414 may be implemented using a general-purpose computer or a general-purpose processor with a computer program that, when executed, performs any of the respective methods, algorithms, and / or instructions described herein. Additionally or alternatively, a dedicated computer / processor may be utilized that, for example, may include other hardware for performing any of the methods, algorithms, or instructions described herein.

[0067] Controller 200, controller 214, processor 422, and / or controller 414 may be one or more of dedicated processors, digital signal processors, microprocessors, controllers, microcontrollers, application processors, central processing units (CPUs), graphics processing units (GPUs), digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays, any other type or combination of integrated circuits, state machines, or any combination thereof in distributed, centralized, cloud-based architectures and / or combinations thereof.

[0068] Generally, the apparatus includes a substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic and force signals regarding the subject, an acoustic interface configured to communicate with the subject, and a processor connected to the plurality of sensors and the acoustic interface. The processor is configured to determine a biosignal from one or more of the acoustic and force signals for monitoring the health state of the subject and to detect the presence of speech in the acoustic signal. The acoustic interface is configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and a verbal command within the detected speech.

[0069] In some embodiments, the processor is further configured to encrypt the digitized acoustic signal by at least one of filtering the digitized acoustic signal to a lower and narrower frequency, masking the digitized acoustic signal using a mask template or a cryptographic key, and converting the digitized acoustic signal using a mathematical formula. In some embodiments, the processor is further configured to perform at least one of comparing the acoustic signal against a dictionary of electronic commands to discard irrelevant conversations, determining the presence of the verbal command, identifying the content of the utterance upon determination of the verbal command, initiating an interactive session with at least one of the subject or another entity based on the verbal command and the content of the utterance via the acoustic interface, and determining a response action based on the verbal command and the content of the utterance. In some embodiments, the acoustic interface is further configured to recognize and respond to verbal commands from a designated individual. In some embodiments, the processor is further configured to compare the acoustic signal against a dictionary of electronic commands to discard irrelevant conversations, determine the presence of the verbal command, analyze the acoustic signal to detect a breathing disorder upon failure to detect the verbal command, and determine a response action for the detection of sleep disordered breathing (SDB). In some embodiments, the plurality of non-contact sensors are configured to capture a force signal from the subject's movement relative to the substrate, and the processor is further configured to perform at least one of cardiac analysis, respiratory analysis, and movement analysis based on the force signal to determine the health state of the subject. In some implementations, when performing a respiratory disorder analysis to determine the health state of the subject, the processor is further configured to fuse the force signal and the acoustic signal based on one or more similarity metrics to generate a fused signal, and use the fused signal, the force signal, and the acoustic signal to detect a sleep disordered breathing (SDB), and determine a response action for the detection of the SDB. In some implementations, the response action is one or more of an audible tone, an audible message, a trigger of a home automation device, a trigger of a voice assistant device, a call to an entity or an emergency service, marking of data for future access, an entry in a database, and a health diagnosis. In some implementations, the processor is further configured to determine the intensity, magnitude, duration, and type of the SDB.

[0070] Generally, the system includes a speech function device configured to communicate with at least one of a subject or an entity associated with the subject, and a device that communicates with the speech function device. The device includes a substrate configured to support the subject, a plurality of non-contact sensors configured to capture an acoustic signal regarding the subject and a force signal from the subject's movement on the substrate, and a processor connected to the plurality of sensors and the acoustic interface. The processor is configured to monitor the health state of the subject based on the force signal and the acoustic signal, and detect a verbal command in the acoustic signal. The speech function device is configured to communicate with at least the subject or the entity based on at least one of a response action required due to the health state of the subject and the detection of the verbal command.

[0071] In some implementations, the processor is further configured to encrypt the digitized acoustic signal by at least one of filtering the digitized acoustic signal to lower and narrower frequencies, masking the digitized acoustic signal using a mask template or a cryptographic key, and transforming the digitized acoustic signal using a mathematical formula. In some implementations, the processor is further configured to perform at least one of comparing the acoustic signal against a dictionary of electronic commands to discard irrelevant conversations, determining the presence of the verbal command, identifying the content of the utterance upon determination of the verbal command, initiating an interactive session with at least one of the subject or the entity based on the verbal command and the content of the utterance via the speech function device, and determining the response action based on the verbal command and the content of the utterance. In some implementations, the speech function device is further configured to recognize and respond to verbal commands from a designated individual. In some implementations, the processor is further configured to perform at least respiratory analysis based on the force signal, compare the acoustic signal against a dictionary of electronic commands to discard irrelevant conversations, determine the presence of the verbal command, and fuse the force signal and the acoustic signal based on one or more similarity metrics to generate a fused signal upon failure to detect the verbal command. The processor is configured to detect sleep disordered breathing (SDB) using the fused signal, the force signal, and the acoustic signal, and determine a response action in response to the detection of the SDB. In some implementations, the processor is further configured to determine the intensity, magnitude, duration, and type of the SDB. In some implementations, the response action is one or more of an audible tone, an audible message, a trigger for a home automation device, a trigger for a voice assistant device, a call to an entity or an emergency service, marking of data for future access, an entry in a database, and a health diagnosis.

[0072] Generally, a method for determining item-specific parameters includes capturing acoustic signals and force signals from a plurality of non-contact sensors placed on a subject on a substrate, determining at least biometric signal information from the acoustic signals and the force signals, detecting the presence of speech in the acoustic signals, and communicating with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and a verbal command found in the detected speech.

[0073] In some implementations, the method further includes encrypting the digitized acoustic signal by at least one of filtering the digitized acoustic signal to lower and narrower frequencies, masking the digitized acoustic signal using a mask template or encryption key, and transforming the digitized acoustic signal using a mathematical formula. In some implementations, the method further includes discarding irrelevant conversations by comparing the acoustic signals to a dictionary of electronic commands, determining the presence of the verbal command, identifying the content of the speech when determining the verbal command, and initiating an interactive session with at least one of the subject or another entity based on the verbal command and the content of the speech and / or determining a response action based on the verbal command and the content of the speech via the acoustic interface. In some implementations, the method further includes recognizing and responding to verbal commands from a designated individual. In some implementations, the method further includes discarding irrelevant conversations by comparing the acoustic signals to a dictionary of electronic commands, determining the presence of the verbal command, analyzing the acoustic signals to detect a breathing disorder in the event of a failure to detect the verbal command, and determining a response action for the detection of sleep disordered breathing (SDB). In some implementations, the method further comprises performing at least one of cardiac analysis, respiratory analysis, and motion analysis based on the force signal to determine the health state of the subject. In some implementations, the method further comprises performing a respiratory disorder analysis to determine the health state of the subject, and the performing step further comprises generating a fused signal by fusing the force signal and the acoustic signal based on one or more similarity metrics, detecting a sleep disordered breathing (SDB) using the fused signal, the force signal, and the acoustic signal, and determining a response action in response to the detection of the SDB. In some implementations, the response action is one or more of an audible tone, an audible message, a trigger of a home automation device, a trigger of a voice assistant device, a call to an entity or emergency service, marking data for future access, an entry in a database, and a health diagnosis. In some implementations, the method further comprises determining the intensity, magnitude, duration, and type of the SDB. In some implementations, the method further comprises performing at least respiratory analysis based on the captured force signal, comparing the acoustic signal to a dictionary of electronic commands to discard irrelevant conversations, determining the presence of the verbal command, generating a fused signal by fusing the force signal and the acoustic signal based on one or more similarity metrics when the detection of the verbal command fails, detecting a sleep disordered breathing (SDB) using the fused signal, the force signal, and the acoustic signal, and determining a response action in response to the detection of the SDB.

[0074] Generally, the apparatus comprises a substrate configured to support a subject, a plurality of non-contact sensors configured to capture force signals regarding the subject, a processor connected to the plurality of sensors and configured to determine a biosignal from the force signals for monitoring the health state of the subject, and an acoustic interface configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and an oral command received via a speech function device.

[0075] Generally, the apparatus comprises a substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic signals and force signals regarding the subject, an acoustic interface configured to communicate with the subject, and a processor connected to the plurality of sensors and the acoustic interface. The processor is configured to determine a biosignal from one or more of the acoustic signals and the force signals for monitoring the health state of the subject and to receive speech detected in the acoustic signals from a speech detection entity. The acoustic interface is configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and an oral command within the detected speech.

[0076] The terms "example", "aspect", or "embodiment" are used in this specification to mean serving as an example, an instance, or an illustration. Any aspect or design described in this specification using one or more of these terms should not necessarily be construed as more preferred or advantageous than other aspects or designs. Rather, the use of the terms "example", "aspect", or "embodiment" is intended to present concepts in a specific aspect. As used in this application, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or". That is, unless otherwise specified or clear from the context, "X includes A or B" is intended to mean any natural inclusive permutation. That is, "X includes A or B" is satisfied when X includes A, when X includes B, or when X includes both A and B. Also, when used in this specification and the appended claims, the articles "a" and "an" should generally be construed to mean "one or more" unless otherwise specified or clear from the context that the singular form is being referred to.

[0077] Although the present disclosure has been described in connection with specific embodiments, it is to be understood that the present disclosure is not limited to the disclosed embodiments, but on the contrary, is intended to cover various modifications and equivalent configurations included within the scope of the appended claims. The claims should be given the broadest interpretation so as to encompass all such modifications and equivalent structures as permitted under the law.

Claims

1. A substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic signals and force signals regarding the subject, an acoustic interface configured to communicate with the subject, a processor connected to the plurality of non-contact sensors and the acoustic interface, wherein to monitor the health state of the subject, determine a biosignal from one or more of the acoustic signal and the force signal, detect the presence of speech in the acoustic signal, generate a fused signal by fusing the force signal and the acoustic signal based on one or more similarity metrics, detect sleep disordered breathing (SDB) using the fused signal, the force signal, and the acoustic signal, and determine a response action in response to the detection of the sleep disordered breathing (SDB), a processor configured as such, comprising, the acoustic interface is configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and a verbal command in the detected speech. An apparatus characterized by this.

2. The processor is further configured to encrypt the digitized acoustic signal by at least one of filtering the digitized acoustic signal to lower and narrower frequencies, masking the digitized acoustic signal using a mask template or a cryptographic key, and transforming the digitized acoustic signal using a mathematical formula. The apparatus according to claim 1, characterized by this.

3. The processor is further configured to compare the acoustic signal against a dictionary of electronic commands to discard irrelevant conversations, determine the presence of the verbal command, specify the content of the speech when the verbal command is determined, and execute at least one of starting a conversational session with at least one of the subject or another entity based on the verbal command and the content of the speech and determining a response action based on the verbal command and the content of the speech, via the acoustic interface. configured as such The apparatus according to claim 1, characterized by this.

4. The acoustic interface is further configured to recognize and respond to verbal commands from a designated individual. The apparatus according to claim 1, characterized by this. [

5. ] A substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic signals and force signals regarding the subject, an acoustic interface configured to communicate with the subject, a processor connected to the plurality of non-contact sensors and the acoustic interface, wherein, for monitoring the health state of the subject, the processor determines a biosignal from one or more of the acoustic signals and the force signals, and detects the presence of speech in the acoustic signals and is configured as such, comprising, the acoustic interface is configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and an oral command in the detected speech, the processor further, compares the acoustic signals against a dictionary of electronic commands to discard irrelevant conversations, determines the presence of the oral command, analyzes the acoustic signals to detect a breathing disorder in the event of a failure to detect the oral command, and determines a response action for the detection of sleep disordered breathing (SDB) and is configured as such characterizing the apparatus. [

6. ] A substrate configured to support a subject, a plurality of non-contact sensors configured to capture acoustic signals and force signals regarding the subject, an acoustic interface configured to communicate with the subject, a processor connected to the plurality of non-contact sensors and the acoustic interface, wherein, for monitoring the health state of the subject, the processor determines a biosignal from one or more of the acoustic signals and the force signals, and detects the presence of speech in the acoustic signals and is configured as such, comprising, the acoustic interface is configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and an oral command in the detected speech, the plurality of non-contact sensors are configured to capture force signals from the subject's movements relative to the substrate The processor is further configured to perform at least one of cardiac analysis, respiratory analysis, and motion analysis based on the force signal to determine the health status of the subject. Device characterized by this. **Claim 7** The response action is one or more of an audible tone, an audible message, a trigger for a home automation device, a trigger for a voice assistant device, a call to an entity or an emergency service, marking of data for future access, a database entry, and a health diagnosis. The device according to claim 1, characterized by this. **Claim 8** The processor is further configured to determine the intensity, magnitude, duration, and type of the sleep disordered breathing (SDB). The device according to claim 1, characterized by this. **Claim 9** A speech function device configured to communicate with at least one of a subject or an entity associated with the subject; A device communicating with the speech function device; Comprising: The device A substrate configured to support the subject; A plurality of non-contact sensors configured to capture an acoustic signal regarding the subject and a force signal from the subject's movement with respect to the substrate; A processor connected to the plurality of non-contact sensors and an acoustic interface, Monitoring the health status of the subject based on the force signal and the acoustic signal, Detecting an oral command in the acoustic signal, Comparing the acoustic signal with a dictionary of electronic commands to discard irrelevant conversations, Determining the presence of the oral command, When the detection of the oral command fails, generating a fusion signal by fusing the force signal and the acoustic signal based on one or more similarity metrics, Detecting a sleep disordered breathing (SDB) using the fusion signal, the force signal, and the acoustic signal, Determining a response action for the detection of the sleep disordered breathing (SDB). A processor configured as such; Having The speech function device is configured to communicate with at least the subject or the entity based on at least one of a response action required due to the health status of the subject and the detection of the oral command. System characterized by this. **Claim 10** The processor is further configured to encrypt the digitized acoustic signal by at least one of filtering the digitized acoustic signal to lower and narrower frequencies, masking the digitized acoustic signal using a mask template or encryption key, and transforming the digitized acoustic signal using a mathematical formula. The system according to claim 9, characterized in that.

11. The processor is further configured to compare the acoustic signal against a dictionary of electronic commands to discard irrelevant conversations, determine the presence of the verbal command, identify the content of the utterance upon determination of the verbal command, initiate an interactive session with at least one of the subject or the entity based on the verbal command and the content of the utterance via the speech function device, and determine the response action based on the verbal command and the content of the utterance, and execute at least one of them. configured to The system according to claim 9, characterized in that.

12. The speech function device is further configured to recognize and respond to verbal commands from a specified individual. The system according to claim 9, characterized in that.

13. The processor is further configured to determine the intensity, magnitude, duration, and type of the sleep disordered breathing (SDB). The system according to claim 9, characterized in that.

14. The response action is one or more of an audible tone, an audible message, a trigger for a home automation device, a trigger for a voice assistant device, a call to an entity or emergency service, marking of data for future access, an entry in a database, and a health diagnosis. The system according to claim 9, characterized in that.

15. A method of operating a device for determining item-specific parameters, comprising: capturing an acoustic signal and a force signal from a plurality of non-contact sensors placed on a subject on a substrate; determining at least biometric signal information from the acoustic signal and the force signal; detecting the presence of an utterance in the acoustic signal; Communicating with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health condition of the subject and a verbal command found in the detected utterance; Comparing the acoustic signal to a dictionary of electronic commands to discard irrelevant conversations; Determining the presence of the verbal command; Analyzing the acoustic signal to detect a breathing disorder when the detection of the verbal command fails; Determining a response action in response to the detection of sleep disordered breathing (SDB); A method characterized by comprising the above.

16. The method further comprises encrypting the digitized acoustic signal by at least one of filtering the digitized acoustic signal to lower and narrower frequencies, masking the digitized acoustic signal using a mask template or cryptographic key, and transforming the digitized acoustic signal using a mathematical formula. The method according to claim 15, characterized by the above.

17. The method further comprises: Comparing the acoustic signal to a dictionary of electronic commands to discard irrelevant conversations; Determining the presence of the verbal command; Identifying the content of the utterance when the verbal command is determined; Starting a dialog session with at least one of the subject or another entity based on the verbal command and the content of the utterance via an acoustic interface, and performing at least one of determining a response action based on the verbal command and the content of the utterance. The method according to claim 15, characterized by comprising the above.

18. The method further comprises: Recognizing and responding to a verbal command from a designated individual The method according to claim 15, characterized by comprising the above.

19. The method further comprises: Performing at least one of cardiac analysis, respiratory analysis, and motion analysis based on the force signal to determine the health condition of the subject The method according to claim 15, characterized by comprising the above.

20. The method further comprises: Performing a respiratory disorder analysis to determine the health condition of the subject comprising, The performing step further comprises: Generating a fused signal by fusing the force signal and the acoustic signal based on one or more similarity metrics. Detecting sleep disordered breathing (SDB) using the fusion signal, the force signal, and the acoustic signal; Determining a response action in response to the detection of the sleep disordered breathing (SDB); having The method according to claim 19, characterized in that.

21. The response action is one or more of an audible tone, an audible message, a trigger for a home automation device, a trigger for a voice assistant device, a call to an entity or an emergency service, marking of data for future access, a database entry, and a health diagnosis The method according to claim 20, characterized in that.

22. The method further includes Determining the intensity, magnitude, duration, and type of the sleep disordered breathing (SDB) The method according to claim 20, characterized by comprising.

23. An operating method of a device for determining item-specific parameters, comprising: Capturing an acoustic signal and a force signal from a plurality of non-contact sensors placed on a subject on a substrate; Determining at least biometric signal information from the acoustic signal and the force signal; Detecting the presence of speech in the acoustic signal; Communicating with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the health state of the subject and a verbal command found in the detected speech; Performing at least respiratory analysis based on the captured force signal; Comparing the acoustic signal to a dictionary of electronic commands to discard irrelevant conversations; Determining the presence of the verbal command; Generating a fusion signal by fusing the force signal and the acoustic signal based on one or more similarity metrics when the detection of the verbal command fails; Detecting sleep disordered breathing (SDB) using the fusion signal, the force signal, and the acoustic signal; Determining a response action in response to the detection of the sleep disordered breathing (SDB) A method characterized by comprising.

24. A bed configured to support a subject; A plurality of non-contact sensors configured to capture a force signal regarding the subject; A processor connected to the plurality of non-contact sensors and configured to determine a biometric signal from the force signal to monitor a sleep disordered breathing (SDB) of the subject An acoustic interface configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the subject's sleep disordered breathing (SDB) and an oral command received via a speech function device. An apparatus, characterized by comprising the same. **Claim 25** A bed configured to support a subject; A plurality of non-contact sensors configured to capture acoustic signals and force signals related to the subject; An acoustic interface configured to communicate with the subject; A processor connected to the plurality of non-contact sensors and the acoustic interface, configured to determine a physiological signal from one or more of the acoustic signals and the force signals for monitoring the subject's sleep disordered breathing (SDB); and configured to receive speech detected in the acoustic signals from a speech detection entity. A processor configured as such. An apparatus comprising the same. The acoustic interface is configured to communicate with at least one of the subject or an entity associated with the subject based on at least one of an action required due to the subject's sleep disordered breathing (SDB) and an oral command within the detected speech. An apparatus, characterized by comprising the same.

Images