Smart sensor and methods thereof

Abstract

A device includes one or more microphones, an output component, and a network interface. The device receives, from the one or more microphones, audio data associated with sound captured within an environment, and determines that the sound is associated with the user. Based on the sound being associated with the user, the device determines one or more biomarkers associated with the sound. A threshold associated with the user having a pathological or behavioral state is determined and when the one or more biomarkers satisfy the threshold, the sound is determined to be indicative of the user having the pathological or behavioral state. The device outputs a first notification on the output component and sends data associated with a second notification to be output on an electronic device associated with a caregiver of the user.

Description

SMART SENSOR AND METHODS THEREOFCROSS-REFERENCE TO RELATED APPLICATIONS

[0001] This application claims priority to U.S. Patent Application No. 18 / 755,471, filed June 26, 2024, entitled “Wearable Speech Therapy Device,” which claims priority to U.S. Provisional Patent Application No. 63 / 524,661, filed July 1, 2023, entitled “Wearable Speech Therapy Devices and Methods for Treating Speech Pathologies, Including Hypophonia associated with Parkinson’s Disease,” U.S. Patent Application No. 18 / 754,955, filed June 26, 2024, entitled “Veterinary Acoustic and Biofeedback Device,” which claims priority to U.S. Provisional Patent Application No. 63 / 524,663, filed July 1, 2023, entitled “Veterinary Acoustic Smart-Detector and Biofeedback Devices and Methods for Detecting and Analyzing Behavioral Activity Signals, for Monitoring Animal Welfare, Diagnosing, Reporting, and Remedying Animal Disease and Distress, and Cueing Animal Behavior to Improve Livestock, Wildlife, and Pet Management and Care,” and U.S. Patent Application No. 18 / 756,857, filed June 27, 2024, entitled “Wellness and Healthcare Monitoring Device,” which claims priority to U.S. Provisional Patent Application No. 63 / 524,667, filed July 1, 2023, entitled “Wearable Smart-Sensor and Biofeedback Devices and Methods for Detecting, Analyzing and Reporting Disease, Health Concerns and Other Adverse Welfare Conditions in Human Subjects,” the entirety of which are herein incorporated by reference.BACKGROUND

[0002] Monitoring wellness, health, and welfare of humans and animals is often labor-intensive and timeconsuming. Despite modem tools and advancements, however, diagnosing, treating, and managing welfare remains insufficient. For example, conventional monitoring and treatment often focuses on costly and time-consuming visits. While devices and / or wearables exist, these devices have limited functionality and inconclusive results. As such, current methods and tools for effective health monitoring are outdated, ineffective, and limited.BRIEF DESCRIPTION OF THE DRAWINGS

[0003] The detailed description is described with reference to the accompanying figures. In the figures, the left-most digit(s) of a reference number identifies the figure in which the reference number first appears. The use of the same reference numbers in different figures indicates similar or identical components or features. The systems depicted in the accompanying figures are not to scale and components within the figures may be depicted not to scale with each other.

[0004] FIG. 1 illustrates an example of a wearable speech therapy device for processing user speech and providing feedback to a user, according to examples of the present disclosure.

[0005] FIG. 2A illustrates an isometric view of the wearable speech therapy device of FIG. 1, according to examples of the present disclosure.

[0006] FIG. 2B illustrates a front view of the wearable speech therapy device of FIG. 1, according to examples of the present disclosure.

[0007] FIG. 3 illustrates an example process in which the wearable speech therapy device FIG. 1 provides feedback to the user, according to examples of the present disclosure.

[0008] FIG. 4 illustrates an example device for monitoring animal welfare to determine one or more pathological states of the animal, according to examples of the present disclosure.

[0009] FIG. 5 illustrates an example process for monitoring animal welfare to determine one or more pathological states of the animal, according to examples of the present disclosure.

[0010] FIG. 6 illustrates an example healthcare monitoring device used to determine a welfare, behavioral, and / or pathological state of a user, according to examples of the present disclosure.

[0011] FIG. 7 illustrates a plurality of healthcare monitoring devices used to determine a welfare, behavioral, and / or pathological state of a user, according to examples of the present disclosure.

[0012] FIG. 8 illustrates an example process in which the wellness and healthcare monitoring device determines a welfare, behavioral, and / or pathological state of a user, according to examples of the present disclosure.DETAILED DESCRIPTIONIllustrative Embodiment of First Device and Methods Thereof

[0013] This application is directed, at least in part, to a wearable speech therapy device that may detect, treat, manage, and / or prevent atypical speech and / or behavior. As used herein, “atypical speech and / or behavior” is intended to cover human behavior, whether volitional or not, primarily speech and language acts, that are non-normal, non-canonical, atypical, disordered, characteristic of disability / disease, or impaired. The term is intended to encapsulate speech and language, such as reading, perception, production, communication, and associated behaviors. In an embodiment, the wearable speech therapy device may receive audio data, such as speech from a user, and process the audio data to provide real-time therapeutic feedback to the user. For example, based on processing the audio data, the wearable speech therapy device may output notification(s) that prompt the user to adjust characteristics of their speech(e.g., tone, prosody, duration, amplitude, etc.). The wearable speech therapy device may continuously capture and process the audio data to provide real-time feedback in a corrective and therapeutic manner. In an embodiment, the wearable speech therapy device may process the audio data to detect low-volume (e.g., low amplitude) speech, such as low-volume speech associated with hypophonia. Users employing the wearable speech therapy device may experience increased speech perceptibility, improved speech production, independence, quality of life, productivity, and mental health.

[0014] In an embodiment, and as will be explained herein, the wearable speech therapy device may determine speech and / or vocal biomarkers associated with the audio data for detecting atypical speech. The speech and / or vocal biomarkers may be compared to reference characteristics, thresholds, etc., to detect when the speech is indicative of atypical speech. For example, the wearable speech therapy device may detect one or more speech and / or vocal biomarkers associated with hypophonia, which may be characterized by low-amplitude speech (e.g., hypophonic speech biomarkers). Atypical speech may also be characterized by breathy and hoarse vocal quality, reduced loudness, and / or reduced pitch and loudness variability.

[0015] In an embodiment, atypical speech may be associated with a variety of diseases and health-care- related conditions, including Parkinson's disease, Alzheimer's disease, Frontotemporal dementia, Vascular dementia, ALS / Lou Gehrig’s disease, hereditary Prior disease, congenital defects, conditions elicited or exacerbated by certain medications, and various forms of injury (e.g., injuries affecting the central nervous system (CNS), including the brain, and trauma impairing vocal anatomy or neuromotor function). In an embodiment, the wearable speech therapy device may associate the speech and / or vocal biomarkers with certain diseases, diagnoses, etc. The wearable speech therapy device may be used for other speech pathological, behavioral, psychomotor, and movement disorders, prodromal symptoms associated with disease or disorder, and pathogenic conditions.

[0016] In an embodiment, the wearable speech therapy device may be a compact device worn on a body of the user. The wearable speech therapy device may be worn in any manner by the user, such as in the form of bracelets, watches, stick-on arrays, eyeglasses, caps, hearing aids, brooches, ties, lanyards, earrings, etc. The wearable speech therapy device may include suitable attachment mechanisms, such as hook and loop, clasps, magnets, cords, chains, etc., for coupling (e.g., pinning, hanging, etc.) in any manner to the user. However, although described herein as being a wearable device worn by the user, in an embodiment, the techniques described herein for treating, detecting, managing, and / or preventing atypical speech may not be limited to such wearable devices. For example, a device (e.g., stand-alonedevice, integrated device, etc.) employing the techniques described herein may be placed on a shelf, counter, wall, etc., within an environment in which the user resides or occupies.

[0017] The wearable speech therapy device may include a housing in which components of the wearable speech therapy device are disposed. The housing may include any shape, such as an elongated cylindrical housing. The housing may be manufactured from any suitable materials, such as plastics, composites, metal, etc., and using any manufacturing process(es), such as injection molding, stamping, blow molding, etc. The housing may also be contaminate-resistant to prevent the ingress of liquid, dust, or contaminants into the housing. Additionally, the wearable speech therapy device may include suitable processing and internal components to detect, treat, manage, and / or prevent atypical speech. For example, the wearable speech therapy device may include sensor(s), microphone(s), battery(ies), processing components (e.g., processors, memory, system on a chip (SoC) or integrated circuits (IC), etc.), etc. disposed within the housing.

[0018] In an embodiment, one or more button(s) may be included to control one or more operation(s) of the wearable speech therapy device. The button(s) may be used to power on / off the wearable speech therapy device, mute the wearable speech therapy device, change setting(s) of the wearable speech therapy device, connect to nearby devices, etc. A user interface may functionality allow the user to interface with the wearable speech therapy device. The user interface may include an organic light emitting diode (O / LED), (in-plane switching [IPS] orthin-fdm-transistor [TFT]) liquid crystal display (LCD), laser video display (LVD), vibrotactile, haptic, or other applications.

[0019] The wearable speech therapy device may include any number of sensor(s) disposed about or within the housing for detecting when the user speaks, voice and speech quality and quantity, environmental sensor input including non-speech acoustic input, and / or when the user is mute (e.g., not speaking). Using the sensor(s), the wearable speech therapy device may determine the speech and / or vocal biomarkers associated with the user’s speech. For example, the wearable speech therapy device may include microphone(s) for collecting audio signals associated with the user’s speech. The audio signals may be processed as data, which is analyzed to determine the speech and / or vocal biomarkers. The wearable speech therapy device may include analog and digital converters in the capture, processing, and storage of signals sampled from the sensors. The signals from the microphone(s) and sensor(s) may be stored in a dataset that includes raw and / or normalized or transformed acoustic wave values.

[0020] In an embodiment, the microphone(s) may include at least two microphones spaced apart from one another within the housing. For example, the housing may include a first end and a second end spacedapart from the first end. A first of the microphones may be located closer to the first end than the second end, while a second of the microphones may be located closer to the second end than the first end. Any number of microphones, however, may be included within the wearable speech therapy device. The microphone(s) may be used for determining the directionality (e.g., time of flight, velocimetry, time-delay analytics, etc.) of the signal and / or whether the signal corresponds to the speech of a user wearing the wearable speech therapy device. For example, the microphone(s) may receive respective audio signals that are representative of sounds within the environment. When worn by the user, the microphones may be spaced apart by different distances from sound sources of the user (e.g., a mouth of the user). For example, when the wearable speech therapy device represents a device that hangs from a neck of the user (e.g., as a necklace), the microphones may be located at different distances from the mouth of the user. As such, as the microphone(s) are spaced apart by a different distance from the mouth of the user, user speech may be discerned from other sources of sound within the environment of the user.

[0021] In an embodiment, other sensors of the wearable speech therapy device, such as piezoelectric sensors, accelerometers, Hall sensors, passive infrared (PIR) sensors, etc., may receive signals that are processed to accumulate and analyze sensor data that is used to determine when the user speaks, user behavior, and / or which is used to determine the speech and / or vocal biomarkers. Still, sensors may additionally or alternatively include global positioning satellite (GPS), gyroscope(s), inertial measurement units (IMU), etc. In an embodiment, GPS may record a location and / or track at-risk users (e.g., persons with dementia) and / or be used to calibrate the wearable speech therapy device (e.g., at different elevations, acoustic environments, language locales, environmental concerns, etc ). The gyroscope, IMU, and / or Hall sensor may be used to detect and alert third parties in the event of falls or accidents. The sensor data generated from the sensor(s), such as an IMU, may be used to determine when the user speaks, for example, based on experienced vibrations, accelerations, inhaling / exhaling, pressure changes, and so forth.

[0022] The wearable speech therapy device may employ voice activity detection (VAD) to determine the speech of the user or to recognize sounds generated by the user or other talkers. VAD may be used to recognize the speech or behavior of the user as compared to other sounds within an environment of the user (e.g., background noise, speech from other persons in the environment, etc.). For example, the wearable speech therapy device may include a VAD component to perform VAD techniques on the audio data (or audio signals) captured by the microphones. The VAD component may utilize one or more VAD algorithms based on channel energy with long and short-term energies, sub-band long and short-termenergies with combination logic, Deep Neural Network (DNN) based VADs, or any other type of VAD algorithms, with shaped and / or hangover windows, to determine whether the audio data is representative of user speech. VAD is representative of one technique to detect speech and laryngeal activity in a signal. Other techniques exist to detect, segregate, label, and classify target speech, other vocalizations, and nonspeech vocalizations. For example, statistical likelihood pattern-matching models may also be employed using diphone, triphone, or phone templates. The wearable speech therapy device may additionally or alternatively employ modern methods of talker identification and (target) talker identification.

[0023] The audio signals generated by the microphones may be cross-correlated. For example, crosscorrelation may applied between audio signals detected by the microphones to determine a time difference of arrival (TDOA) for the correlated signals. In other words, the differently spaced microphones from the mouth of the user have different time of arrival (TOA) values. This allows the calculation of an angle from which the detected signals originate relative to a line intersecting the two microphones. Because the mouth of the user is approximately in line with the two microphones (e.g., when the wearable speech therapy device is worn as a necklace), a threshold may be applied on the TDOA, and when the TDOAs exceed this threshold, the audio signals may be determined to be speech of the user, and non-speech sound, behavior of users or others, or speech of another talker.

[0024] Once the user speech is determined, using VAD, for example, the audio signals may be further processed. For example, given that the wearable speech therapy device may be used in a plurality of diverse environments of the user, user speech may be detected based on a signal-to-noise ratio (SNR). To determine SNR, a decibel level of the speech of the user is determined, and the decibel level of the speech of the user is subtracted from a decibel level of a background or non-target-user information. In an embodiment, the decibel level of the user may be determined by the following equation. dB dBe dBu lQio = 10Ao" + 10Vo“ (1) where: dB represents the current decibel reading (which includes the speech of the user and environmental noise; dBe represents the decibel level of the environment, and dBu represents the decibel level of the user.

[0025] Subsequently, a signal to noise (SNR) ratio may be determined by the following equation.where: dBe represents the decibel level of the environment, and dBu represents the decibel level of the user.

[0026] In an embodiment, the decibel level of the user and the decibel level of the environment may be determined by calibrating the microphone closest in proximity to the mouth of the user and using a sound level meter. The microphone may be placed next to a sound level meter, and the log of the root mean square (log RMS) may be recorded over an interval of time (e.g., 0.5 seconds), along with a corresponding decibel reading on the sound level meter. Once data is collected over the decibel range of interest (e.g., 40-70dB), a linear regression is applied to obtain a function that takes logRMS input and outputs decibel values. Using this calculation, the decibel level of the environment may be determined by averaging the decibel readings over a specified time frame (e.g., 2 seconds, 3 seconds, etc.) of noise that was determined to be environmental noise as calculated by equation (2) above. Further, the time analysis window may be empirically determined, user-specific, dynamic, adjustable, and / or modulated according to other factors. The time window may be modulated to affect performance according to user needs. For example, if the user is in a known noisy environment, such as a public event, the time window may be adjusted to make the performance of the wearable speech therapy device more or less sensitive. The audio data may be analyzed after determining SNR to improve VAD and / or accuracy in detecting hypophonia.

[0027] However, as noted above, sensor data generated by the sensor(s) may be used in combination with the audio data to determine when the user is speaking. For example, the sensor(s) may detect vocal vibrations by detecting changes in pressure, acceleration, temperature, strain, or force. Times associated with generating the audio data and the sensor data may be associated with one another to correlate the sensor data with the audio data. To effectively detect vocalization, the sensor(s) may be positioned in proximity to the clavicle or thyroid / cricoid cartilage or near vocal folds of the user to better detect movement or pressure changes associated with vocalization.

[0028] Once speech of the user is detected, the audio data and / or sensor data may be analyzed to determine characteristics of the user speech. The characteristics may include the speech and / or vocal biomarkers, where the speech and / or vocal biomarkers may identify or be associated with characteristics of the user speech. In an embodiment, the speech and / or vocal biomarkers may be associated with how the user speaks, such as pitch, intonation, tone, pauses, phonation, and / or changes associated with the user’s speech. The language, speech, and / or vocal biomarkers may also include features such as articulation, decreased energy in the higher parts of a harmonic spectrum, imprecise articulation of vowels andconsonants, fundamental frequency, voicing, windowed and absolute syllable / sonorant-peak rates, SNR, temporal and spectral voice characteristics, frequency, spectral / cepstral cues, vocal quality / stability (e.g., shimmer, jitter, harmonic to noise ratio), prosody, temporal output, amplitude stability, and / or motor processes.

[0029] In an embodiment, the speech and / or vocal biomarkers may be associated with certain portions of the user speech. For example, when the user pauses during speech, portions of the user speech that include the pauses may be identified (e.g., via metadata). In an embodiment, the speech and / or vocal biomarkers may be characterized by duration, frequency, decibel level, amplitude, energy, etc. Although described as speech and / or vocal biomarkers, the sensor data captured by the sensor(s) may be used to determine the speech and / or vocal biomarkers associated with the user’s speech. Additionally, the speech and / or vocal biomarkers may be stored to provide pathology estimation and behavioral characteristics, and sample comparative algorithms may be used to diagnose, qualify, classify, quantify, and monitor the speech pathological condition, status, and therapeutic progress.

[0030] In an embodiment, the wearable speech therapy device may analyze the audio data, the sensor data, and / or the SNR to determine the speech and / or vocal biomarkers for use in detecting atypical speech, such as low-volume speech associated with hypophonia. The amplitude of the user speech may be compared to one or more thresholds. In an embodiment, the thresholds may be based at least in part on the SNR and / or the user and may be dynamically determined. For example, if the amplitude associated with user speech is less than the threshold, the wearable speech therapy device may detect and classify hypophonia. Comparatively, if the user speech is greater than the threshold, hypophonia may not be detected and classified as such. In an embodiment, the wearable speech therapy device may determine any number of speech and / or vocal biomarkers for detecting atypical speech, such as hypophonia in the user’s speech. The speech and / or vocal biomarkers may be compared to respective thresholds to determine whether the speech and / or vocal biomarkers are indicative of hypophonia. The speech and / or vocal biomarkers, in addition to indicative low volume and known pathologies of the speech production physiology, may also include other vocal and acoustic features and values, both normal / typical and pathological, to improve operative flexibility and accuracy and therapeutic utility of the wearable speech therapy device.

[0031] In an embodiment, the speech and / or vocal biomarkers may be compared to reference characteristics of speech from users with hypophonia as well as from users without hypophonia. Through this comparison, a statistical likelihood may be determined whether the speech and / or vocal biomarkersare indicative of hypophonia or not indicative of hypophonia. In an embodiment, the wearable speech therapy device may determine voice fingerprints or vocal phenotype. The use of the voice fingerprints may increase an accuracy in detecting speech of the user, or discerning speech of the user from other sounds emanating within the environment. Moreover, the voice fingerprints may be used to more accurately detect atypical speech, such as hypophonia. For example, in an embodiment, the speech and / or vocal biomarkers may be used to generate voice fingerprints associated with the user, and the voice fingerprint may be analyzed to determine whether the voice fingerprint is associated with or includes hypophonia. The voice fingerprint may characterize the user speech. In an embodiment, the voice fingerprint may be compared to one or more reference fingerprints to detect hypophonia. In addition to fingerprints of target speech behavior, other audio and acoustic fingerprint / phenotype biomarkers may be used in the wearable speech therapy device. For example, characteristic environmental auditory events such as crowd noises, door slamming, and household implements (e.g., blender, toilet flushing, alarms, etc.) have unique acoustic signatures that may be used in combination or solely to better interpret the auditory and acoustic environment of the wearable speech therapy device or the user. There may be audio or acoustic biomarkers or enviro-markers used in the determination and classification of signals.

[0032] The wearable speech therapy device may output the notification(s) in an attempt to signal the user to correct their speech and / or behavior. The notification(s) may be output in real-time or substantially real-time to provide near-instantaneous feedback to the user to correct their speech and / or behavior. In an embodiment, the notification(s) may be audible, visual, olfactory, electronic, haptic, and / or any combination thereof. For example, the wearable speech therapy device may include a speaker that outputs audio (e.g., deeps, tones, instructions, etc.), lighting elements that illuminate (e.g., patterns, colors, etc.), and / or a motor that vibrates to prompt the user to correct their speech. By outputting the notification(s), the wearable speech therapy may provide real-time therapeutic assistance to alert the user to correct their speech and / or behavior. For example, the notification(s) may prompt or inform the user to raise the volume of their speech. The wearable speech therapy device may output the notification(s) for a predetermined amount of time, continuously, and / or until the speech and / or behavior of the user is corrected. Using this feedback, the user may raise the volume of their speech, and the wearable speech therapy device may subsequently capture additional audio data and / or sensor data that is analyzed to determine whether the user has corrected their speech and / or behavior. As such, the notification(s) may serve to signal the user to correct their speech and / or behavior based on objective measurements. In addition to real-time feedback as described above, asynchronous reports may also be delivered to the userand stored in the software. For example, a daily histographic digest of performance and behavior may be delivered via email or via software (e.g., a mobile application) to the user for offline use. Performance data may also be delivered to a database collection performance characteristics from many users, constituting the basis for additional software and algorithm changes. The application and / or software may be embedded into the hardware / device or may be external software capable of being instantiated into external hardware such as a mobile phone, tablet, or computer.

[0033] In an embodiment, the wearable speech therapy device may tailor the notification(s) based on one or more preference(s) stored in association with a user profile of the user. For example, the user profile may indicate a type of notification(s) to output, as well as their associated intensity (e.g., volume, luminosity, etc.), duration, type (e.g., sound-haptic-sound), patterns (e.g., low haptic-medium haptic-low haptic, etc.), combinations thereof, etc. The wearable speech therapy device may access the preference(s) as stored in the user profile when outputting the notification(s).

[0034] In addition to processing the audio data and / or the sensor data for detecting speech and behavioral disorder, the wearable speech therapy device may utilize other data. For example, the wearable speech therapy device may have access to user data stored within a user profile of the user. The user data may indicate demographics of the user (e g., sex, age, disease state, audiological status, medical history, occupation, residence and care data, socioeconomic status, etc.), environmental factors (e.g., time of day, medication schedule, auditory and acoustic environment, etc.), a schedule of the user (e.g., when the user is sleeping, working, exercise, eating, socially interacting, etc.), medications of the user (e.g., medication type and medication schedules), etc. This user data may additionally, or alternatively, be used to treat, detect, manage, and / or prevent atypical speech and / or behavior such as hypophonia. For example, incorporating user-specific, treatment-related, non-vocal data may be used to more effectively treat, manage, and / or prevent hypophonia. Another example is using data or user input to aid the user in scheduling appointments or follow up visits with health care providers (e.g., speech language pathologist, physician, occupational or physical therapist, dentist, neurologist, or similar) or other professionals. The wearable speech therapy device and software may, based on inputs, give suggestions, advice, or recommendations including referrals based on data from user or database information. The wearable speech therapy device may also afford the user to interact with their existing disease specialist for monitoring or services.

[0035] In an embodiment, the wearable speech therapy device may utilize artificial intelligence (Al) and / or machine-learning (ML) model(s) to detect speech or behavioral characteristics. To clarify, Al andML techniques, which are described and detailed with examples below, are used as encompassing cover terms for “smart” and algorithmic collection, processing, analyzing, databasing, managing, summarizing, and using data in an automated and large-scale fashion, in this case frequently speech and behavioral data. The ML model(s) may be trained from a database (e.g., historical data, such as audio data that includes atypicality such as hypophonia and / or typical speech and behavior normalized data) to analyze the audio data captured by the sensor(s) and microphone(s) for identifying the speech and / or vocal biomarkers. The ML model(s) may also be trained from other sensor data that was indicative of atypicality and / or where atypicality was not present. The ML model(s) may determine speech and / or vocal biomarkers and may assess the speech and / or vocal biomarkers in comparison to information stored in the database (e.g., reference characteristics, reference voice fingerprint, etc.) to determine whether atypicality was detected in the user speech and / or behavior.

[0036] As part of the ML model(s) analyzing the audio data and / or the sensor data, the ML model(s) may label the characteristic(s) of the speech and / or behavior, for example, the speech and / or vocal biomarkers, to indicate whether the characteristic(s) are associated with atypicality. In an embodiment, an output of the ML model(s) may indicate whether the speech contains hypophonia. In an embodiment, the ML model(s) may determine the characteristics(s) for comparison to respective references for determining whether the speech contains hypophonia or is free of hypophonia. The ML model(s) may use any number of character! stic(s) for determining whether the speech contains hypophonia or is free of hypophonia.

[0037] In an embodiment, the ML model(s) may generate scores that are associated with the characteristic(s) of the typical or atypical speech. For example, the ML model(s) may generate scores associated with the speech and / or vocal biomarkers. If the scores are greater than a threshold, the ML model(s) may have a threshold confidence that the speech is associated with disorder or atypicality. Atypicality may be assessed with regard to the model parameters, and a discrete or probabilistic classification will be generated. The classification may be stored, processed, and used for feedback (as needed).

[0038] In an embodiment, the wearable speech therapy device may be calibrated during or between uses. Calibration may yield more sensitive, situationally adjusted, and therapeutically useful when detecting atypicality. The calibration and sensitization may operate by combining, for example, detection and coordinate processing of vocal amplitude along with one or more discrete user voice characteristics (e.g., non-amplitude vocal biomarkers) for a user, during a use session or between use sessions, to refine or otherwise modify subsequent detection of user-specific speech pathological data for output thenotification(s). Calibration may also involve detecting sound from other sources within the environment of the user, such as non-speech background noise (e.g., any substantial, persistent or intermittent background noise occurring in the environment, such as a television, appliances, etc.). For example, the wearable speech therapy device may detect sound from multiple sources, including the user and at least one non-user source, and through differential detection and / or processing, the wearable speech therapy device may self-adjust or calibrate, whether in real-time or throughout repeated use sessions to more accurately assign user vocal signals as user-originated, and more effectively distinguish third party and other background acoustic signals from sources other than the user.

[0039] Sounds generated from sources other than the user may be filtered to allow for more accurate detection of atypical speech and / or behavior. In an embodiment, the wearable speech therapy device may assign portions of the audio data as being user-originated and non-user-originated to more effectively detect atypicality, such as hypophonia. In an embodiment, the notification(s) may be based at least in part on a level of non-user-originated audio sources, for example, to inform the user that conversation or other background noise is occurring, requiring a greater correction of low-amplitude speech.

[0040] As described above, the wearable speech therapy device may provide real-time feedback to the user when atypical speech and / or behavior is detected. However, the wearable speech therapy device may record audio data over periods of time (e.g., hours, days, etc.) to provide a history of the user’s speech. The audio data and / or the detection of atypical speech and / or behavior may be stored in association with the user profile, a database, etc. Such data may be used by the user, care provider, and / or a speech pathologist to further refine the therapeutic benefits made possible by the wearable speech therapy device. Moreover, the audio data and / or the sensor may be used to retrain the ML model(s), for example, to increase an accuracy of detecting user speech and / or atypical speech and / or behavior in future instances. The audio data and / or sensor data may also be stored in memory of the device and provided (e.g., downloaded) to other devices.

[0041] Although the wearable speech therapy device is described as processing the audio data and / or the sensor data, in an embodiment, the wearable speech therapy device may communicatively couple to one or more separate devices for processing the data. For example, the wearable speech therapy device may communicatively couple to a mobile device, laptop, etc. of the user, whereby the mobile device may process the audio data and / or the sensor data to determine the speech and / or vocal biomarkers for determining atypical speech and / or behavior. The wearable speech therapy device may send the audio data and / or the audio data to the mobile device for processing, and the mobile device may transmitnotifications back to the wearable speech therapy device for output. Microphone(s) and / or sensor(s) of the mobile device may additionally or alternatively be used to detect atypical speech and / or behavior.

[0042] Additionally, the wearable speech therapy device and / or the mobile device may communicatively couple to remote computing resource(s) (e.g., cloud) for processing the collected data. Any level of split processing may be performed by the wearable speech therapy device, the mobile device, the remote computing resource(s), and / or other devices, systems, networks, etc. The wearable speech therapy device may include suitable network interface(s), such as Bluetooth, Cellular, Wi-Fi, etc. for communicatively coupling with such devices.

[0043] In an embodiment, the wearable speech therapy device may transition between states, such as an off / sleep state and an on / awake state. Transitioning between the states may conserve battery power of the wearable speech therapy device. For example, after a predetermined amount of time, or latent period (e.g., five or ten minutes) transpires without detected sounds, or VAD, the wearable speech therapy device may power off certain components (e.g., network interfaces, VO components, etc.).

[0044] The present disclosure provides an overall understanding of the principles of the structure, function, device, and system disclosed herein. One or more examples of the present disclosure are illustrated in the accompanying drawings. Those of ordinary skill in the art will understand and appreciate that the devices, the systems, and / or the methods specifically described herein and illustrated in the accompanying drawings are non-limiting embodiments. The features illustrated or described in connection with one embodiment, or instance, may be combined with the features of other embodiments or instances. Such modifications and variations are intended to be included within the scope of the disclosure and appended claims.

[0045] Vocal amplitude may be determined based on detected values from the sensor(s) 116, which give both negative and positive values that are not normalized. To resolve these problems, the detected values are squared and then the square root is computed (to determine an absolute value from the combined negative and positive values) and these data are normalized to fit into an interpretable range. Raw vocal data from the sensor(s) 116 are typically channeled through an analog-to-digital converter (comprising a sensor transducer and a modulator). The sensor transducer receives raw data from a plurality of sensors. The analog-to-digital converter generates a set of raw data and normalized data comprising raw data and normalized acoustic wave data processed from each sensor. The processor receives the set of raw data and normalized data and processes it to generate processed speech and behavioral data, and based on processing results a user biofeedback signal is either generated or not generated.

[0046] In an embodiment, other examples of the sensor(s) 1 16 of the wearable speech therapy device 102, such as piezoelectric sensors, accelerometers, etc. may generate signals that are processed to generate sensor data 124 that is used to determine when the user 100 speaks, and / or which is used to determine the speech and / or vocal biomarkers 110. The sensor(s) 116 may additionally or alternatively include global positioning satellite (GPS), gyroscope(s), inertial measurement units (IMU), etc. The sensor data 124 generated from the sensor(s) 116, such as an IMU, may be used to determine when the user 100 speaks, for example, based on experienced vibrations, accelerations, inhaling / exhaling, pressure changes, and so forth. The sensor(s) 116 may be calibrated to optimize function and increase processing fidelity. The sensor data 124 collected by the sensor(s) 116 may be used in combination with the audio data 104 to determine when the user 100 is speaking. Timestamps associated with the audio data 104 and the sensor data 124 may be associated with one another to correlate the sensor data 124 with the audio data 104. To effectively detect vocalization, the sensor(s) 116 may be positioned in proximity to the clavicle or thyroid / cricoid cartilage, or near vocal folds of the user 100 to better detect movement or pressure changes associated with vocalization.

[0047] Once user speech 106 is detected, the audio data 104 and / or sensor data 124 may be analyzed to determine the speech and / or vocal biomarkers 110. In an embodiment, the speech and / or vocal biomarkers 110 may be associated with how the user 100 speaks, such as pitch, intonation, tone, pauses, phonation, changes associated with the user speech 106, articulation, decreased energy in the higher parts of a harmonic spectrum, imprecise articulation of vowels and consonants, fundamental frequency, voicing, windowed and absolute syllable / sonorant-peak rates, SNR, temporal and spectral voice characteristics, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, and / or motor processes. In an embodiment, the speech and / or vocal biomarkers 110 may be associated with certain portions of the user speech 106. For example, when the user 100 pauses, portions of the user speech 106 (or the audio data 104) that include the pauses may be identified (e.g., via metadata). In an embodiment, the speech and / or vocal biomarkers 110 may be characterized by duration, frequency, decibel level, amplitude, energy, etc.

[0048] In an embodiment, the wearable speech therapy device 102 may process the audio data 104 and / or the sensor data to detect low-volume (e.g., low amplitude) speech, such as low-volume speech associated with hypophonia or other atypical speech and / or behavior. The low-volume speech, for example, may be indicated within the speech and / or vocal biomarkers 110. The speech and / or vocal biomarkers 110 may indicate an amplitude of the user speech 106, and the amplitude of the user speech 106 may be comparedto one or more thresholds or reference character! stic(s) indicated of low-volume speech associated with hypophonia. In an embodiment, the thresholds may be based at least in part on SNR calculated form the audio and / or specific(s) of the user 100 (e.g., baseline volume). In an embodiment, the threshold(s) to which the amplitude of the user speech 106 is compared, or more generally the threshold(s) to which the speech and / or vocal biomarkers 110 are compared, may be dynamically determined.

[0049] As an example, if an amplitude associated with user speech 106 is less than the threshold, the wearable speech therapy device 102 may detect hypophonia. Comparatively, if the amplitude of the user speech 106 is greater than the threshold, hypophonia may not be detected. In an embodiment, the wearable speech therapy device 102 may determine any number of the speech and / or vocal biomarkers 110 for comparison to respective thresholds for detecting hypophonia in the user speech 106. In this manner, the speech and / or vocal biomarkers 110 may be compared to reference characteristics of speech from users with hypophonia as well as from users without hypophonia. Through this comparison, a statistical likelihood may be determined whether the speech and / or vocal biomarkers 110 are indicative of hypophonia, or not indicative of hypophonia.

[0050] In an embodiment, the wearable speech therapy device 102 may determine voice fingerprint(s) 126 (or phenotype). The use of the voice fingerprint(s) 126 may increase an accuracy in detecting the user speech 106, or discerning speech of the user 100 from other sounds emanating within the environment. Moreover, the voice fingerprint(s) 126 may be used to more accurately detect atypical speech and / or behavior within the user 100. For example, in an embodiment, the speech and / or vocal biomarkers 110 may be used to generate voice fingerprint(s) 126 (e.g., audio signatures, acoustic fingerprint, voice signature, voiceprint, etc.) associated with the user 100, and the voice fingerprint(s) 126 may be analyzed to determine whether the voice fingerprint(s) 126 is associated with, or includes, disordered speech or behavior. Generally, the voice fingerprint(s) 126 may characterize the user speech 106. In an embodiment, the voice fingerprint(s) 126 may be compared to one or more reference fingerprints to detect atypical and / or disordered speech or behavior. For example, the voice fingerprint(s) 126 may be compared against reference audio voiceprints associated with hypophonia. Each voice fingerprint(s) 126 may indicate speech and / or vocal biomarkers 110, and if a similarity between the voice fingerprint(s) 126 and a stored reference audio voiceprint exists, hypophonia may be detected.

[0051] In an embodiment, the wearable speech therapy device 102 may output notification(s) 128 in instances where atypical speech and / or behavior is detected and / or where atypical speech and / or behavior is not detected. The notification(s) 128 may provide feedback to the user 100 in a therapeutic manner,thereby assisting the user 100 in speech therapy. For example, if atypical speech and / or behavior is not detected, the notification(s) 128 may provide indications of such, thereby informing the user 100 of their proper speech. Comparatively, if atypical speech and / or behavior is detected, the notification(s) 128 may provide indications of such, thereby informing the user 100 to correct or modulate their speech. The notification(s) 128 may be output in real-time, or substantially real-time, to provide near-instantaneous feedback to the user 100.

[0052] In an embodiment, the notification(s) may be associated with the speech and / or vocal biomarkers 110 that are detected. For example, if the user speech 106 has a low-amplitude, the notification(s) 128 may indicate such. As another example, if the user speech 106 is indicative of stammering, the notification(s) 128 may indicate such. As yet another example, if the user speech 106 is not well articulated (e.g., slurring), etc. the notification(s) 128 may indicate such. As such, the notification(s) 128 may convey specific information and / or prompts to the user 100 to correct their speech.

[0053] In an embodiment, the notification(s) 128 may be audible, visual, haptic, and / or a combination thereof. The wearable speech therapy device 102 may include I / O component(s) 130 that output the notification(s) 128. For example, lighting elements (e.g., LEDs) may signal the user 100 by outputting light at different brightness levels, flashing at different durations, frequencies and / or specific sequences (e.g., numeric sequences of long and short flashes) in different modes, and / or by lighting in different colors. Speaker(s) may output an audible signal, such as a continuous prompt tone, sounding at different pitches or loudness in different modes, sounding at different durations, frequencies and / or specific sequences (e.g., numeric sequences of loud and soft, or high and low pitch sounds) in different modes. In an embodiment, the speaker may additionally or alternatively be activated to provide programmed voice instructions to the user 100, for example, to provide more specific and personalized speech therapy instructions (e.g., prompting the user 100 “low speech detected, speak up please”, or “high background noise detected”, etc.)

[0054] The notification(s) 128 may also be output by a haptic motor. For example, the haptic motor may vibrate to prompt the user 100 via a continuous tactile stimulus. The vibration may be low, medium or high intensity, by intermittent tactile stimuli of different durations, frequencies and / or specific sequences (e.g.. numeric sequences of intermittent tactile stimuli differing in periodicity and / or intensity) to communicate specific information / prompts to the user 100.

[0055] Other output components of the wearable speech therapy device may include displays, touchscreens, etc. Example input devices may include button(s), switches, toggles, etc.

[0056] By outputting the notification(s) 128, the wearable speech therapy device 102 may provide realtime therapeutic assistance to alert the user 100 to correct their speech. For example, the notification(s) 128 may prompt or inform the user 100 to raise their volume. The wearable speech therapy device 102 may output the notification(s) 128 for a predetermined amount of time, continuously, and / or until the user speech 106 is corrected. For example, using the feedback provided by the notification(s) 128, the user may raise their volume. The wearable speech therapy device 102 may subsequently capture the audio data 104 and / or sensor data 124 that is analyzed to determine whether the user 100 has corrected their speech. As such, the notification(s) 128 may serve to signal the user to correct their speech based on objective measurements.

[0057] In an embodiment, the wearable speech therapy device 102 may tailor the notification(s) 128 based on one or more preference(s) stored in association with a user profile 132 of the user 100. For example, the user profile 132 may indicate a type of the notification(s) 128 to output, as well as their associated intensity (e.g., volume, luminosity, etc.), duration, type, patterns, combinations thereof, etc. The wearable speech therapy device 102 may access the preference(s) as stored in the user profile 132 when outputting the notification(s) 128.

[0058] In an embodiment, the wearable speech therapy device 102 may store or have access to machine- learned (ML) model(s) 134. The ML model (s) 134 may be trained to identify the user speech 106, or behavior, and / or may be trained to analyze the audio data 104 and / or the sensor data 124 to determine whether the user speech 106 corresponds to atypical speech and / or behavior. In an embodiment, the ML model(s) 134 may determine the speech and / or vocal biomarkers 110 for determining whether the user speech 106 is associated with hypophonia. In an embodiment, the ML model(s) 134 may determine or generate score(s) associated with the audio data 104 and / or the sensor data 124, to determine whether the user speech 106 contains atypical speech and / or behavior. In an embodiment, the score(s) may relate to a probability or likelihood that the user speech 106 is associated with or contains atypical speech and / or behavior. In other words, the score(s) output by the ML model(s) 134 may be machine-learned scores.

[0059] Machine learning generally involves processing a set of examples (called “training data”) to train a machine learning model(s). A machine learning model(s), once trained, is a learned mechanism that can receive new data as input and estimate or predict a result as output. For example, a trained machine learning model may comprise a classifier that is tasked with classifying unknown input (e.g., an unknown audio) as one of multiple class labels. In some cases, a trained machine learning model is configured to implement a multi-label classification task. Additionally, or alternatively, a trained machine learningmodel may be trained to infer a probability or a set of probabilities, for a classification task based on unknown data received as input. In the context of the present disclosure, the unknown input may be the audio data 104 and / or the sensor data 124 that is associated with the user speech 106, and the ML model(s) 134 may be tasked with outputting the score that indicates or otherwise relates to, a probability of the user speech 106 containing atypical speech and / or behavior. Additionally, the scores may indicate or otherwise relate to a probability of the speech and / or vocal biomarkers 110 for determining whether the user speech 106 is associated with atypical speech and / or behavior such as hypophonia. Other data, such as behavior data of the user 100 may be provided as an input to the ML model(s) 134.

[0060] The training data that is used to train ML model(s) 134 may include various types of data. In general, training data for machine learning may include two components, features and labels. However, in an embodiment, the training data used to train the ML model(s) 134 may be unlabeled. Accordingly, the ML model(s) 134 may be trainable using any suitable learning technique, such as supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, and so on. The features included in the training data may be represented by a set of features, such as in the form of an n-dimensional feature vector of quantifiable information about an attribute of the training data. The following is a list of example features that can be included in the training data for training the ML model(s) 134 described herein. However, it is to be appreciated that the following list of features is non-exhaustive, and features used in training may include additional features not described herein, and, in some cases, some, but not all, of the features listed herein. Example features included in the training data may include, without limitation, pitch, intonation, tone, pauses, phonation, changes associated with the user speech 106, articulation, decreased energy in the higher parts of a harmonic spectrum, imprecise articulation of vowels and consonants, fundamental frequency, voicing, windowed and absolute syllable / sonorant-peak rates, SNR, temporal and spectral voice characteristics, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, and / or motor processes. In an embodiment, the features included within the training data may be associated with user speech that contains atypical speech and / or behavior and / or user speech that does not contain atypical speech and / or behavior.

[0061] In an embodiment, as part of the training process, weights may be applied to a set of features included in the training data, as derived from the historical data. In an embodiment, the weights that are set during the training process may apply to parameters that are internal to the ML model(s) 134 (e.g., weights for neurons in a hidden-layer of a neural network). These internal parameters of the ML model(s) 134 may or may not map one-to-one with individual input features of the set of features. The weightsmay indicate the influence that any given feature, parameter, or characteristic has on the score that is output by the ML model(s) 134.

[0062] The ML model(s) 134 may represent a single model or an ensemble of base-level machine learning models and may be implemented as any type of machine learning model. For example, suitable machine learning models for use with the techniques and systems described herein include, without limitation, neural networks, tree-based models, support vector machines (SVMs), kernel methods, random forests, splines (e.g., multivariate adaptive regression splines), hidden Markov model (HMMs), Kalman fdters (or enhanced Kalman filters), Bayesian networks (or Bayesian belief networks), expectation maximization, genetic algorithms, linear regression algorithms, nonlinear regression algorithms, logistic regressionbased classification models, or an ensemble thereof. An “ensemble” can comprise a collection of machine learning models whose outputs (predictions) are combined, such as by using weighted averaging or voting. The individual machine learning models of an ensemble can differ in their expertise, and the ensemble can operate as a committee of individual machine learning models that is collectively “smarter” than any individual machine learning model of the ensemble.

[0063] In an embodiment, and as noted above, the ML model(s) 134 may identify or determine the characteristic(s) 108, such as the speech and / or vocal biomarkers 110, for comparison against the reference characteristic(s). The ML model(s) 134 may learn to identify complex relationships between the speech and / or vocal biomarkers 110, user data, etc. for determining whether the user speech 106 is associated with atypical speech and / or behavior. For example, the ML model(s) 134 may learn to associate certain speech and / or vocal biomarkers 110 with whether the user speech 106 includes or is free from atypical speech and / or behavior.

[0064] In an embodiment, the ML model(s) 134 may learn to predict that the user speech 106 contains atypical speech and / or behavior by attributing corresponding score(s) to the user speech 106. In this manner, user speech 106 with low scores, for example, (e.g., below threshold) may not include atypical speech and / or behavior, or the speech and / or vocal biomarkers 110 may not be indicative of atypical speech and / or behavior, while the user speech 106 with high scores (e.g., above threshold) may hypophonia. Although the use of a threshold is described as one example way of providing labeling (i.e., replaced or maintained), other techniques are contemplated, such as clustering algorithms, or other statistical approaches that use the trust scores may be used. The ML model(s) 134 is / are retrainable with new data to adapt the ML model(s) 134.

[0065] In addition to processing the audio data 104 and / or the sensor data 124 to detect atypical speech and / or behavior, the wearable speech therapy device 102 may utilize other data. For example, the wearable speech therapy device 102 may have access to user data stored within the user profile 132. The user data may indicate demographics of the user 100, environmental factors, a schedule of the user 100, medications of the user 100, etc. This user data may additionally, or alternatively be used to treat, detect, manage, and / or prevent atypical speech and / or behavior. For example, incorporating user-specific, treatment-related, non-vocal data may be used to more effectively treat, manage, and / or prevent atypical speech and / or behavior. In an embodiment, the speech and / or vocal biomarkers 110 may be determined via a correlation with the user data.

[0066] The speech and / or vocal biomarkers 110 may be stored in association with the user profile 132. The user profile 132 may also store the audio data 104. Such data may be used by the user 100, care provider, and / or a speech pathologist to further refine therapeutic benefits made possible by the wearable speech therapy device 102. The audio data 104 and / or sensor data 124 may also be stored in the memory 114 of the wearable speech therapy device 102 and provided (e.g., downloaded) to other devices.

[0067] As shown, the wearable speech therapy device 102 may be communicatively coupled to one or more devices, such as a mobile device 136 and / or remote computing resource(s) 138 over network(s) 140. The network(s) 140 may include any viable communication technology, such as wired and / or wireless modalities and / or technologies. The network(s) 140 may include any combination of Personal Area Networks (PANs), Local Area Networks (LANs), Campus Area Networks (CANs), Metropolitan Area Networks (MANs), extranets, intranets, the Internet, short-range wireless communication networks (e.g., ZigBee, Bluetooth, etc.) Wide Area Networks (WANs) - both centralized and / or distributed - and / or any combination, permutation, and / or aggregation thereof. The wearable speech therapy device 102 includes suitable network interface(s) 142 for communicating over the network(s) 140.

[0068] In an embodiment, the mobile device 136 may process the audio data 104 and / or the sensor data 124 to determine the speech and / or vocal biomarkers 110 for detecting atypical speech and / or behavior. The wearable speech therapy device 102 may send the audio data 104 and / or the sensor data 124 to the mobile device 136 for processing, and the mobile device 136 may transmit notifications back to the wearable speech therapy device 102 for output. Microphone(s) and / or sensor(s) of the mobile device 136 may additionally or alternatively be used to detect atypical speech and / or behavior. Additionally, the wearable speech therapy device 102 and / or the mobile device 136 may communicatively couple to the remote computing resource(s) 138 for processing the audio data 104 and / or the sensor data 124. Any levelof split processing may be performed by the wearable speech therapy device 102, the mobile device 136, the remote computing resource(s) 138, and / or other devices, systems, networks, etc.

[0069] Although described herein as being a wearable device worn on the user 100, in an embodiment, the techniques described herein for treating, detecting, managing, and / or preventing hypophonia may not be limited to such wearable devices. For example, non-wearable devices may employ the techniques described herein.

[0070] The remote computing resource(s) 138 may be implemented as one or more servers and may, in an embodiment, form a portion of a network-accessible computing platform implemented as a computing infrastructure of processors, storage, software, data access, etc. that is maintained and accessible via a network such as the Internet. The remote computing resource(s) 138 does not require end-user knowledge of the physical location and configuration of the system that delivers the services. Common expressions associated with the remote computing resource(s) 138 may include “on-demand computing”, “software as a service (SaaS)”, “platform computing”, “network-accessible platform”, “cloud services”, “data centers”, etc.

[0071] As used herein, a processor, such as the processor(s) 112 may include multiple processors and / or a processor having multiple cores. Further, the processor(s) 112 may comprise one or more cores of different types. For example, the processor(s) 112 may include application processor units, graphic processing units, and so forth. In one implementation, the processor(s) 112 may comprise a microcontroller and / or a microprocessor. The processor(s) 112 may include a graphics processing unit (GPU), a microprocessor, a digital signal processor or other processing units or components known in the art. Alternatively, or in addition, the functionally described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that may be used include field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system- on-a-chip systems (SOCs), complex programmable logic devices (CPLDs), etc. Additionally, each of the processor(s) 112 may possess its own local memory, which also may store program components, program data, and / or one or more operating systems.

[0072] Memory, such as the memory 114 may include volatile and nonvolatile memory, removable and non-removable media implemented in any method or technology for storage of information, such as computer-readable instructions, data structures, program component, or other data. Such memory may include but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, RAID storage systems, or any other medium which can be used to store the desired information and which can be accessed by a computing device. The memory may be implemented as computer-readable storage media (“CRSM”), which may be any available physical media accessible by the processor(s) to execute instructions stored on the memory. In one basic implementation, CRSM may include random access memory (“RAM”) and Flash memory. In other implementations, CRSM may include but is not limited to, read-only memory (“ROM”), electrically erasable programmable read-only memory (“EEPROM”), or any other tangible medium that can be used to store the desired information and which can be accessed by the processor(s) 112. The memory 114 is an example of non-transitory computer-readable media. The memory 114 may store an operating system and one or more software applications, instructions, programs, and / or data to implement the methods described herein and the functions attributed to the various systems.

[0073] FIG. 2A and 2B illustrate various views of the wearable speech therapy device 102, according to examples of the present disclosure. In an embodiment, the wearable speech therapy device 102 may be a compact device worn on a body of the user 100. For example, the wearable speech therapy device 102 may be worn in any manner on the user 100 such as in the form of bracelets, watches, stick-on arrays, eyeglasses, caps, hearing aids, brooches, ties, lanyards, earrings, etc. The wearable speech therapy device 102 may include suitable attachment mechanisms, such as hook and loop, clasps, magnets, cords, chains, etc. for coupling (e.g., pinning, hanging, etc.) in any manner to the user 100. As illustrated and as discussed in FIG. 1, the wearable speech therapy device 102 may be worn around a neck of the user 100 (e.g., via a lanyard, for example).

[0074] The wearable speech therapy device 102 may include a housing 200 in which components of the wearable speech therapy device 102 are disposed. The housing 200 may include any shape, such as an elongated cylindrical housing. The housing 200 may be manufactured from any suitable materials, such as plastics, composites, metal, etc., and using any manufacturing processes, such as injection molding, stamping, blow molding, etc. The housing 200 may also be water-resistant to prevent the ingress of contaminates into the housing. Internal components are disposed within the housing 200, such as the microphone(s) 118, sensor(s) 116, processor(s) 112, rigid or flexible PCBs that may be stand-alone or embedded or integrated, battery, etc.

[0075] The housing 200 may include a first end 202 and a second end 204 spaced apart from the first end 202 (e.g., in the Z-direction). The first end 202, when worn by the user 100, may be disposed closer tothe mouth of the user 100 as compared to the second end 204. For example, the housing 200 may hang from the first end 202, via a lanyard (not shown), coupling to an attachment mechanism 216 (e.g., clasp, hoop, etc.). When worn, the housing 200 (or more generally the wearable speech therapy device 102) may be vertically disposed relative to an anatomical axis of the user 100. The microphone(s) 118 and the mouth of the user 100 may be positioned along approximately a straight line, with one of the microphone(s) 118 being located nearer to the mouth than the other. The first of the microphone(s) 118 may be located closer to the first end 202 as compared to the second of the microphone(s) 118, while the second of the microphone(s) 118 may be located closer to the second end 204 as compared to the first end 202. The microphone(s) 118 may be located along a central axis of the wearable speech therapy device 102 (e.g., between the first end 202 and the second end 204).

[0076] Sound may be routed to the microphone(s) 118, respectively, via port(s) 206 disposed through the housing 200. The port(s) 206 may route sound to other sensor(s) 116. For example, the housing 200 may include a front surface 208 that defines the port(s) 206. A first of the port(s) 206 may channel sound to the first of the microphone(s) 118, while a second of the port(s) 206 may channel sound to the second of the microphone(s) 118. Additionally, the port(s) 206 may also be sealed or covered with an acoustic mesh or membrane material that prevents or substantially prevents the ingress of water or moisture into the interior of the wearable speech therapy device 102, while allowing sound to permeate there through and reach the microphone(s) and / or sensor(s) 116. For example, in an embodiment, the mesh or membrane material may include polytetrafluoroethylene (PTFE), silicone rubber, metal, and / or a combination thereof. Although the port(s) 206 are described and shown as being defined by the front surface 208 of the housing 200, the port(s) 206 may be disposed on other sides, surface(s), etc. of the housing 200.

[0077] When worn by the user 100, a back surface 210 of the housing 200 may be disposed against the body of the user 100. As the back surface 210 may be in contact (e.g., rest against, abut, etc.) the body of the user 100, the sensor(s) 116 may be arranged to detect the user speech 106, for example, based on experienced vibrations, accelerations, inhaling / exhaling, pressure changes, and so forth. Additionally, although a particular disposition of the wearable speech therapy device 102 is described when worn by the user 100, the wearable speech therapy device 102 may be horizontally or vertically disposed on the user 100.

[0078] The wearable speech therapy device 102, as discussed above, may include I / O component(s) 130. For example, the wearable speech therapy device 102 may include lighting element(s) 212 that output light. The lighting element(s) 212 may output the notification(s) 128 associated with providing feedbackto the user 100. The lighting element(s) 212 may output any combination of light, color, intensity, etc. As shown, the wearable speech therapy device 102 may include three of the lighting element(s) 212, however, more than or less than three of the lighting element(s) 212 may be included. Additionally, although the lighting element(s) 212 are shown being located on the front surface 208, the lighting element(s) 212 may be located elsewhere. Additionally, the lighting element(s) 212 may be implemented as a LED strip, or LED bar, as compared to being separate lighting elements disposed through the housing 200.

[0079] The housing 200 may also define one or more orifices 214 through which sound is emitted via a speaker of the wearable speech therapy device 102. The one or more orifices 214 may be located along one or more sides of the housing 200. The speaker may output audio corresponding to the notification(s) 128 for providing feedback to the user 100. Mesh or other material may cover the one or more orifices 214 to allow sound to exit the wearable speech therapy device while at the same time, inhibiting the ingress of liquids, moisture, or debris.

[0080] In an embodiment, one or more button(s) may be included to control one or more operation(s) of the wearable speech therapy device 102. The button(s) may be used to power on / off the wearable speech therapy device 102, mute the wearable speech therapy device 102, change setting(s) of the wearable speech therapy device 102, connect to nearby devices, etc.

[0081] FIG. 3 illustrates an example process 300 (e.g., method) related to determining whether the user speech 106 is associated with atypical speech and / or behavior, according to examples of the present disclosure. The process 300 described herein is illustrated as collections of blocks in logical flow diagrams, which represent a sequence of operations or acts, some or all of which may be implemented in hardware, software, or a combination thereof. In the context of software, the blocks may represent computer-executable instructions stored on one or more computer-readable media that, when executed by one or more processors, program the processors to perform the recited operations or acts. Generally, computer-executable instructions include routines, programs, objects, components, data structures and the like that perform particular functions or implement particular data types. The order in which the blocks are described should not be construed as a limitation, unless specifically noted. Any number of the described blocks may be combined in any order and / or in parallel to implement the process 300 or alternative processes, and not all of the blocks need be executed. For discussion purposes, the process 300 is described with reference to the environments, devices, architectures, diagrams, and systems described in the examples herein, such as, for example those described with respect to FIGS. 1-2B,although the process 300 may be implemented in a wide variety of other environments, architectures, and systems.

[0082] At 302, the process 300 may include receiving audio data from one or more microphone(s) of a first device. For example, the wearable speech therapy device 102 may receive the audio data 104 as generated from the microphone(s) 118. In an embodiment, the audio data 104 may be at least partially processed or unprocessed (e.g., filtered). In an embodiment, rather than receiving audio data 104, the wearable speech therapy device 102 may receive audio signals.

[0083] At 304, the process 300 may include receiving sensor data from one or more sensor(s) of the first device. For example, the wearable speech therapy device 102 may receive the sensor data 124 as generated from the sensor(s) 116. In an embodiment, the sensor data 124 may be at least partially processed or unprocessed (e.g., filtered). In an embodiment, rather than receiving sensor data 124, the wearable speech therapy device 102 may receive sensor signals from the sensor(s) 116. The sensor(s) may include accelerometer(s), piezoelectric sensor(s), IMUs, etc.

[0084] At 306, the process 300 may include determining whether the audio data and / or the sensor data is indicative of user speech or relevant behavior. For example, while various sounds may be present within an environment, the audio data 104 and / or the sensor data 124 may be processed to determine when the sounds correspond to the user speech 106. As discussed herein, the wearable speech therapy device 102 may be configured to identify the user speech 106 to provide therapeutic feedback to the user 100. Accordingly, to provide the therapeutic feedback, the user speechl06 is first detected.

[0085] Any combination of the audio data 104 and / or the sensor data 124 may be used to determine whether the user is speaking. For example, the sensor data 124 may indicate vibrations, accelerations, inhaling / exhaling, pressure changes, and so forth. In an embodiment, the audio data 104 may be processed using VAD, TOA, etc. to determine whether the user 100 is speaking. If at 306 the process 300 determines that the user speech 106 was not detected, the process 300 may follow the “NO” route and loop to 302. Here, the wearable speech therapy device 102 may continue to receive the audio data 104 and / or the sensor data 124 for determining whether the user speech 106 was detected. Comparatively, if at 306 the process 300 determines that the user speech 106 was detected, the process 300 may follow the “YES” route and proceed to 308.

[0086] At 308, the process 300 may include determining speech and / or vocal biomarker(s) associated with the user speech. For example, if the user speech 106 was detected, the process 300 may include determining the speech and / or vocal biomarker(s). In an embodiment, the speech and / or vocalbiomarker(s) may be associated with determining whether the user speech is associated with atypical speech and / or behavior. Any number of the speech and / or vocal biomarkers 110 may be determined. As non-limiting examples, the speech and / or vocal biomarkers 110 may include pitch of the user speech 106, intonations in the user speech 106, tones associated with the user speech 106, pauses in the user speech 106, phonations associated with the user speech 106, changes associated with the user speech 106, articulations associated with the user speech 106, decreased energy in the higher parts of a harmonic spectrum of the user speech 106, imprecise articulation of vowels and consonants of the user speech 106, fundamental frequency of the user speech 106, voicing, windowed and absolute syllable / sonorant-peak rates, SNR, temporal and spectral voice characteristics, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, and / or motor processes. In an embodiment, the speech and / or vocal biomarkers 110 may be associated with detecting low-amplitude speech or atypical speech and / or behavior.

[0087] In an embodiment, ML model(s) 134 may be used to process the audio data 104 and / or the sensor data 124 for determining the speech and / or vocal biomarkers 110. For example, the ML model(s) 134 may be previously trained to determine the speech and / or vocal biomarkers 110. In an embodiment, the ML model(s) 134 may receive the audio data 104 and / or the sensor data 124 as an input, and output an indication associated with whether the audio data 104 and / or the sensor data 124 are indicative of the speech and / or vocal biomarkers 110. In an embodiment, the ML model(s) 134 may output a score associated with the speech and / or vocal biomarkers 110, where the score may indicate likelihood or probability of the user speech containing the speech and / or vocal biomarkers 110.

[0088] At 310, the process 300 may include determining whether the speech and / or vocal biomarker(s) satisfy threshold(s). For example, the speech and / or vocal biomarkers 110 may be compared to reference speech and / or vocal biomarker(s), or character! stic(s) that are indicative of atypical speech and / or behavior. Any number of the speech and / or vocal biomarkers 110 may be considered when determining whether the user speech 106 contains atypical speech and / or behavior. In an embodiment, the ML model(s) may be used for determining whether the speech and / or vocal biomarkers 110 are indicative of satisfying threshold(s) associated with the user speech 106 containing atypical speech and / or behavior.

[0089] As non-limiting examples, the speech and / or vocal biomarkers 110 may be associated a pitch, amplitude, and / or tone associated with the user speech 106, and the pitch, amplitude, and / or tone may be compared to respective threshold(s). Based on such comparison, for example, the process 300 may determine whether the pitch, amplitude, tone, etc. are indicative of atypical speech and / or behavior. If at310 the process 300 determines that the speech and / or vocal biomarkers 110 do not satisfy the threshold(s), the process 300 may follow the “NO” route and proceed to 312.

[0090] At 312, the process 300 may include causing output of one or more first notification(s). For example, the I / O component(s) 130 of the wearable speech therapy device 102 may output notification(s) 128, where the notification(s) 128 may indicate that the user speech 106 does not contain atypical speech and / or behavior. For example, the lighting element(s) 212 may illuminate green, the speakers) may output audio tones, and so forth. In an embodiment, the first notification(s) may indicate that the user 100 is speaking properly, for example, as a way to provide feedback to the user 100 that their speech is correct and does not include indications of atypical speech and / or behavior. From 312, the process 300 may proceed to 302, whereby the process 300 may continue to receive the audio data 104 and / or the sensor data 124 for determining whether the user speech 106 is indicative of atypical speech and / or behavior.

[0091] Returning to 310, if the process 300 determines that the speech and / or vocal biomarker(s) satisfy the threshold(s), the process 300 may follow the “YES” route and proceed to 314. At 314, the process 300 may include causing output of one or more second notification(s). For example, the VO component(s) 130 of the wearable speech therapy device 102 may output notification(s) 128, where the notification(s) 128 may indicate that the user speech 106 contains atypical speech and / or behavior. For example, the lighting element(s) 212 may illuminate red, the speaker(s) may output audio tones, and so forth. In an embodiment, the second notification(s) may be different than the first notification(s) as a way to signal to the user 100 to correct their speech. In this manner, the second notification(s) may provide feedback to the user 100 to enable the user 100 to correct their speech.

[0092] At 316, the process 300 may include sending at least a portion of the audio data and / or the sensor data to a second device. For example, the wearable speech therapy device 102 may send the audio data 104 and / or the sensor data 124 and / or the sensor data 124 to the mobile device 136 and / or the remote computing resource(s) 138. In an embodiment, the mobile device 136 and / or the remote computing resource(s) 138 may further process the audio data 104 and / or the sensor data 124, may store / log the audio data 104 and / or the sensor data 124 to maintain a history of the user speech 106 associated with the user 100, etc. From 316, the process 300 may loop to 302, whereby the wearable speech therapy device 102 may continue to process and analyze the user speech 106.Illustrative Embodiment of Second Device and Methods Thereof

[0093] This application is directed, at least in part, to monitoring animal welfare for determining one or more pathological, physiological, anatomical, behavioral, physical, or locative states of the animal, according to examples of the present disclosure. In an embodiment, sensor data generated from one or more sensors may be used to determine animal welfare for detecting behavioral and / or pathological states, such as animal disease, distress, etc. to provide care, treatment, and / or management. For example, the sensor data may be used to determine laryngeal / vocal biomarkers, physical biomarkers, environmental biomarkers, and / or other audio markers associated with the animal or an environment of the animal. The markers may be used to determine the welfare of the animal for deducing pathological states of the animal, for example, by comparing the biomarkers to one or more reference markers. The animal welfare, behavioral, and / or pathological states may be determined in real-time and / or over a period of time (e.g., day, month, year, etc.). In an embodiment, based at least in part on the welfare, behavioral, and / or the pathological states, a caregiver, owner, etc. associated with the animal may be notified to provide care, treatment, etc.

[0094] In an embodiment, the techniques described herein may be implemented within, or at least in part, by a device. In an embodiment, the device may be considered a veterinary acoustic and biofeedback device. For example, as will be discussed herein, the veterinary acoustic and biofeedback device may, in an embodiment, generate or utilize sensor data associated with the acoustics of the animal, animal behavior, physiological state, or an environment of the animal to provide feedback to the animal, a caregiver (e.g., owner, veterinarian, etc.) of the animal, and so forth. However, as will be discussed herein, although mentioned as an acoustic device, the veterinary acoustic and biofeedback device may generate or utilize other non-acoustic data (e.g., including biological assessment data, visual data, behavioral data, geolocation, environmental data, etc.) associated with the animal to determine the welfare, behavioral, and pathological states.

[0095] The veterinary acoustic and biofeedback device may be a compact, wearable, or peripheral device. For example, the veterinary acoustic and biofeedback device may be worn, attached, mounted, etc., to the animal. In an embodiment, the veterinary acoustic and biofeedback device may be small, non-invasive, non-intrusive, lightweight, and discreet in nature. In an embodiment, the veterinary acoustic and biofeedback device may be embodied within a harness, collar, stick-on arrays, saddle, tag, etc., that is worn by the animal. Additionally, or alternatively, the veterinary acoustic and biofeedback device may be attached or worn by the animal using attachment mechanisms, such as hook and loop, clasps, magnets, cords, chains, clips, belts and buckles, cords, pins, adhesives, etc. for coupling (e.g., pinning, hanging,etc.) in any manner to the animal. Alternatively, the veterinary acoustic and biofeedback device may be sub-dermally implanted, injected, or ingested.

[0096] The veterinary acoustic and biofeedback device, however, may also and / or additionally be separate from the animal (i.e., not worn by the animal), such as placed in proximity to the animal. For example, animals may be housed in barns, stables, beds, pens, cages, etc. Here, the veterinary acoustic and biofeedback device (or a component thereof) may be positioned relative to the animal or group of animals or a group of animals to monitor the animal and / or the environment of the animal. In an embodiment, the veterinary acoustic and biofeedback device may be mounted on a wall or door of a pen, for example, attached to a fixture or piece of equipment within an animal enclosure, attached to fence(s) within a pasture, and so forth.

[0097] The veterinary acoustic and biofeedback device may include a housing in which components of the veterinary acoustic and biofeedback device are disposed. The housing may include any shape, such as an elongated cylindrical housing. The housing may be manufactured from any suitable materials, such as plastics, composites, metal, etc., and using any manufacturing process(es), such as injection molding, stamping, blow molding, etc. The housing may also be contaminate-resistant to prevent the ingress of liquid, dust, or contaminants into the housing. As introduced above, the housing may be configured to be worn by the animal, for example, through the use of a harness (e.g., collar), or may be separate from the animal.

[0098] As will be discussed herein, the veterinary acoustic and biofeedback device may include suitable processing and internal components to detect the animal welfare and behavior, as well as to treat, manage, and / or prevent pathological and / or physiological states. For example, the veterinary acoustic and biofeedback device may include one or more sensor(s), microphone(s), battery(ies), processing components (e.g., processors, memory, system on a chip (SoC) or integrated circuits (IC), etc.), etc. disposed within the housing. In an embodiment, the veterinary acoustic and biofeedback device may include VO components, such as a display for outputting information associated with the animal. The display may include an organic light emitting diode (O / LED), (in-plane switching (IPS) or thin-film- transistor (TFT)) liquid crystal display (LCD), laser video display (LVD), vibrotactile, haptic, or other applications. In an embodiment, the device (or a component of a system) may even include a camera whereby still and or video image data may be collected for confirmation or collaboration in connection with the audio and / or other data gathered.

[0099] In an embodiment, a plurality of veterinary acoustic and biofeedback devices may be used in conjunction with one another, such as in a system, to determine the animal welfare and / or pathological, biological, and / or behavioral states. For example, a first veterinary acoustic and biofeedback device (i.e., a first system component) worn on the animal may be used in conjunction with a second veterinary acoustic and biofeedback device (i.e., a second system component) placed within a pasture. In such instances, the veterinary acoustic and biofeedback devices may be the same or different, for example, in size, functionality, capabilities, etc. For example, the first veterinary acoustic and biofeedback device may include first sensors that are different from the second sensors of the second veterinary acoustic and biofeedback device. In instances in which a plurality of the veterinary acoustic and biofeedback devices are used, the veterinary acoustic and biofeedback device(s) may be communicatively coupled to one another via one or more network interface(s) (e.g., Bluetooth, Cellular, etc.). Any combination of veterinary acoustic and biofeedback device(s) worn on the animal and / or off the animal may be used, or by different animals may be used.

[0100] The veterinary acoustic and biofeedback device may be used in association with any type of animal. For example, the animals may be any suitable animal, such as livestock (e.g., cattle, sheep, pigs, etc.), pets (e.g., horses, dogs, cats, parrots, etc.), wildlife subjects (e.g., wolves, whales, dolphins, waterfowl, birds, etc.), and so forth. In an embodiment, the veterinary acoustic and biofeedback device may be designed or preconfigured for the animal. For example, a size, type, material, etc. of a harness may be dependent upon the animal. In addition, the placement of the veterinary acoustic and biofeedback device on the animal may be dependent upon the type of animal. For example, to detect welfare, behavioral, and / or pathological states, the veterinary acoustic and biofeedback device may be placed at predetermined positions on the animal. For example, in bovine and ovine applications, the large animals tolerate a collar-type device, in which case acoustic sensing of laryngeal / vocal activity is conveniently proximal to the animal’s larynx and airway physiology. Other applications, such as for poultry or waterfowl, may require leg-mounted devices with special attention to the weight of the overall device relative to animal size and proportion. In another application, penned or cloistered animals may be monitored with devices mounted to the penning structure. Free-range or grazing / roaming animals may require exclusively body-worn devices. These necessitated physical mounting demands may require adaptations to device characteristics such as power management, number of sensors, orientation of device, and so on.

[0101] The veterinary acoustic and biofeedback device may include any number of sensor(s) to monitor the animal and / or the environment of the animal for determining the welfare, behavioral, and / or pathological states. Example sensor(s) include but are not limited to, microphone(s), piezoelectric sensor(s), accelerometer(s), temperature sensor(s), heart rate sensor(s), humidity sensor(s), inertial measurement unit (IMU) sensor(s), GPS sensor(s), blood oxygen sensor(s), Hall sensor(s), passive infrared (PIR) sensor(s), etc. In an embodiment, the sensor(s) may be used to identify activity and / or nonactivities of the animal, as well as to characterize the activity and / or non-activity. For example, the sensor(s) may be used to determine activities of the animal (e.g., vocalizations, respiration, movement, eating, sleeping, etc.) and / or non-activates of the animal (e.g., non-vocalizations, lameness, etc.). Moreover, the sensor(s) may be used to determine vocalizations and / or behavior of the animal for deducing the welfare, behavioral, and / or pathological states.

[0102] As will be discussed herein, the microphone(s) collect audio data, and the sensor(s) collect sensor data that is analyzed to determine the laryngeal / vocal biomarkers, physical biomarkers, behavioral biomarkers, environmental biomarkers, and / or other audio and physiological markers associated with the animal or an environment of the animal. The veterinary acoustic and biofeedback device may include analog and digital converters that assist in the capture, processing, and storage of signals sampled from the microphone(s) and sensor(s), respectively. The signals from the microphone(s) and sensor(s) may be stored in a dataset, including raw and / or normalized or transformed values. The audio signals or sensor signals may be processed as data, which is analyzed to determine the vocal biomarkers, physical biomarkers, behavioral data, environmental biomarkers, and / or other audio markers.

[0103] In an embodiment, the vocal biomarkers may be associated with vocalizations and laryngeal activity (e.g., sound, communications, respiratory activity, etc.) produced by the animals. In an embodiment, the vocalizations may be associated with pain (e.g., groaning, grunting, labored breathing, etc.), distress (e.g., predators in area), social (e.g., signaling isolation / separation of a calf from its mother, mating calls, nursing or mating sounds, etc.). In an embodiment, characteristics of the vocalizations may be determined for use in determining the welfare, behavioral, and / or pathological states. In doing so, the vocalizations may be used to determine a health of the animal for providing treatment and / or responding to the animal.

[0104] The vocalizations may be based at least in part on the type of animal and / or the environment of the veterinary acoustic and biofeedback device. In an embodiment, the vocal biomarkers may be determined based at least in part on characteristic(s) of the audio, such as tone, prosody, duration,amplitude, etc. In an embodiment, the audio data generated by the microphone(s) may be processed to determine the vocal biomarkers. Additionally, or alternatively, sensor data generated by the sensor(s) may be processed to determine the vocal biomarkers. For example, sound captured by the microphones and data indicative of inhaling, exhaling, or vocalizing (e.g., via vocal cord vibration, for example) generated by the sensor(s) may be processed to determine the vocal biomarkers.

[0105] In an embodiment, the physical biomarkers may be associated with physical activity of the animals, such as walking, calving, mating, avoidance behavior, inter-animal interactions, volitional activity including laryngeal vocalizations, searching or avoiding food or water sources, learned behaviors, etc. The environment biomarkers may be associated with weather prediction, seasonal variation, estimation of precipitation, identification of dangerous weather patterns (e.g., electrical storms, flooding, etc.), and existence in the environment of other hazards (e.g., proximity to traffic noise, predator sounds, rival animals fighting, etc.), humidity, presence of other animals, etc. The audio markers may be associated with physiological, anatomical, behavioral, or mental state of the animal. There is growing scientific literature and worldwide concern over the welfare of animals as indicated by their behavior, activity, and interactions. Both physical and mental health of animals are measured as a function of animal behavior both individually and within-group. Using predictors as mentioned herein, objective indicators are collected via sensors in the device, used in the classification / discrimination models, and interpreted as actionable information for the animal directly or for their caretakers. For example, increased lowing frequency and increased amplitude in a post-natal cow may indicate acute separation of cow and calf as perceived by the cow. This indication may be monitored by the device and constitute a cause for alerting a caregiver. Other biomarkers, such as individual or situational biomarkers, may be used to characterize the experiences, activities, behaviors, and actions of the animal. In an embodiment, the vocal biomarkers, physical biomarkers, environmental biomarkers, and / or other audio markers may be generally referred to herein as “animal biomarkers.”

[0106] As introduced above, the veterinary acoustic and biofeedback device may include the microphone(s) for collecting audio signals associated with vocalizations, sounds, etc., generated from the animal or sounds generated from other sources within the environment. In an embodiment, the microphone(s) may include at least two microphones spaced apart from one another within the housing. For example, the housing may include a first end and a second end spaced apart from the first end. A first of the microphones may be located closer to the first end than the second end, while a second of the microphones may be located closer to the second end than the first end. Any number of microphones,however, may be included within the veterinary acoustic and biofeedback device. In an embodiment, the microphone(s) may be used for determining a directionality (e.g., time of flight, velocimetry, time-delay analytics, etc.) of the signal and / or whether the signal corresponds to the sound generated from the animal wearing the veterinary acoustic and biofeedback device (e.g., as compared to background noise, other animals, predators, weather, etc.). For example, the microphone(s) may receive respective audio signals that are representative of sounds within the environment.

[0107] When worn by the animal, the microphones may be spaced apart by different distances from sound sources of the animal (e.g., the mouth of the animal, digestive system, etc.). For example, when the veterinary acoustic and biofeedback device represents a device that attaches as a collar, the microphones may be located at different distances from the mouth of the veterinary acoustic and biofeedback device to discern sound from the animal as compared to other sources of sound within the environment. In an embodiment, the veterinary acoustic and biofeedback device may employ suitable techniques to detect, segregate, label, and classify animal sounds or other vocalizations. For example, audio signatures may be used to correlate sounds of the animal and / or the environment with certain known sounds. Additionally, statistical likelihood pattern-matching models may also be employed using templates of known animal laryngeal vocal activity or by machine comparisons with known, idealized, or structured samples.

[0108] The audio signals generated by the microphones may be cross-correlated. For example, crosscorrelation may apply between audio signals detected by the microphones to determine a time difference of arrival (TDOA) for the correlated signals. In other words, the differently spaced microphones from the mouth of the animal have different time of arrival (TOA) values. This allows the calculation of an angle from which the detected signals originate relative to a line intersecting the two microphones.

[0109] In operation, the sensor(s) of the veterinary acoustic and biofeedback device may collect signals or data that is analyzed to determine the animal biomarkers. For example, a microphone may collect audio signals, or audio data, that is analyzed to determine vocalizations, acoustics, sound, etc. of the animal. Vocalizations of the animal may include not only sounds generated via a mouth of the animal (e.g., moaning, grunting, etc.) but sound generated by digestion, body parts, foot-fall, etc. of the animal. Moreover, the microphones may capture sound or other acoustics of an environment of the animal (e.g., weather, sounds of other animals, etc.).

[0110] In an embodiment, the veterinary acoustic and biofeedback device may be configured to identify sounds emanating or generated from the animal, as compared to sounds generated from other sources within the environment of the animal. For example, the veterinary acoustic and biofeedback device mayutilize the sensor(s) to determine the sounds of the animal, their amplitude, signal-to-noise ratio, and so forth. Sounds may also be associated with the animal, as compared to other sources of sound, based on detecting vocal vibrations by detecting changes in pressure, breathing, acceleration, temperature, strain, force, etc. Times associated with generating the audio data and the sensor data may be associated with one another to correlate the sensor data with the audio data.[OHl] In addition to acoustics, accelerometers may generate sensor signals or sensor data to determine behaviors of the animal, such as movements (e.g., bedding, running, calving, digestion, etc.). Other sensor(s) may be used to classify the behavior of the animal and the environment of the animal. For example, heart rate sensor(s) may be used to determine the health of the animal (e.g., electrocardiogram, photoelectric pulse wave, blood pressure measurement, phonocardiography, etc.), weather sensor(s) may be used to determine weather conditions (e.g., temperature of the animal and / or environment), and so forth. Based at least in part on the audio data and / or the sensor data, the animal biomarker(s) associated with the animal may be determined. Any combination of the audio data and / or the sensor data may be used to determine the animal biomarker(s) for determining the welfare, behavioral, and / or pathological states.

[0112] The animal biomarkers may be compared against reference animal biomarkers or thresholds, for example, to determine the animal welfare. In an embodiment, the pathological and / or behavioral states of the animal may be determined based at least in part on the animal welfare. In an embodiment, the reference biomarkers may be considered diagnostic / classificatory biomarkers in that the animal biomarkers are compared to the reference biomarkers to determine welfare, behavioral, and / or pathological details. As an example, poultry have been observed to express up to thirty distinct vocalizations corresponding to different pathological states. Capturing sounds of the animal allows for the determination of whether the poultry, for example, includes a pathological state that warrants care, treatment, etc. As such, sounds (e.g., pathological sounds) emitted from the animal may be detected and analyzed to identify and inform a caregiver regarding the pathological state of the animal, whether normal, suspect, or determined pathological or distress experienced by the animal. In an embodiment, these different vocalizations may have, or be characterized by, reference audio signatures. As will be discussed herein, audio signatures of the vocalizations of the animal may be compared against reference audio signatures to determine the welfare, behavioral, and / or pathological state of the animal.

[0113] Depending upon the type of animal, comparable sets of reference biomarkers may be used for determining the animal welfare, biological, behavioral, and / or pathological state. For example, theveterinary acoustic and biofeedback device may have reference biomarkers for cattle, dogs, birds, etc. for comparison to the animal biomarker(s). The comparison between the animal biomarkers and the reference animal biomarkers or thresholds may determine a statistical likelihood of pathological states of the animal. Moreover, the veterinary acoustic and biofeedback device may be preconfigured to monitor for certain types of welfare, behavioral, and / or pathological states. For example, the veterinary acoustic and biofeedback device may be preconfigured to monitor for fowl cholera within poultry, as compared to other diseases (e.g., salmonellosis). For example, bovine respiratory disease (BRD) or porcine respiratory disease complex (PRDC) may have specific laryngeal, vocal, and respiratory signatures and identifiable biomarkers identified by the device. In an embodiment, the veterinary acoustic and biofeedback device may be configured to monitor for any number of pathological states.

[0114] In an embodiment, the animal welfare may be associated with healthy, sick, diseased, etc. In an embodiment, animal welfare may be used to classify or determine the pathological states of the animal. Pathological states may be related to, but are not limited to, respiration, feeding, water intake, movement, digestive, gastrointestinal, mating or calving, disease, metabolism, gate (e.g., lameness, mastitis, metritis, ketosis, etc.), and so forth. These pathological states may be correspondingly utilized to identify, characterize, diagnose, and treat, etc. diseases (e.g., states, factors, etc ), disabilities, distresses, welfare disturbances, pathologies, activities, congenital defects, effects of medications, environmental effects, accidents, harmful physical events, etc. associated with the animal.

[0115] In an embodiment, the veterinary acoustic and biofeedback device may generate audio fingerprint(s) of the animals based at least in part on the animal biomarkers. The audio fingerprint(s) may be individual to the animals and permit the animals to be distinguished from one another, for example, to know which animal amongst a group of animals produced a sound or contributed data to the device’s sensor array. This allows the sound to be associated with a certain animal to accurately determine the welfare, behavioral, and / or pathological state of a particular animal.

[0116] The vocal fingerprint or phenotype may also improve the accuracy of determining welfare, behavioral, and / or pathological states of the animal. For example, in an embodiment, the animal biomarkers may be used to generate audio fingerprint(s) associated with the animal, and the audio fingerprint(s) may be analyzed to determine whether the audio fingerprint(s) is associated with the particular behavioral and / or pathological state. In other words, the audio fingerprint(s) may characterize the sounds of the animal, and the audio fingerprint(s), for example, may be compared to one or more reference fingerprints to determine whether the animal is calving, running, ill with disease, having boweldischarge, etc., based on a comparison to reference fingerprints associated with the welfare, behavioral, and / or pathological states. In an embodiment, the use of the audio fingerprint(s) may discern sound generated from the animal from other sounds emanating within the environment.

[0117] In addition to audio fingerprint(s) of the animal, other audio and acoustic fingerprint / phenotype biomarkers may be used in the veterinary acoustic and biofeedback device. For example, characteristic environmental auditory events such as stampeding, weather (e.g., lighting), predators (e.g., howling), etc., have unique acoustic signatures that may be used in combination or solely to better interpret the auditory and acoustic environment of the veterinary acoustic and biofeedback device.

[0118] In an embodiment, the types of pathological states that are detected may be based at least in part on the characteristic(s) of the animal. For example, the animal biomarkers may be based at least in part on other characteristic(s) of the animal, such as age, sex, breed, species, etc. For cattle, diseases such as BVD, BVDV, and PRDC, among others, may be checked. In addition, a history of the animal may be accessed when determining or to determine the animal welfare, behavioral, and / or pathological states. For example, a disease state, audiological status, medical history, medications, etc. may be used to determine the behavior and / or pathological states of the animal.

[0119] In an embodiment, the veterinary acoustic and biofeedback device may output the notification(s) on devices, or transmit the notification(s) to other devices associated with the welfare, behavioral, and / or pathological states. In an embodiment, the notification(s) may be output in real-time or substantially realtime. In an embodiment, the notification(s) may be audible, visual, olfactory, electronic, haptic, and / or any combination thereof. For example, the veterinary acoustic and biofeedback device may include a speaker that outputs audio (e.g., deeps, tones, instructions, etc.), lighting elements that illuminate (e.g., patterns, colors, etc.), and / or a motor that vibrates, etc. By outputting the notification(s), the veterinary acoustic and biofeedback device may provide real-time alerts to the caregiver of the animal. For example, a caregiver tending to their animals may be notified as to the animal welfare, behavioral, and / or pathological states. In an embodiment, the notification(s) may be output when the welfare, behavioral, and / or pathological states are concerning, troublesome, etc. The caregiver, for example, may determine when, or how, the notification(s) are output. For example, the caregiver may specify that based on the welfare, behavioral, and / or pathological states, when the notifications are to be output. The veterinary acoustic and biofeedback device may output the notification(s) for a predetermined amount of time, continuously, and / or until the caregiver responds, aids, etc., the animal.

[0120] In addition to real-time feedback as described above, asynchronous reports may also be delivered to the caregiver and stored in the software. For example, a daily histography digest of performance and behavior may be delivered via email or software (e.g., a mobile application) to the caregiver for offline use. Performance data may also be delivered to a database collection of performance characteristics constituting the basis for additional software and algorithm changes. The application and / or software may be embedded into the hardware / device or may be external software capable of being instantiated into external hardware such as a mobile phone, tablet, or computer.

[0121] The veterinary acoustic and biofeedback device may continuously generate or receive signals from the microphone(s) and / or sensor(s) for determining the welfare, behavioral, and / or pathological states. This allows for the determination of continual and present behavioral and / or pathological states for use in determining whether the behavioral and / or pathological states are ongoing or have been resolved. In an embodiment, the veterinary acoustic and biofeedback device may determine the behavioral and / or pathological states in real-time or may determine the veterinary acoustic and biofeedback device over a period of time.

[0122] In an embodiment, the veterinary acoustic and biofeedback device may utilize artificial intelligence (Al) and machine learning (ML) models to improve the accuracy of determining the animal welfare, behavioral, and / or pathological states. To clarify, Al and ML techniques, which are described and detailed with examples below, are used as encompassing cover terms for “smart” and algorithmic collection, processing, analyzing, databasing, managing, summarizing, and using data in an automated and large-scale fashion. The ML model(s) may be trained from a database (e.g., historical data, such as audio data that includes respiration, feeding, water intake, movement, digestive, gastrointestinal, mating or calving, disease, metabolism, gait (e.g., lameness, mastitis, metritis, ketosis, etc.), disabilities, distresses, welfare disturbances, pathologies, activities, congenital defects, effects of medications, environmental effects, accidents, harmful physical events, etc. of the animal) to analyze the data captured by the sensor(s) and microphone(s) for identifying the animal biomarkers, determining the animal welfare, behavioral, and / or pathological states. The ML model(s) may determine the animal biomarkers and may assess the animal biomarkers in comparison to the information stored in the database (e.g., reference characteristics, reference vocal fingerprint, etc.) to determine the welfare, behavioral, and / or pathological states.

[0123] As part of the ML model(s) analyzing the audio data and / or the sensor data, the ML model(s) may label the character! stic(s) of the vocalizations and / or behavior of the animal, for example, the animalbiomarkers, to indicate whether the characteristic(s) are associated certain animal welfare and / or pathological states. In an embodiment, an output of the ML model(s) may indicate whether the vocalizations and / or behavior are associated with a particular animal welfare and / or pathological state (e.g., healthy, diseased, calving, injured, etc.). In an embodiment, the ML model(s) may determine the characteristic(s) for comparison to respective references for determining whether the vocalizations and / or behavior are indicative of, or are not indicative of, a particular welfare, behavioral, and / or pathological state. The ML model(s) may use any number of characterise c(s) for determining whether the vocalizations and / or behavior are indicative of, or are not indicative of, the particular welfare, behavioral, and / or pathological state.

[0124] In an embodiment, the ML model(s) may generate scores that are associated with the characterise cl's) of the welfare, behavioral, and / or pathological states. For example, the ML model(s) may generate scores associated with the animal biomarkers, animal welfare, behavioral, and / or pathological states. If the scores are greater than a predetermined threshold (which varies depending on the particular data and desired information relayed in return), the ML model(s) may have a threshold confidence that the characteristics are associated with a particular pathological state of the animal, for example. In an embodiment, the ML model(s) may compare an audio fingerprint(s) of the animal with a reference audio fingerprint. A pathological state may be assessed with regard to the model parameters, and a discrete or probabilistic classification will be generated. The classification may be stored, processed, and used for feedback (as needed).

[0125] Although the veterinary acoustic and biofeedback device is described as processing, analyzing, and / or determining the animal biomarkers, welfare, , behavioral, and / or pathological states, etc., in an embodiment, the veterinary acoustic and biofeedback device may communicatively couple to one or more computing devices, resource(s), etc. which may additionally or alternatively process, analyze, and / or determine the animal biomarkers, animal welfare, pathological states, etc. For example, the veterinary acoustic and biofeedback device may receive sensor data that is transmitted to remote computing resource(s) for analysis. As introduced above, the remote computing resource(s) may receive the sensor data from any number of the veterinary acoustic and biofeedback device(s), separate device(s) (e.g., weather and / or temperature device), etc.). In an embodiment, the remote computing resource(s) may analyze the sensor data, among other data, to determine the animal biomarkers, welfare, behavioral, and / or pathological states, etc. Thereafter, the remote computing resource(s) may transmit notification(s) to the veterinary acoustic and biofeedback device, the device of the caregiver, and so forth. Any level of splitprocessing may be performed by the veterinary acoustic and biofeedback device, the device of the caregiver, the remote computing resource(s), and / or other devices, systems, networks, etc.

[0126] The present disclosure provides an overall understanding of the principles of the structure, function, device, and system disclosed herein. One or more examples of the present disclosure are illustrated in the accompanying drawings. Those of ordinary skill in the art will understand and appreciate that the devices, the systems, and / or the methods specifically described herein and illustrated in the accompanying drawings are non-limiting embodiments. The features illustrated or described in connection with one embodiment, or instance, may be combined with the features of other embodiments or instances. Such modifications and variations are intended to be included within the scope of the disclosure and appended claims.

[0127] FIG. 4 illustrates an example environment 400 including an animal 402 wearing a veterinary acoustic and biofeedback device 404, according to examples of the present disclosure. Details of the veterinary acoustic and biofeedback device 404 are discussed herein, however, generally, the veterinary acoustic and biofeedback device 404 may determine biomarkers 406 of the animal 402 for detecting a welfare, behavioral, and / or pathological states 408 of the animal 402. Detection of the welfare, behavioral, and / or pathological states 408 allows for a caregiver 410 (e.g., owner, veterinarian, rancher, etc.) to treat and / or manage the animal 402.

[0128] The veterinary acoustic and biofeedback device 404 may be a compact, wearable, or peripheral device. In an embodiment, the veterinary acoustic and biofeedback device 404 may be worn on a body of the animal 402. In an embodiment, the veterinary acoustic and biofeedback device may be small, non- invasive, non-intrusive, lightweight, and discreet in nature. As shown in FIG. 4, the veterinary acoustic and biofeedback device 404 may be worn as a collar around a neck of the animal 402. However, in an embodiment, the veterinary acoustic and biofeedback device 404 may be embodied within a harness, stick- on arrays, saddle, tag, etc. that is worn by the animal 402. Additionally, or alternatively, the veterinary acoustic and biofeedback device 404 may be attached or worn by the animal 402 using attachment mechanisms, such as hook and loop, clasps, magnets, cords, chains, clips, belts and buckles, cords, pins, adhesives, etc. for coupling (e.g., pinning, hanging, etc.) in any manner to the animal 402. Still, in an embodiment, the veterinary acoustic and biofeedback device 404 may be sub-dermally implanted, injected, or ingested.

[0129] In an embodiment, the veterinary acoustic and biofeedback device 404, (and / or, as discussed above, a component of a system), may be separate from the animal 402 (i.e., not worn by the animal 402),placed in proximity to the animal 402. For example, the animal 402 may be housed in a barn, stable, bed, pen, cage, etc. Here, the veterinary acoustic and biofeedback device 404 may be positioned relative to the animal 402, or a group of animals, to monitor the animal 402 or the environment 400 of the animal 402. In an embodiment, the veterinary acoustic and biofeedback device 404 may be mounted on a wall or door of a pen, for example, attached to a fixture or piece of equipment within an animal enclosure, attached to fence(s) within a pasture, and so forth.

[0130] The veterinary acoustic and biofeedback device 404 may be used in association with any type of animal 402. For example, as shown in FIG. 4, the veterinary acoustic and biofeedback device 404 may be used in conjunction with a cow. However, other suitable animals may include livestock (e.g., sheep, pigs, etc.), pets (e.g., horses, dogs, cats, parrots, etc.), wildlife subjects (e.g., wolves, whales, dolphins, waterfowl, birds, etc.), and so forth. In an embodiment, the veterinary acoustic and biofeedback device 404 may designed or preconfigured for the animal 402. For example, a size, type, material, etc. of a harness may be dependent upon the animal 402. Moreover, in an embodiment, the animal 402 may be free to roam, graze, traverse within the environment 400, or may be housed within a barn, within fences, etc. In either scenario, the veterinary acoustic and biofeedback device 404 may be configured to monitor the animal 402 for the welfare, behavioral, and / or pathological states 408, and may utilize communication functions compatible with remote transmission componentry dispersed in the region where the animal 402 inhabits.

[0131] The veterinary acoustic and biofeedback device 404 may capture, generate, or receive audio data 412, such as vocalizations, sound, noises, etc. of the animal 402 or the environment 400 of the animal 402 and process the audio data 412 to detect the welfare, behavioral, and / or pathological states 408. Based on processing the audio data 412, the veterinary acoustic and biofeedback device 404 may provide real-time feedback to the caregiver 410 that prompts the caregiver 410 to treat or manage the animal 402, for example, based at least in part on the welfare, behavioral, and / or pathological states 408. For example, the caregiver 410 may provide assistant to the animal 402, administer vaccines, medications, etc. In an embodiment, the veterinary acoustic and biofeedback device 404 may continuously capture and process the audio data 412 to provide real-time feedback and further may incorporate additional video data where video may be real-time, constantly streaming, or triggered to record by a signal from the veterinary acoustic and biofeedback device 404.

[0132] In an embodiment, the veterinary acoustic and biofeedback device 404 may process the audio data 412 to determine characteristic(s) 414 of the vocalizations of the animal 402. The characterise c(s) 414may be compared to reference characteristics, thresholds, etc. for detecting when the vocalizations are indicative of certain welfare, behavioral, and / or pathological states 408 or include character! stic(s) associated with certain welfare, behavioral, and / or pathological states 408. In an embodiment, the characteristic(s) 414 may include or be represented by biomarkers 406, where the biomarkers 406 may identify or be associated with the characteristic(s) 414. As an example, the veterinary acoustic and biofeedback device 404 may detect one or more of the biomarkers associated with BVD or with PRDC or with Influenza A Virus (IAV) including subtypes such as H5N4, H5N3, and H7N9 for determining the welfare, behavioral, and / or pathological states 408. For example, in domesticated poultry or wild avian applications, respiratory pathology or irregularities could be detected by the veterinary acoustic and biofeedback device and software and reported to caretakers to promote action including vaccination or culling of infected individuals. In another example, the laryngeal and vocal productions of swine infected with PRDC could be monitored for symptomatic labored breathing, irregular ingressive and egressive respiration rates, foreign substance (such as mucous) expulsion, and so forth to determine the welfare, pathology, and behavior of animals. Increased probability of symptomatic laryngeal activity could be coupled with geographical location or inter-individual location tracking to estimate the likelihood of disease transmission and afford the caretakers or others opportunity to intervene or track changes in behavior, health, or well-being.

[0133] In an embodiment, the biomarkers 406 may include vocal biomarkers, physical biomarkers, environmental biomarkers, and / or audio biomarkers. In an embodiment, the vocal biomarkers may be associated with vocalizations (e.g., sound, communications, etc.) produced by the animal 402. In an embodiment, the vocalizations may be associated with any sounds generating or emanating from the animal 402, such as pain (e.g., groaning), distress (e.g., predators in area), social (e.g., signaling isolation / separation of a calf from its mother, mating calls, nursing or mating sounds, etc.), flatulence and excretion of solid and liquid animal waste, and so forth. In an embodiment, the vocal biomarkers may be determined based at least in part on character! stic(s) of the audio, such as tone, prosody, duration, amplitude, etc. In an embodiment, the physical biomarkers may be associated with physical activity of the animals, such as walking, calving, mating, voiding of waste, etc. The environment biomarkers may be associated with weather prediction, estimation of precipitation, identification of dangerous weather patterns (e.g., electrical storms, flooding, etc.), and existence in the environment of other hazards (e.g., proximity to traffic noise, predator sounds, rival animals fighting, etc.), humidity, etc. The audio markers may be associated with physiological, anatomical, behavioral, or mental state of the animal 402. The welfare ofanimals is indicated by their behavior, activity, and interactions. The physical and mental health of animals are measured as a function of animal behavior both inter- and intra-individually. Using predictors as mentioned herein, objective indicators are collected via the sensor(s) 448 of the veterinary acoustic and biofeedback device 404, used in the classification / discrimination models, and interpreted as actionable information for the animal 402 directly or for the caregiver 410. For example, increased lowing frequency and increased amplitude in a post-natal cow may indicate acute separation of cow and calf as perceived by the cow. This indication may be monitored by the veterinary acoustic and biofeedback device 404 and constitute cause for alerting the caregiver 410. Other biomarkers, such as individual or situational biomarkers, may be used to characterize experiences, activities, behaviors, and actions of the animal 402.

[0134] In an embodiment, the biomarkers 406 may be determined based at least in part on the audio data 412 generated by microphone(s) 424. The audio data 412 includes but it not limited to spectral and temporal analyses of the sampled signal including fundamental frequency, harmonics of vocal tract including formant frequencies, respiratory support, amplitude characteristics, source-fdter characteristics of vocal fold and laryngeal activity coupled to upper airway morphophysiology and anatomy, and vocal amplitude. These features are characterized and interpreted in combination, to infer physiological details of the animal 402 (e.g., vocal tract length, oral aperture, laryngeal port motor control, lingual shape characteristics, etc.), physio-temporal details (e.g., fine- and gross- motor control, etc.), and range / tendency / error / entropy / stability in the observed distributions in the signal. Additionally, or alternatively, sensor data 416 generated by sensor(s) 448 may be processed to determine the biomarkers 406. For example, sound captured by the microphone(s) 424 and data indicative of inhaling, exhaling, and vocalizing (e.g., via vocal cord vibration, for example) generated by the sensor(s) 448 may be processed to determine the biomarkers 406.

[0135] The veterinary acoustic and biofeedback device 404 is shown including processor(s) 420 and memory 422, where the processor(s) 420 may perform various functions and operations associated with detecting the welfare, behavioral, and / or pathological states 408, and the memory 422 may store instructions executable by the processor(s) 420 to perform the operations described herein. The veterinary acoustic and biofeedback device 404 may include any number of the sensor(s) 448 for detecting the vocalizations, behavior, etc. of the animal to determine the welfare, behavioral, and / or pathological states 408. For example, using the sensor(s) 448, the veterinary acoustic and biofeedback device 404 may determine the biomarkers 406 associated with the animal 402.

[0136] In an embodiment, the microphone(s) 424 may include at least two microphones. Any number of microphone(s) 424, however, may be included within the veterinary acoustic and biofeedback device 404. The microphone(s) 424 may be used for determining a directionality (e.g., time of flight, time-delay analytics, etc.) of sound and / or whether the sound corresponds to the animal 402 and / or other sources (e.g., other animals, weather, etc.). The microphone(s) 424 may generate respective audio signals that are representative of sound within the environment 400. In an embodiment, the microphone(s) 424 may be spaced apart by a different distance from a mouth of the animal 402, for example, to discern sounds that are generated by the animal 402 from other sources of sound within the environment 400.

[0137] The veterinary acoustic and biofeedback device 404 may include processing component(s) 426, such as echo cancellation, noise reduction, beamforming, and the like, to enable processing of the vocalizations and behavior of the animal 402. In an embodiment, the processing component(s) 426 may include a beamforming module used to process audio signals generated by the microphone(s) 424. Beamforming or spatial filtering is a signal processing technique for directional signal reception. Signals generated by the microphone(s) 424 may be processed in such a way that signals at particular angles experience constructive interference while others experience destructive interference. The beamforming techniques form multiple directional signals corresponding to different directions or orientations within the environment 400.

[0138] In an embodiment, the veterinary acoustic and biofeedback device 404 may self-adjust or calibrate (in real-time or over a course of repeated use sessions). Calibration may yield more sensitive or situationally adjusted results when determining the welfare, behavioral, and / or pathological states 408. The calibration and sensitization may operate by combining, for example, detection and coordinate processing of amplitude for an animal to refine or otherwise modify subsequent detection of the welfare, behavioral, and / or pathological states 408.

[0139] In an embodiment, other examples of the sensor(s) 448 of the veterinary acoustic and biofeedback device 404, such as piezoelectric sensors, accelerometers, etc., may generate signals that are processed to generate sensor data 416 that are used to determine the biomarkers 406. The sensor(s) 448 may additionally or alternatively include global positioning satellite (GPS), gyroscope(s), inertial measurement units (IMU), etc. The sensor data 416 generated from the sensor(s) 448, such as an IMU, may be used to determine vocalizations of the animal, for example, based on experienced vibrations, accelerations, inhaling / exhaling, pressure changes, and so forth. The sensor(s) 448 may be calibrated to optimize function and increase processing fidelity. The sensor data 416 collected by the sensor(s) 448 may be usedin combination with the audio data 412. Timestamps associated with the audio data 412 and the sensor data 416 may be associated with one another to correlate the sensor data 416 with the audio data 412. In an embodiment, to effectively detect vocalization, the sensor(s) 448 may be positioned in proximity to the clavicle (or furcula in birds, etc.) or thyroid / cricoid area (or equivalent), or near vocal production mechanism of the animal 402 to better detect movement or pressure changes associated with vocalization and / or respiration.

[0140] The audio data 412 and / or sensor data 416 may be analyzed to determine the biomarkers 406. In an embodiment, the biomarkers 406 may be associated with the vocalizations of the animal 402, such as pitch, intonation, tone, pauses, phonation, energy changes in the harmonic spectrum, fundamental frequency, voicing, windowed and absolute sonorant-peak rates, SNR, temporal and spectral voice characteristics, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, and / or motor processes. In an embodiment, the biomarkers 406 may be characterized by duration, frequency, decibel level, amplitude, energy, etc.

[0141] The biomarkers 406 may also be associated with movements of the animal 402 (e.g., bedding, running, walking, etc.), activities of the animal 402 (e.g., calving, drinking, eating, etc.), the environment (e.g., weather, temperature, etc.), etc.

[0142] In an embodiment, the veterinary acoustic and biofeedback device 404 may process the audio data 412 and / or the sensor data 416 to determine the biomarkers 406 for use in detecting the welfare, behavioral, and / or pathological states 408, such as diseases, distress, etc. The vocalizations of the animal 402 may be used to deduce the welfare, behavioral, and / or pathological states 408, or stated alternatively, the vocalizations of the animal 402 may be compared against characteristics indicative of the welfare, behavioral, and / or pathological states 408. In addition, the activities or non-activities of the animal 402 may be used to determine the welfare, behavioral, and / or pathological states 408, or whether the behavior of the animal 402 is indicative of the welfare, behavioral, and / or pathological states 408. In an embodiment, the biomarkers 406 may be compared to one or more predetermined thresholds or reference characteristic(s) indicative of the welfare, behavioral, and / or pathological states 408, for example, to determine whether the behavior of the animal 402 is indicative of a certain pathological state. In this manner, the biomarkers 406 may be compared to reference characteristics of animals with the welfare, behavioral, and / or pathological states 408 and without the welfare, behavioral, and / or pathological states 408. Through this comparison, a statistical likelihood may be determined whether the vocalizations and / or behavior of the animal 402 is indicative of the welfare, behavioral, and / or pathological states 408.

[0143] In an embodiment, the veterinary acoustic and biofeedback device 404 may determine audio fingerprint(s) 428 (or phenotype). The use of the audio fmgerprint(s) 428 may increase an accuracy in detecting vocalizations of the animal 402, discerning a type of vocalizations of the animal 402, or discerning other sounds emanating within the environment 400. Moreover, the audio fingerprint(s) 428 may be used to more accurately detect the biomarkers 406, and / or the welfare, behavioral, and / or pathological states 408. For example, in an embodiment, the biomarkers 406 may be used to generate audio fingerprint(s) 428 (e.g., audio signatures, acoustic fingerprint, etc.) associated with the animal 402, and the audio fingerprint(s) 428 may be analyzed to determine whether the audio fingerprint(s) 428 is associated with, or includes, characteristics of the welfare, behavioral, and / or pathological states 408. In an embodiment, the audio fingerprint(s) 428 may be compared to one or more reference fingerprints to detect the welfare, behavioral, and / or pathological states 408. Capturing sounds of the animal 402 allows for the determination of whether the poultry, for example, includes a pathological state that warrants care, treatment, etc. Each of the audio fingerprint(s) 428 may indicate biomarkers 406, and if a similarity between the audio fmgerprint(s) 428 and a stored reference audio voiceprint exists, the welfare, behavioral, and / or pathological states 408 may be detected. The audio fingerprint(s) 428 may be compared to any reference audio voiceprint for determining whether any number of welfare, behavioral, and / or pathological states 408 exist.

[0144] In an embodiment, the veterinary acoustic and biofeedback device 404 may output notification(s) 430 in instances where welfare, behavioral, and / or pathological states 408 are detected. The notification(s) 430 may provide feedback to the caregiver 410 such that the caregiver 410 may administer aid, assist, treat, etc. the welfare, behavioral, and / or pathological state 408. For example, if the animal 402 has a disease, the caregiver 410 may administer medicine or other intervention. The notification(s) 430 may be output in real-time, or substantially real-time, to provide near-instantaneous feedback to the caregiver 410. In an embodiment, the notification(s) 430 may be associated with the biomarkers 406 that are detected. This may permit the caregiver 410 to make informed decisions about the administered care, or whether to not to administer care. In an embodiment, the notification(s) 430 may be output on a mobile device 432 of the caregiver 410 and / or at the veterinary acoustic and biofeedback device 404.

[0145] In an embodiment, the notification(s) 430 may be audible, visual, haptic, and / or a combination thereof. The veterinary acoustic and biofeedback device 404 may include VO component(s) 434 that output the notification(s) 430. For example, lighting elements (e.g., LEDs) may output light at different brightness levels, flashing at different durations, frequencies and / or specific sequences (e.g., numericsequences of long and short flashes) in different modes, and / or by lighting in different colors. Speaker(s) may output an audible signal, such as a continuous prompt tone, sounding at different pitches or loudness in different modes, sounding at different durations, frequencies and / or specific sequences (e.g., numeric sequences of loud and soft, or high and low pitch sounds) in different modes. In an embodiment, these notification(s) 430 may assist in locating the animal 402 within the environment 400, and / or deducing which animal 402 amongst a group (e.g., herd, flock, etc.) of animals needs treatment. Other output components of the veterinary acoustic and biofeedback device 404 may include displays, touchscreens, etc. Example input devices may include button(s), switches, toggles, etc.

[0146] In an embodiment, the veterinary acoustic and biofeedback device 404 may store or have access to machine-learned (ML) model(s) 436. The ML model(s) 436 may be trained to analyze the audio data 412 and / or the sensor data 416 to determine whether the welfare, behavioral, and / or pathological states 408 of the animal 402.

[0147] In an embodiment, the ML model(s) 436 may determine the biomarkers 406 for determining the welfare, behavioral, and / or pathological states 408 of the animal 402. In an embodiment, the ML model(s) 436 may determine or generate score(s) associated with the audio data 412 and / or the sensor data 416, to determine whether the audio data 412 and / or the sensor data 416 are indicative of certain welfare, behavioral, and / or pathological states 408. In an embodiment, the score(s) may relate to a probability or likelihood that the vocalizations and / or behavior of the animal 402 is associated with certain welfare, behavioral, and / or pathological states 408. In other words, the score(s) output by the ML model(s) 436 may be machine-learned scores. The welfare, behavioral, and / or pathological states 408 may be based at least in part on the type of animal 402.

[0148] Machine learning generally involves processing a set of examples (called “training data”) to train a machine learning model(s). A machine learning model(s), once trained, is a learned mechanism that can receive new data as input and estimate or predict a result as output. For example, a trained machine learning model may comprise a classifier that is tasked with classifying unknown input (e.g., an unknown audio) as one of multiple class labels. In some cases, a trained machine learning model is configured to implement a multi-label classification task. Additionally, or alternatively, a trained machine learning model may be trained to infer a probability or a set of probabilities, for a classification task based on unknown data received as input. In the context of the present disclosure, the unknown input may be the audio data 412 and / or the sensor data 416 that is associated with the animal 402, and the ML model(s) 436 may be tasked with outputting the score that indicates or otherwise relates to, a probability of the animal402 having a particular welfare, behavioral, and / or pathological state 408. Additionally, the scores may indicate or otherwise relate to a probability of the biomarkers 406 for determining whether the vocalizations and / or behavior of the animal 402 is associated with the welfare, behavioral, and / or pathological states 408. Other data, such as medications of the animal 402, a history of the animal 402, etc. may be provided as an input to the ML model(s) 436.

[0149] The training data that is used to train ML model(s) 436 may include various types of data. In general, training data for machine learning may include two components, features and labels. However, in an embodiment, the training data used to train the ML model(s) 436 may be unlabeled. Accordingly, the ML model(s) 436 may be trainable using any suitable learning technique, such as supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, and so on. The features included in the training data may be represented by a set of features, such as in the form of an n-dimensional feature vector of quantifiable information about an attribute of the training data. The following is a list of example features that can be included in the training data for training the ML model(s) 436 described herein. However, it is to be appreciated that the following list of features is non-exhaustive, and features used in training may include additional features not described herein, and, in some cases, some, but not all, of the features listed herein. Example features included in the training data may include, without limitation, pitch, intonation, tone, pauses, articulation, decreased energy in the harmonic spectrum, fundamental frequency, SNR, temporal and spectral vocal characteristics, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, and / or motor processes. In an embodiment, the features included within the training data may be associated with vocalizations and / or behavior of the animal 402 that contains welfare, behavioral, and / or pathological states 408 and / or does not contain welfare, behavioral, and / or pathological states 408.

[0150] In an embodiment, as part of the training process, weights may be applied to a set of features included in the training data, as derived from the historical data. In an embodiment, the weights that are set during the training process may apply to parameters that are internal to the ML model(s) 436 (e.g., weights for neurons in a hidden layer of a neural network). These internal parameters of the ML model(s) 436 may or may not map one-to-one with individual input features of the set of features. The weights may indicate the influence that any given feature, parameter, or characteristic has on the score that is output by the ML model(s) 436.

[0151] The ML model(s) 436 represent a single model or an ensemble of base-level machine learning models and may be implemented as any type of machine learning model. For example, suitable machinelearning models for use with the techniques and systems described herein include, without limitation, neural networks, tree-based models, support vector machines (SVMs), kernel methods, random forests, splines (e.g., multivariate adaptive regression splines), hidden Markov model (HMMs), Kalman filters (or enhanced Kalman filters), Bayesian networks (or Bayesian belief networks), expectation maximization, genetic algorithms, linear regression algorithms, complex networks, nonlinear regression algorithms, logistic regression-based classification models, or an ensemble thereof. An “ensemble” can comprise a collection of machine learning models whose outputs (predictions) are combined, such as by using weighted averaging or voting. The individual machine learning models of an ensemble can differ in their expertise, and the ensemble can operate as a committee of individual machine learning models that is collectively “smarter” than any individual machine learning model of the ensemble.

[0152] In an embodiment, and as noted above, the ML model(s) 436 may identify or determine the characteristic(s) 414, such as the biomarkers 406, for comparison against the reference characteristic(s). The ML model(s) 436 may learn to identify complex relationships between the biomarkers 406 for determining whether the vocalizations and / or behavior is associated with the welfare, behavioral, and / or pathological states 408. For example, the ML model(s) 436 may learn to associate certain biomarkers 406 with whether the animal 402 has, or does not have, a certain welfare, behavioral, and / or pathological state 408.

[0153] In an embodiment, the ML model(s) 436 may learn to predict that the vocalizations and / or behavior is indicative of a welfare, behavioral, and / or pathological state 408 by attributing corresponding score(s). In this manner, biomarkers 406, for example, with low scores (e g., below threshold) may not include a particular welfare, behavioral, and / or pathological states 408, while the biomarkers 406 with high scores (e.g., above threshold) may include a particular welfare, behavioral, and / or pathological states 408. Although the use of a threshold is described as one example way of providing labeling (i.e., replaced or maintained), other techniques are contemplated, such as clustering algorithms, or other statistical approaches that use the trust scores may be used. The ML model(s) 436 is / are retrainable with new data to adapt the ML model(s) 436.

[0154] In addition to processing the audio data 412 and / or the sensor data 416 to detect the welfare, behavioral, and / or pathological states 408, the veterinary acoustic and biofeedback device 404 may utilize other data. For example, the veterinary acoustic and biofeedback device 404 may have access to data stored within a profile 438. The data may indicate demographics of the animal 402, environmental factors, medications of the animal 402, etc. This data may additionally, or alternatively be used to treat, detect,manage, and / or prevent the welfare, behavioral, and / or pathological states 408. For example, incorporating user-specific, treatment-related, non-vocal data may be used to more effectively treat and manage the welfare, behavioral, and / or pathological states 408.

[0155] The biomarkers 406 may be stored in association with the profile 438. The profile 438 may also store the audio data 412 and / or the sensor data 416 in association with the animal 402. Such data may be used by the caregiver 410 to further refine therapeutic benefits made possible by the veterinary acoustic and biofeedback device 404. The audio data 412 and / or sensor data 416 may also be stored in the memory 422 of the veterinary acoustic and biofeedback device 404 and provided (e.g., downloaded) to other devices.

[0156]

[0030] As shown, the veterinary acoustic and biofeedback device 404 may be communicatively coupled to one or more devices, such as the mobile device 432 and / or remote computing resource(s) 440 over network(s) 442. The network(s) 442 may include any viable communication technology, such as wired and / or wireless modalities and / or technologies. The network(s) 442 may include any combination of Personal Area Networks (PANs), Local Area Networks (LANs), Campus Area Networks (CANs), Metropolitan Area Networks (MANs), extranets, intranets, the Internet, short-range wireless communication networks (e.g., ZigBee, Bluetooth, etc.) Wide Area Networks (WANs) - both centralized and / or distributed - and / or any combination, permutation, and / or aggregation thereof. The veterinary acoustic and biofeedback device 404 includes suitable network interface(s) 444 for communicating over the network(s) 442.

[0157] In an embodiment, the veterinary acoustic and biofeedback device 404 may send the audio data 412 and / or the sensor data 416 to the remote computing resource(s) 440 for processing. The remote computing resource(s) 440 may receive the audio data 412 and / or the sensor data 416 from any number of veterinary acoustic and biofeedback device(s) 404 disposed within the environment 400, or other environments. The remote computing resource(s) 440 may have access to the components (e.g., ML model(s) 436) or additional components for processing the audio data 412 and / or the sensor data 416. In an embodiment, the remote computing resource(s) 440 may process the audio data 412 and / or the sensor data 416, determine the biomarkers 406, determine the welfare, behavioral, and / or pathological states 408, and then communicate the notification(s) 430 to the veterinary acoustic and biofeedback device 404, the mobile device 432, or other devices. Any level of split processing may be performed by the veterinary acoustic and biofeedback device 404, the mobile device 432, the remote computing resource(s) 440, and / or other devices, systems, networks, etc.

[0158] The remote computing resource(s) 440 may be implemented as one or more servers and may, in an embodiment, form a portion of a network-accessible computing platform implemented as a computing infrastructure of processors, storage, software, data access, etc. that is maintained and accessible via a network such as the Internet. The remote computing resource(s) 440 does not require end-user knowledge of the physical location and configuration of the system that delivers the services. Common expressions associated with the remote computing resource(s) 440 may include “on-demand computing”, “software as a service (SaaS)”, “platform computing”, “network-accessible platform”, “cloud services”, “data centers”, etc.

[0159] As used herein, a processor, such as the processor(s) 420 may include multiple processors and / or a processor having multiple cores. Further, the processor(s) 420 may comprise one or more cores of different types. For example, the processor(s) 420 may include application processor units, graphic processing units, and so forth. In one implementation, the processor(s) 420 may comprise a microcontroller and / or a microprocessor. The processor(s) 420 may include a graphics processing unit (GPU), a microprocessor, a digital signal processor or other processing units or components known in the art. Alternatively, or in addition, the functionally described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that may be used include field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system- on-a-chip systems (SOCs), complex programmable logic devices (CPLDs), etc. Additionally, each of the processor(s) 420 may possess its own local memory, which also may store program components, program data, and / or one or more operating systems.

[0160] Memory, such as the memory 422 may include volatile and nonvolatile memory, removable and non-removable media implemented in any method or technology for storage of information, such as computer-readable instructions, data structures, program component, or other data. Such memory may include but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD- ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, RAID storage systems, or any other medium which can be used to store the desired information and which can be accessed by a computing device. The memory may be implemented as computer-readable storage media (“CRSM”), which may be any available physical media accessible by the processor(s) to execute instructions stored on the memory. In one basic implementation, CRSM may include random access memory (“RAM”) and Flash memory. In otherimplementations, CRSM may include but is not limited to, read-only memory (“ROM”), electrically erasable programmable read-only memory (“EEPROM”), or any other tangible medium that can be used to store the desired information and which can be accessed by the processor(s) 420. The memory 422 is an example of non-transitory computer-readable media. The memory 422 may store an operating system and one or more software applications, instructions, programs, and / or data to implement the methods described herein and the functions attributed to the various systems.

[0161] FIG. 5 illustrates an example process 500 (e.g., method) related to determining whether vocalizations and / or behavior of the animal 402 are associated with certain welfares, behavioral, and / or pathological states 408, according to examples of the present disclosure. The process 500 described herein is illustrated as collections of blocks in logical flow diagrams, which represent a sequence of operations or acts, some or all of which may be implemented in hardware, software, or a combination thereof. In the context of software, the blocks may represent computer-executable instructions stored on one or more computer-readable media that, when executed by one or more processors, program the processors to perform the recited operations or acts. Generally, computer-executable instructions include routines, programs, objects, components, data structures and the like that perform particular functions or implement particular data types. The order in which the blocks are described should not be construed as a limitation, unless specifically noted. Any number of the described blocks may be combined in any order and / or in parallel to implement the process 500 or alternative processes, and not all of the blocks need be executed. For discussion purposes, the process 500 is described with reference to the environments, devices, architectures, diagrams, and systems described in the examples herein, such as, for example those described with respect to FIG. 4, although the process 500 may be implemented in a wide variety of other environments, architectures, and systems.

[0162] At 502, the process 500 may include receiving audio data from one or more microphone(s) of a first device. For example, the veterinary acoustic and biofeedback device 404 may receive the audio data 412 as generated from the microphone(s) 424. In an embodiment, the audio data 412 may be at least partially processed or unprocessed (e.g., filtered). In an embodiment, rather than receiving audio data 412, the veterinary acoustic and biofeedback device 404 may receive audio signals.

[0163] At 504, the process 500 may include receiving sensor data from one or more sensor(s) of the first device. For example, the veterinary acoustic and biofeedback device 404 may receive the sensor data 416 as generated from the sensor(s) 448. In an embodiment, the sensor data 416 may be at least partially processed or unprocessed (e.g., filtered). In an embodiment, rather than receiving sensor data 416, theveterinary acoustic and biofeedback device 404 may receive sensor signals from the sensor(s) 418. The sensor(s) may include accelerometer(s), piezoelectric sensor(s), IMUs, GPS, etc. The veterinary acoustic and biofeedback device 404 may receive the sensor data 416 from any number of sensor(s) 448, and / or the number of sensor(s) 448 that generate the sensor data 416 may be based at least in part on an application of the veterinary acoustic and biofeedback device 404, the animal 402, a welfare, behavioral, and / or pathological state 408 being examined, etc.

[0164] At 506, the process 500 may include determining biomarker(s) associated with the animal. In an embodiment, the biomarker(s) 406 may be associated with determining whether welfare, behavioral, and / or pathological states 408 are detected. Any number of the biomarkers 406 may be determined, and the biomarkers 406 may be associated with vocalizations of the animal 402, activity of the animal 402, an environment of the animal 402, etc. As non-limiting examples, the biomarkers 406 may include a pitch of vocalizations of the animal 402, intonations of vocalizations of the animal 402, tones associated with the vocalizations of the animal 402, changes associated with the vocalizations of the animal 402, weather within the environment 400, movement of the animal 402, eating habits associated with the animal 402, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, and / or motor processes. In an embodiment, the biomarkers 406 may be acoustic-based or nonacoustic based.

[0165] In an embodiment, the ML model(s) 436 may be used to process the audio data 412 and / or the sensor data 416 for determining the biomarkers 406. For example, the ML model(s) 436 may be previously trained to determine the biomarkers 406. In an embodiment, the ML model(s) 436 may receive the audio data 412 and / or the sensor data 416 as an input, and output an indication associated with whether the audio data 412 and / or the sensor data 416 are indicative of the biomarkers 406. In an embodiment, the ML model(s) 436 may output a score associated with the biomarkers 406, where the score may indicate likelihood or probability of the vocalizations and / or behavior of the animal 402 being associated with the biomarkers 406.

[0166] At 508, the process 500 may include determining whether the biomarker(s) satisfy predetermined threshold(s). For example, the biomarkers 406 may be compared to reference biomarker(s), or characteristic(s) that are indicative of the welfare, behavioral, and / or pathological states 408. Any number of the biomarkers 406 may be considered when determining whether the vocalizations and / or behavior are indicative of the welfare, behavioral, and / or pathological states 408. In an embodiment, the ML model(s) 436 may be used for determining whether the biomarkers 406 are indicative of satisfyingthreshold(s) associated with the vocalizations and / or behavior containing certain pathological states. As non-limiting examples, the biomarkers 406 may be associated with a pitch, amplitude, and / or tone associated with the vocalizations of the animal 402, and the pitch, amplitude, and / or tone may be compared to respective threshold(s). Based on such comparison, for example, the process 500 may determine whether the pitch, amplitude, tone, etc. are indicative of certain welfare, behavioral, and / or pathological states 408. Moreover, in an embodiment, audio fingerprint(s) 428 may be determined for comparison with one or more reference signatures to determine the welfare, behavioral, and / or pathological states 408 of the animal 402. If, at 508, the process 500 determines that the biomarkers 406 do not satisfy the threshold(s), the process 500 may follow the “NO” route and proceed to 510.

[0167] At 510, the process 500 may include causing output of one or more first notification(s). For example, the I / O component(s) 434 of the veterinary acoustic and biofeedback device 404 may output notification(s) 430, where the notification(s) 430 may indicate that the vocalizations and / or behavior is not associated with a particular welfare, behavioral, and / or pathological states 408. In an embodiment, the notification(s) 430 may be output at the veterinary acoustic and biofeedback device 404, or the mobile device 432 of the caregiver 410. From 510, the process 500 may proceed to 502, whereby the process 500 may continue to receive the audio data 412 and / or the sensor data 416 for determining the welfare, behavioral, and / or pathological states 408.

[0168] Returning to 508, if the process 500 determines that the biomarker(s) 406 satisfy the threshold(s), the process 500 may follow the “YES” route and proceed to 512. At 512, the process 500 may include causing output of one or more second notification(s). For example, the I / O component(s) 434 of the veterinary acoustic and biofeedback device 404 may output notification(s) 430, where the notification(s) 430 may indicate the welfare, behavioral, and / or pathological states 408.

[0169] At 514, the process 500 may include sending at least a portion of the audio data and / or the sensor data to a second device. For example, the veterinary acoustic and biofeedback device 404 may send the audio data 412 and / or the sensor data 416 to the mobile device 432 and / or the remote computing resource(s) 440. In an embodiment, the mobile device 432 and / or the remote computing resource(s) 440 may further process the audio data 412 and / or the sensor data 416, may store / log the audio data 412 and / or the sensor data 416 to maintain a history of the animal 402. In addition, the mobile device 432 may output the notification(s) 430 indicative of the welfare, behavioral, and / or pathological states 408. From 514, the process 500 may loop to 502, whereby the veterinary acoustic and biofeedback device 404 may continue to process and analyze vocalizations and / or behavior of the animal 402.Illustrative Embodiment of Third Device and Methods Thereof

[0170] This application is directed, at least in part, to monitoring user welfare to determine one or more behavioral and / or pathological states of the user, according to examples of the present disclosure. In an embodiment, sensor data generated from one or more sensors may be used to determine user welfare for detecting behavioral and / or pathological states, such as diseases, distresses, accidents, injuries, etc., to provide care, treatment, and / or management. In an embodiment, any number of sensors may be used to acoustically and non-acoustically monitor the user. For example, the sensors may capture data indicative of sounds generated by the user, movements of the user, vitals of the user, behavior of the wearer, etc. The sensor data may be analyzed to determine biomarkers associated with the user, where the biomarkers are used to determine a welfare of the user for deducing behavioral and / or pathological states of the user. In an embodiment, the biomarkers may be compared to one or more reference biomarkers to deduce the behavioral and / or pathological states. The user welfare and / or pathological states may be determined in real-time and / or over a period of time (e.g., day, month, year, etc.). In an embodiment, based at least in part on the welfare, behavioral, and / or pathological state, a caregiver may be notified to provide care, treatment, etc.

[0171] In an embodiment, the techniques described herein may be implemented within, or at least in part, by a device. In an embodiment, the device may be considered a wellness and healthcare monitoring device. For example, as will be discussed herein, the wellness and healthcare monitoring device may, in an embodiment, generate or utilize sensor data associated with acoustics of the user, non-acoustics of the user, and / or an environment of the user as a way to monitor the user and / or provide feedback to the user, a caregiver (e.g., physician, therapist, etc.) of the user, and so forth. However, although described as a wellness and healthcare monitoring device, the systems, methods, and techniques described herein may be utilized within applications other than healthcare.

[0172] In an embodiment, the wellness and healthcare monitoring device may be a compact device worn by the user. The wellness and healthcare monitoring device may be worn in any manner by the user, such as in the form of bracelets, watches, stick-on arrays, eyeglasses, caps, hearing aids, brooches, ties, lanyards, earrings, etc. The wellness and healthcare monitoring device may include suitable attachment mechanisms, such as hook and loop, clasps, magnets, cords, adhesives, chains, etc., for coupling (e.g., pinning, hanging, etc.) in any manner to the user. However, although described herein as a device worn by the user, in an embodiment, the techniques described herein may not be limited to such devices. For example, a device (e.g., stand-alone device, integrated device, etc.) employing the methods describedherein may be placed on a shelf, counter, wall, etc., within an environment where the user resides or occupies.

[0173] The wellness and healthcare monitoring device may include a housing in which components of the wellness and healthcare monitoring device are disposed. The housing may include any shape, such as cylindrical, square-shaped, etc. The housing may be manufactured from suitable materials, such as plastics, composites, metal, etc., and using any manufacturing process(es), such as injection molding, stamping, blow molding, etc. The housing may also be contaminate-resistant to prevent the ingress of liquid, dust, or contaminants into the housing. Additionally, the wellness and healthcare monitoring device may include suitable processing and internal components to detect and manage welfare, behavioral, and / or pathological states of the user. For example, as will be discussed herein, the wellness and healthcare monitoring device may include sensor(s), microphone(s), battery(ies), processing components (e.g., processors, memory, system on a chip (SoC), or integrated circuits (IC), etc.), etc. disposed within the housing. In an embodiment, the wellness and healthcare monitoring device may include I / O components, such as buttons, switches, lighting elements, and / or a display for outputting information associated with the user. The display may include an organic light emitting diode (O / LED), (in-plane switching (IPS) or thin-film-transistor (TFT)) liquid crystal display (LCD), laser video display (LVD), vibrotactile, haptic, or other applications.

[0174] As introduced above, the wellness and healthcare monitoring device may diagnose, manage, and / or monitor the welfare, behavioral, and / or pathological state or changes of the user. In an embodiment, the wellness and healthcare monitoring device may be administered to the user to diagnose, manage, and / or monitor the welfare, behavioral, and / or pathological state or changes in state. While the cause of certain welfare, behavioral, and / or pathological states may be unknown, characteristic(s) associated with the welfare, behavioral, and / or pathological states are often known. The wellness and healthcare monitoring device includes the sensor(s) to determine the characteristic(s) of the user, which, in turn, are compared to reference characteristics for determining the welfare, behavioral, and / or pathological states.

[0175] The wellness and healthcare monitoring device may include any number of sensor(s) to monitor the user and / or the environment of the user for determining the welfare, behavioral, and / or pathological states. Example sensor(s) include but are not limited to microphone(s), piezoelectric sensor(s), accelerometer(s), camera(s) or optical sensors, temperature sensor(s), heart rate sensor(s), humidity sensor(s), inertial measurement unit (IMU) sensor(s), gyroscope, ECC sensor(s), blood glucose sensor(s), breathing sensor(s), geo- and spatial location sensors such as GPS or RFID sensor(s), blood oxygensensor(s), Hall sensor(s), electrodermal sensors such as galvanic skin response, passive infrared (PIR) sensor(s), etc. In an embodiment, the sensor(s) may be used to identify vitals associated with the user, activities of the user, and non-activities of the user, as well as to characterize the vitals, activities, and / or non-activity. For example, the sensor(s) may be used to determine active or volitional activities of the user (e.g., vocalizations, movement, eating, sleeping, etc.) and / or non-activates, passive vegetative, automatic, autonomic (sympathetic, parasympathetic, enteric) / involuntary activities / states of the user (e.g., non-vocalizations, breathing, heart rate, blood oxygen level, etc.). Moreover, the sensor(s) may be used to determine any number of sounds or signals generated by or emanating from the user, whether audio, vocals, speech, murmuring, mumbling, etc.

[0176] The sensor(s) may capture audio data and / or sensor data that is analyzed to determine the character! stic(s) of the user. In an embodiment, the characteristic(s) may be associated with the sounds, behavior, vitals, activities, and / or non-activities of the user. In an embodiment, the character! stic(s) may be used to determine biomarker(s) associated with vocalizations and / or sounds of the user. For example, the biomarker(s) may be associated with pitch, intonation, tone, pauses, phonation, and / or changes associated with speech of the user. The biomarker(s) may also include features such as articulation, decreased energy in the higher parts of a harmonic spectrum, imprecise articulation of vowels and consonants, fundamental frequency, voicing, windowed and absolute syllable / sonorant-peak rates, SNR, temporal and spectral voice characteristics, frequency, spectral / cepstral cues, vocal quality / stability (e.g., shimmerjitter, harmonic to noise ratio), prosody, temporal output, amplitude stability, etc. Biomarker(s) may also be associated with non-acoustic characteristic(s) of the user, such as motor processes, vitals, EEG, EKG, motion, fatigue, physical activity, posture, gait, location, as well as environment biomarkers such as temperature, air quality, location, date / time, etc.

[0177] The wellness and healthcare monitoring device may include analog and digital converters that assist in capturing, processing, and storing signals collected from the microphone(s) and sensor(s), respectively. The signals from the microphone(s) and sensor(s) may be stored in a dataset including raw and / or normalized or transformed values. The audio signals or sensor signals may be processed as data, which is analyzed to determine the biomarkers.

[0178] In an embodiment, the microphone(s) may include at least two microphones spaced apart from one another within the housing. For example, the housing may include a first end and a second end spaced apart from the first end. A first of the microphones may be located closer to the first end than the second end, while a second of the microphones may be located closer to the second end than the first end. Anynumber of microphones, however, may be included within the wellness and healthcare monitoring device. In an embodiment, the microphone(s) may be used to determine a directionality (e.g., time of flight, velocimetry, time-delay analytics, etc.) of the signal and / or whether the signal corresponds to the sounds of the user wearing the wellness and healthcare monitoring device. For example, the microphone(s) may receive respective audio signals that represent sounds within the environment. When worn by the user, the microphones may be spaced apart by different distances from sound sources of the user (e.g., mouth of the user, stomach, etc.). As such, the microphone(s) are spaced apart by a different distance from the sound sources of the user to discern sounds generated by the user from other sources of sound within the environment.

[0179] The audio signals captured by the microphone(s) may be cross-correlated with one another. For example, cross-correlation may applied between audio signals detected by the microphones to determine a time difference of arrival (TDOA) for the correlated signals. In other words, the differently spaced microphones from the mouth of the user have different time of arrival (TOA) values. This allows the calculation of an angle from which the detected signals originate relative to a line intersecting the two microphones.

[0180] In an embodiment, the wellness and healthcare monitoring device may be configured to identify sounds emanating or generated from the user, as compared to sounds generated from other sources within an environment of the user. As introduced above, the sound of the user may include vocalizations (e.g., speech, moaning, grunting, etc.) generated from a mount of the user, but sound generated by digestion, body parts, etc. of the user. Moreover, the microphone(s) may capture sound or other acoustics of an environment of the user (e.g., weather, sounds of other people, etc.). In an embodiment, sounds may also be associated with the user, as compared to other sources of sound, based on detecting vocal vibrations through changes in pressure, breathing, acceleration, temperature, strain, force, etc. The wellness and healthcare monitoring device may detect vocal pathological or distress biomarkers, such as sonic or other detectable human vocal biomarkers associated with a targeted disease condition or welfare disorder. In an embodiment, the wellness and healthcare monitoring device detects one or more sonic markers associated with disease or distress, characterized by a diagnostic vocal, non-vocal acoustic or non-acoustic detected signal.

[0181] Capturing sounds of the user allows for the determination of whether the user includes a behavioral and / or pathological state that warrants care, intervention, treatment, etc. As such, sounds (e.g., pathological sounds) emitted from the user may be detected and analyzed to identify and inform acaregiver regarding the pathological state of the user, whether normal, suspect, or determined pathological or distress experienced by the user. In an embodiment, these different vocalizations may have or be characterized by reference signatures. As will be discussed herein, in an embodiment, signatures (e.g., audio, motion, etc.) may be compared against reference signatures to determine the welfare, behavioral, and / or pathological state of the user.

[0182] In addition to acoustics, accelerometers may capture sensor signals, or sensor data, to determine behaviors of the users, such as movements (e.g., running, digestion, etc.), vitals (e.g., temperature, pulse rate, respiration rate, blood pressure, etc.), activities (e.g., breathing, heart rate, congestion, coughing, etc.), and so forth. For example, heart rate sensor(s) may be used to determine health of the user (e.g., electrocardiogram, photoelectric pulse wave, blood pressure measurement, phonocardiography, etc.), weather sensor(s) may be used to determine weather conditions (e.g., temperature of the user and / or environment), and so forth.

[0183] The biomarkers may be determined from audio data generated by microphone(s) of the wellness and healthcare monitoring device, and / or sensor data generated from sensor(s) of the wellness and healthcare monitoring device. The biomarkers may be compared against reference biomarkers or thresholds, for example, to determine the welfare, behavioral, and / or pathological state of the user. In an embodiment, the behavioral and / or pathological state of the user may be determined based at least in part on the welfare. In an embodiment, the reference biomarkers may be considered diagnostic or classificatory biomarkers in that the biomarkers are compared to the reference biomarkers to determine the welfare, behavioral, and / or pathological state or changes in state of the user. For example, the wellness and healthcare monitoring device may detect one or more biomarkers associated with congenital heart disease, which may be characterized by heart murmurs, rapid breathing, shortness of breath, rapid heartbeat, etc. (e.g., congenital heart disease biomarkers). Here, the wellness and healthcare monitoring device may determine characteristic(s) of the user for determining the biomarkers, and in turn, may compare the biomarkers singly or in combination / interaction to reference biomarkers indicative of congenital heart disease. Based on this comparison, the wellness and healthcare monitoring device may determine whether the characteristic(s) of the user are indicative of congenital heart disease.

[0184] The wellness and healthcare monitoring device may be useful to detect any number of types of pathological conditions, distress states, behavior, and a wide range of other activities and circumstances detectable from a user and / or their environment. In an embodiment, the wellness and healthcare monitoring device may be configured to detect the particular welfare, behavioral, and / or pathological stateof the user. In this sense, the wellness and healthcare monitoring device may be tailored to the user (e.g., with a known disease, for example) or may be used to diagnose welfare, behavioral, and / or pathological states. Depending upon the welfare(s), behavioral, and / or pathological state(s), comparable sets of reference biomarkers may be employed.

[0185] For example, the wellness and healthcare monitoring device may be specifically programmed and / or optimized for the specific biomarkers and / or character! stic(s) that are already established or known for a behavioral and / or pathological state for a given disease, as well as the disease, the user, and / or the environments. For example, continuing with the above example, if a user has congenital heart disease, the wellness and healthcare monitoring device may be programmed to monitor for characteristics (e.g., heart rate, breathing patterns, movement, etc.) indicative of congenital heart disease, as compared to other diseases. Moreover, in an embodiment, the wellness and healthcare monitoring device (or the sensors) may be calibrated to each disease, behavioral and / or pathological state of the disease, the user, the environment, location, etc. The comparison between the biomarkers and the reference biomarkers or thresholds may determine a statistical likelihood of the welfare, behavioral, and / or pathological state of the user.

[0186] As introduced above, reference biomarker(s), whether acoustic or non-acoustic, may be captured, stored as sample data, and analyzed in conjunction with audio data and / or sensor data detected by the wellness and healthcare monitoring device to detect, diagnose, and manage a wide range of welfares, behavioral, and / or pathological states. Example diseases include, but are not limited to, infectious diseases, as well as metabolic disorders, gate disorders, respiratory diseases, gastrointestinal diseases, a wide range of non-pathological behaviors and activities (both volitional and autonomic), etc. Early detection and monitoring of these diseases and other adverse welfare states greatly enhance human health and welfare and may minimize the need for costly medical intervention. In an embodiment, the welfare of the user may be associated with healthy, sick, diseased, etc. In an embodiment, the welfare of the user may be used to classify or determine the behavioral and / or pathological states of the user. Pathological states may be related to, but are not limited to, respiration, feeding, water intake, movement, digestive, gastrointestinal, disease, metabolism, gate, and so forth. These pathological states may be correspondingly utilized to identify, characterize, diagnose, treat, etc. diseases (e.g., states, factors, etc.), disabilities, distresses, welfare disturbances, pathologies, activities, congenital defects, effects of medications, environmental effects, accidents, harmful physical events, etc. associated with the user. Sensor and sampled data may also be used to identify and classify non-pathological states that are also ofinterest to behaviors, activities, states, changes, and observations of the device user and those interacting with the device user. Biomarkers of non-pathological behavior and states are also of interest and can be the target of classification and “diagnostic” determinations that contribute to interventions, suggestions, analyses, and so on for bearing on quality-of-life, improved performance, reflection, monitoring, aptitude, assessment, tracking of behavior / conditions, and other data.

[0187] In an embodiment, the wellness and healthcare monitoring device may generate fingerprint(s), unique signature(s), or phenotype(s) based at least in part on the biomarkers. The fingerprint(s) may be individual to the user and permit the user to be distinguished from one another. In an embodiment, the use of the fingerprint(s) may discern sound generated from the user from other sounds emanating within the environment. This allows the sound to be associated with a certain user to accurately determine the welfare, behavioral, and / or pathological state of a particular user. The fingerprint(s) may also improve the accuracy of determining welfare, behavioral, and / or pathological states. For example, in an embodiment, the biomarkers may be used to generate fingerprint(s) associated with the user, and the fingerprint(s) may be analyzed to determine whether the fingerprint(s) is associated with behavioral and / or pathological states. In an embodiment, the fmgerprint(s) may characterize the sounds, activities, movements, vitals, and the fmgerprint(s), for example, and may be compared to one or more reference fingerprint(s) to determine whether the welfare, behavioral, and / or pathological states based on a comparison to reference fingerprint(s).

[0188] The fmgerprint(s) may be based at least in part on the audio data and / or the sensor data. In doing so, acoustics and non-acoustic characteristic(s) may be used to indicate a certain welfare, behavioral, and / or pathological state of the user. For example, fingerprint(s) may associate acoustic sounds (e.g., breathing) with non-acoustic sounds (e.g., an acceleration of the chest during breathing) for comparison to characterise c(s) of the user to determine the welfare, behavioral, and / or pathological states. In other words, audio data and / or sensor data may be used to characterize the user, thereby generating a fingerprint associated with the sounds, non-sounds, vitals, movement, etc. of the user, and such fingerprint may be compared to reference fingerprints to determine the welfare, behavioral, and / or pathological states. Through this comparison, like characteristics (e.g., sound to sound, movement to movement, etc.) may be compared with one another to determine the welfare, behavioral, and / or pathological states.

[0189] The wellness and healthcare monitoring device may utilize other data associated with the user to determine the welfare, behavioral, and / or pathological state of the user. For example, a profile of the user may store information such as age, sex, medications, past medical history, location, etc. This informationmay be used to increase accuracy in detecting the welfare, behavioral, and / or pathological state of the user. For example, the information may be considered when determining the welfare, behavioral, and / or pathological state of the user. That is, if a user is taking medications and is prone to certain diseases based on prior diagnoses, the wellness and healthcare monitoring device may utilize such information when determining the welfare, behavioral, and / or pathological states.

[0190] The wellness and healthcare monitoring device may output the notification(s) in an attempt to notify the user of their welfare, behavioral, and / or pathological state. The notification(s) may also be sent to other devices. For example, a caregiver of the user may receive the notification(s) as a way to be notified about welfare and / or pathological states, diagnose the welfare and / or pathological states, treat the welfare and / or pathological states, and / or manage the welfare and / or pathological states.

[0191] The notification(s) may be output in real-time, or substantially real-time, to provide near- instantaneous feedback to the user regarding their welfare, behavioral, and / or pathological state. In an embodiment, the notification(s) may be audible, visual, olfactory, electronic, haptic, etc., and / or any combination thereof. For example, the wellness and healthcare monitoring device may include a speaker that outputs audio (e.g., deeps, tones, instructions, etc.), lighting elements that illuminate (e.g., patterns, colors, etc.), and / or a motor that vibrates. Noted above, the notification(s) may also be output on a display of the wellness and healthcare monitoring device. The wellness and healthcare monitoring device may output the notification(s) for a predetermined amount of time, until the user responds, etc. In an embodiment, the wellness and healthcare monitoring device may tailor the notification(s) based on one or more preference(s) stored in association with a profile of the user. For example, the profile may indicate a type of notification(s) to output, as well as their associated intensity (e.g., volume, luminosity, etc.), duration, type (e.g., sound-haptic-sound), patterns (e.g., low haptic-medium haptic-low haptic, etc.), combinations thereof, etc. The wellness and healthcare monitoring device may access the preference(s) as stored in the profile when outputting the notification(s).

[0192] In addition to real-time feedback as described above, asynchronous reports may also be delivered to the user and stored in the software. For example, a daily histography digest of performance and behavior may be delivered via email or software (e.g., a mobile application) to the user. Performance data may also be delivered to a database collection performance characteristics from many users constituting the basis for additional software and algorithm changes. The application and / or software may be embedded into the wellness and healthcare monitoring device or may be external software capable of being instantiated into external hardware such as a mobile phone, tablet, or computer. Although thetechniques are described herein as being performed by the wellness and healthcare monitoring device, the techniques may be embodied as software or an application operating on a device (e.g., mobile phone). Suitable APIs may interact and / or utilize sensor(s) of device(s), such as camera(s), microphone(s), location, temperature, etc. to carry out the operations described herein.

[0193] In an embodiment, the wellness and healthcare monitoring device may utilize artificial intelligence (Al) and machine learning (ML) models to improve the accuracy of determining the welfare, behavioral, and / or pathological states of the user. To clarify, Al and ML techniques, which are described and detailed with examples below, are used as encompassing cover terms for “smart” and algorithmic collection, processing, analyzing, databasing, managing, summarizing, and using data in an automated and large- scale fashion. The ML model(s) may be trained from a database (e.g., historical data, such as audio data that includes respiration, feeding, water intake, movement, digestive, gastrointestinal, disease, metabolism, gate, vitals, disabilities, distresses, welfare disturbances, pathologies, activities, congenital defects, effects of medications, environmental effects, accidents, harmful physical events, etc. of the user) to analyze the data captured by the sensor(s) and microphone(s) for identifying the biomarkers and determining the welfare, behavioral, and / or pathological states. The ML model(s) may determine the biomarkers and assess the biomarkers in comparison to information stored in the database (e.g., reference characteristics, reference fingerprint(s), etc.) to determine the welfare, behavioral, and / or pathological states.

[0194] As part of the ML model(s) analyzing the audio data and / or the sensor data, the ML model(s) may label the characteristic(s) of the activities, non-activities, and / or behavior of the user singly or in combination, for example, the biomarkers, to indicate whether the character! stic(s) are associated certain welfare, behavioral, and / or pathological states. In an embodiment, an output of the ML model(s) may indicate whether the user has a particular welfare, behavioral, and / or pathological state (e.g., healthy, diseased, injured, etc.). In an embodiment, the ML model(s) may determine the characteristic(s) for comparison to respective reference characteristics to determine whether activities, non-activities, and / or behaviors of the user are indicative of, or are not indicative of, a particular welfare, behavioral, and / or pathological state. The ML model(s) may use any number of character! stic(s) for determining the welfare, behavioral, and / or pathological states.

[0195] In an embodiment, the ML model(s) may generate scores, metrics, classifications, or determinations that are associated with the characteristic(s) of the welfare, behavioral, and / or pathological states. For example, the ML model(s) may generate scores associated with the biomarkers, welfare,behavioral, and / or pathological states. For example, if the scores are greater than a predetermined threshold (which varies depending on the particular data and desired information relayed in return), the ML model(s) may return a quantitative confidence that the characteristics are associated with a particular welfare, behavioral, and / or pathological states. In an embodiment, the ML model(s) may compare fingerprint(s) / phenotype(s) of the user, as generated from the audio data and / or sensor data, against one or more reference fingerprints. A behavioral and / or pathological state may be assessed with regard to the model parameters, and a discrete or probabilistic classification will be generated. The classification may be stored, processed, and used for feedback (as needed).

[0196] In an embodiment, the ML model(s) may receive or be trained from characteristic(s) that are specific to any number of diseases. The ML model(s), having been trained, may receive the audio data and / or sensor data to predict and / or identify the behavioral and / or pathological states of the user. In an embodiment, the ML model(s) may be compared character! stic(s) of the audio data and / or the sensor data with reference character! stic(s) associated with the diseases, respectively. As an example, in the context of opioid use, the ML model(s) may identify correlations (e.g., factors, variables, features, etc.) to predict and / or determine whether a user has an opioid addiction. Such correlations may be based on a given condition or state associated with a disease, disorder, condition, socioeconomic status, location, etc.

[0197] The ML model(s) may be trained to determine characteristic(s) of the audio data and / or sensor data, or character! stic(s) of an individual behavioral and / or pathological state. The ML model(s) may be trained from data collected from users with similar demographics (e.g., sex, age, ethnicity, etc.). For example, the ML model(s) may be trained from amplitudes associated with the audio data, sex, race, age, disease state, audiological status, medical history, socioeconomic status (SES), geographical location, etc., as well as environmental factors (e.g., time of day, medication schedule, auditory and acoustic environment, etc.). The ML model(s) may be dynamic and retrained to improve an accuracy in determining welfare, behavioral, and / or pathological states.

[0198] Although the wellness and healthcare monitoring device is described as processing, analyzing, and / or determining the characteristic(s) of the user, the biomarkers, welfare, behavioral, and / or pathological states, etc., in an embodiment, the wellness and healthcare monitoring device may communicatively couple to one or more computing devices, resource(s), etc. which may additionally or alternatively process, analyze, and / or determine the characteristic(s) of the user, the biomarkers, and welfare, behavioral, and / or pathological states. For example, the wellness and healthcare monitoring device may receive or capture sensor data that is further transmitted to remote computing resource(s) foranalysis. The remote computing resource(s) may receive the sensor data from any number of the wellness and healthcare monitoring device(s) or separate device(s). In an embodiment, the remote computing resource(s) may analyze the sensor data, among other data, to determine the biomarkers, welfare, pathological states, etc. Thereafter, the remote computing resource(s) may transmit notification(s) to the wellness and healthcare monitoring device, the device of the caregiver, healthcare provider, other organization, and so forth. Any level of split processing may be performed by the wellness and healthcare monitoring device, the device of the caregiver, the remote computing resource(s), and / or other devices, systems, networks, etc.

[0199] In an embodiment, the wellness and healthcare monitoring devices may be used in conjunction with one another, such as in a system, to determine the user welfare, behavior, and / or pathological states or activity. For example, a first wellness and healthcare monitoring device (i.e., a first system component) worn on the user may be used in conjunction with a second wellness and healthcare monitoring device (i.e., a second system component) placed within an environment of the user. In such instances, the wellness and healthcare monitoring device may be the same or different, for example, such as in size, functionality, capabilities, etc. For example, the first wellness and healthcare monitoring device may include first sensors that are different from the second sensors of the second wellness and healthcare monitoring device. In instances where a plurality of the wellness and healthcare monitoring devices are used, the wellness and healthcare monitoring devices may be communicatively coupled to one another via one or more network interface(s) (e.g., Bluetooth, Cellular, etc.). Any combination of the wellness and healthcare monitoring devices worn on the user and / or off the user may be used.

[0200] The wellness and healthcare monitoring device may be used within a plurality of environments and / or for any suitable purposes. For example, the wellness and healthcare monitoring device may find use within healthcare, neurology, speech pathology, speech training, physical therapy, occupational therapy, professional caregiving, physical therapy, occupational therapy, occupational safety, sporting events, home environment, leisure activities, during periods of rest or activity, social gatherings / events, during meal preparation, personal hygiene care, sleeping, reading, etc. The wellness and healthcare monitor device may be utilized to determine any disease, such as Parkinson's disease, Alzheimer's disease, Frontotemporal dementia, Vascular dementia, ALS, stroke, hereditary Prior disease, etc. The wellness and healthcare monitoring device may be utilized to determine prodromal, symptomatic, unknown, suspected, or other markers and biomarkers of disease and disorder. The wellness and healthcare monitoring device may be utilized to determine and assist in non-disease related behavior and states,including classifications and descriptions of states, activities, and behaviors that contribute to wellness, well-being, quality of life, mobility, active lifestyles, independent living, mood, psychiatric / psychological / psychosocial well-being, and so on.

[0201] The present disclosure provides an overall understanding of the principles of the structure, function, device, and system disclosed herein. One or more examples of the present disclosure are illustrated in the accompanying drawings. Those of ordinary skill in the art will understand and appreciate that the devices, the systems, and / or the methods specifically described herein and illustrated in the accompanying drawings are non-limiting embodiments. The features illustrated or described in connection with one embodiment, or instance, may be combined with the features of other embodiments or instances. Such modifications and variations are intended to be included within the scope of the disclosure and appended claims.

[0202] FIG. 6 illustrates an example user 600 wearing a wellness and healthcare monitoring device 602, according to examples of the present disclosure. Details of the wellness and healthcare monitoring device 602 are discussed herein, however, generally, the wellness and healthcare monitoring device 602 may determine biomarkers 604 of the user 600 for detecting welfare, behavior, and / or pathological states 606 of the user 600. Detection of the welfare, behavior, and / or pathological states 606 allows for a caregiver 608 (e.g., physician, therapist, etc.) to treat and / or manage the user 600.

[0203] The wellness and healthcare monitoring device 602 may be a compact, wearable, or peripheral device. In an embodiment, the wellness and healthcare monitoring device 602 may be worn on a body of the user 600. In an embodiment, the wellness and healthcare monitoring device 602 may be small, non- invasive, non-intrusive, lightweight, and discreet in nature. As shown in FIG. 6, the wellness and healthcare monitoring device 602 may be worn as a necklace or lanyard. However, the wellness and healthcare monitoring device 602 may be attached or worn by the user 600 using mechanisms such as hook and loop, clasps, magnets, cords, chains, clips, belts and buckles, cords, pins, adhesives, etc. for coupling (e.g., pinning, hanging, etc.) in any manner to the user 600. Still, in an embodiment, the wellness and healthcare monitoring device 602 may be sub-dermally implanted, injected, or ingested. As such, although shown in a particular embodiment in FIG. 6, the wellness and healthcare monitoring device 602 may be attached or worn by the user 600 in other manners.

[0204] In an embodiment, the wellness and healthcare monitoring device 602 may be separate from the user 600, such as not being worn by the user 600, placed in proximity to the user 600 or caregiver 608.For example, the wellness and healthcare monitoring device 602 may be positioned relative to the user, such as on a shelf, stand, etc. to monitor the user600 and / or an environment of the user 600.

[0205] The wellness and healthcare monitoring device 602 may capture, generate, or receive audio data 660, such as vocalizations, sound, noises, etc. of the user 600 or the environment and process the audio data 660 to detect the welfare, behavior, and / or pathological states 606. Based on processing the audio data 660, the wellness and healthcare monitoring device 602 may provide real-time feedback to the caregiver 608 that prompts the caregiver 608 to treat or manage the user 600, for example, based at least in part on the welfare, behavior, and / or pathological states 606. For example, the caregiver 608 may provide assistance to the user 600, administer vaccines, medications, etc. In an embodiment, the wellness and healthcare monitoring device 602 may continuously capture and process the audio data 660 to provide real-time feedback and / or triggered to capture the audio data 660.

[0206] In an embodiment, the wellness and healthcare monitoring device 602 may process the audio data 660 to determine character! stic(s) 662 of the vocalizations of the user 600. The characteristic(s) 662 may be compared to reference characteristics, thresholds, etc. for detecting when the vocalizations indicate certain welfare, behavior, and / or pathological states 606 or include characteristic(s) associated with certain welfare, behavior, and / or pathological states 606. In an embodiment, the characteristic(s) 662 may include or be represented by biomarkers 604, where the biomarkers 604 may identify or be associated with the characteristic(s) 662. As an example, the wellness and healthcare monitoring device 602 may detect one or more of the biomarkers associated with congenital heart disease, diabetes, measles, diphtheria, etc. For example, the wellness and healthcare monitoring device 602 may detect a simultaneous combination of increased electrodermal activity and decreased heart rate in a user known to be of a particular age and race. These factors, in combination, may increase the probability of acute risk and prompt intervention or notification to the user 600, the caregiver 608, and / or a healthcare provider. For example, the wellness and healthcare monitoring device 602 may detect slowed or slurred speech, increased response latency to conversational partners, or increased syllable durations in speech productions, thus prompting notification or documentation of the biomarkers 604 in combination or alone. In this example, speech production and language behavior may be assessed in combination with pharmaceutical intake, pharmacotherapeutic interventions, or other medications to give feedback or information to the user 600 and / or those designated such as caregivers or researchers.

[0207] In an embodiment, the biomarkers 604 may include vocal biomarkers, physical biomarkers, environmental biomarkers, and / or audio biomarkers. In an embodiment, the vocal biomarkers may beassociated with vocalizations (e.g., sound, communications, etc.) produced by the user 600. In an embodiment, the vocalizations may be associated with any sounds generated or emanating from the user 600, such as pain (e.g., groaning), distress (e.g., crying), social (e.g., conversation, rate, loudness, etc.), stomach grumbling, and so forth. In an embodiment, the vocal biomarkers may be determined based at least in part on character! stic(s) of the audio, such as tone, prosody, duration, amplitude, etc. In an embodiment, the physical biomarkers may be associated with physical activity of the user 600, such as walking, vitals, etc. The environmental biomarkers may be associated with the weather, a location of the user 600, temperature, humidity, etc. The audio biomarkers may be associated with a physiological, anatomical, behavioral, or mental state of the user 600. Other biomarkers, such as individual or situational biomarkers may be used to characterize experiences, activities, behaviors, actions, of the user 600. For example, the wellness and healthcare device 602 may classify a reduction in communicative contributions of the user 600 in noisy environments, during specific times of the day, or with specific interlocutors or companions. This complex behavioral data may contribute the ability to bear on meaningful quality of life, wellness, well-being interventions, suggestions, logging, or analyses. In the preceding example, a notification reminder may be produced to the user 600 to increase the frequency and amplitude of speech production and language contributions when a specific companion is detected by the wellness and healthcare device 602. This alert / notification could be produced in real-time and in retrospective fashion for later, off-line analysis.

[0208] In an embodiment, the welfare of the user 600 may be indicated by their behavior, activity, and interactions. The physical and mental health of the user 600 may be measured as a function of the behavior of the user 600, both inter- and intra-individually. Using predictors as mentioned herein, objective indicators are collected via sensor(s) 664 of the wellness and healthcare monitoring device 602, used in the classification / discrimination models, and interpreted as actionable information for the user 600 directly or for the caregiver 608.

[0209] In an embodiment, the biomarkers 604 may be determined based at least in part on the audio data 660 generated by microphone(s) 666. The audio data 660 includes but is not limited to spectral and temporal analyses of the sampled signal including fundamental frequency, harmonics of vocal tract including formant frequencies, respiratory support, amplitude characteristics, source-filter characteristics of vocal fold and laryngeal activity coupled to upper airway morphophysiology and anatomy, and vocal amplitude. These features are characterized and interpreted in combination, to infer physiological details of the user 600 (e.g., vocal tract length, oral aperture, laryngeal port motor control, lingual shapecharacteristics, etc.), physio-temporal details (e.g., fine- and gross- motor control, etc.), and range / tendency / error / entropy / stability in the observed distributions in the signal. Additionally, or alternatively, sensor data 668 generated by the sensor(s) 664 may be processed to determine the biomarkers 604. For example, sound captured by the microphone(s) 666 and sensor data 668 indicative of inhaling, exhaling, vocalizing (e.g., via vocal cord vibration, for example) generated by the sensor(s) 664 may be processed to determine the biomarkers 604.

[0210] The wellness and healthcare monitoring device 602 is shown including processor(s) 620 and memory 622, where the processor(s) 620 may perform various functions and operations associated with detecting the welfare, behavior, and / or pathological states 606, and the memory 622 may store instructions executable by the processor(s) 620 to perform the operations described herein. The wellness and healthcare monitoring device 602 may include any number of the sensor(s) 664 for detecting the vocalizations, behavior, etc. of the user 600 to determine the welfare, behavior, and / or pathological states 606.

[0211] In an embodiment, the microphone(s) 666 may include at least two microphones. Any number of microphone(s) 666, however, may be included within the wellness and healthcare monitoring device 602. The microphone(s) 666, in addition to capturing sound associated with the user 600, may be used for determining a directionality (e.g., time of flight, time-delay analytics, etc.) of sound and / or whether the sound corresponds to the user 600 and / or other sources (e.g., background, weather, etc.). The microphone(s) 666 may generate respective audio signals that are representative of sound within the environment. In an embodiment, the microphone(s) 666 may be spaced apart by a different distance from a mouth of the user 600, for example, to discern sounds that are generated by the user 600 from other sources of sound within the environment.

[0212] The wellness and healthcare monitoring device 602 may include speech-processing component(s) 624, such as speech recognition, digital signal processing (DSP), automatic speech processing and recognition (ASP and ASR), natural language processing (NLP), echo cancellation, noise reduction, beamforming, and the like to enable processing of vocalizations of the user 600. In an embodiment, the speech-processing component(s) 624 may include a multi-dimensional non-linear beamforming module used to process audio signals captured by the microphone(s) 666. Beamforming or spatial (multiplex) filtering is a signal processing technique for directional signal reception. Signals generated by the microphone(s) 666 may be processed in such a way that signals at particular angles experience constructive interference while others experience destructive interference. The beamforming techniquesform multiple directional signals corresponding to different directions or orientations within the environment associated with the vocalizations. As sounds are received from a particular direction, the directional signal (e.g., formed beam) associated with that direction tends to exhibit more energy or signal strength than the other signals (or beams).

[0213] In an embodiment, the speech-processing component(s) 624 may be configured to perform speech recognition, such as automatic speech recognition (ASR), natural language processing (NLP), artificial neural networks (ANNs) such as large language models (LLMs), and / or natural language understanding (NLU) on the audio data 660 to identify words or phrases associated with the user speech, or may be configured to provide the audio data 660 to another device for performing the ASR, NLU, and / or NLP on the audio data 660. The speech-processing component s) 624 may also reduce noise, identify same voice, double-talk, speaker segregation, talker in noise, differentiation in the cocktail party effect, echo, and / or attenuate any signal components associated with noise other than the user 600. In turn, after processing the audio signals, a clean-high quality audio signal (or the audio data 660) may be generated that is associated with the talker or the user 600.

[0214] The speech-processing component(s) 624 may also include a VAD component 626 that detects speech of the user 600, as compared to other sounds within the environment (e g., background noise, speech from other persons in the environment, etc.) and / or other talkers. For example, the VAD component 626 may confirm that the sounds generated within the environment correspond to vocalizations of the user 600. The VAD component 626 may perform VAD techniques on the audio data 660 utilizing one or more VAD algorithms based on channel energy with long and short-term energies, sub-band long and short-term energies with combination logic, deep neural network (DNN) based VADs, or any other type of VAD algorithms, with hangover windows, to determine whether the audio data 660 is representative of vocalizations of the user 600. DNN approaches including multilayer perceptrons (MLPs), recurrent neural networks (RNNs), convolutional neural networks (CNNs), and post processing including hysteretic / durati on-based thresholding, noise cancellation, or complex extension filtering may be employed.

[0215] More particularly, in an embodiment, the VAD component 626 may employ binary labeling, e.g., with speech and non-speech frames labeled 6 and 0, respectively. Using multiple microphone(s), VAD can be applied to each channel independently. Subsequently, each outputted binary label can be unified with a frame-by-frame logical sum operation. TDOA estimation is done, for example, using a generalized correlation method to estimate time delays. Using VAD in combination with TDOA in the wellness andhealthcare monitoring device 602 having multiple microphone(s) 666, as described, readily distinguishes when the user 600 is speaking or not speaking. Calculating signal-to-noise ratio (SNR) may be based on a postprocessing algorithm such as hysteretic thresholding applied after the VAD is determined to discern which parts of detected sounds are of the user 600 and which parts are environmental noise or other talkers.

[0216] In an embodiment, other examples of the sensor(s) 664 of the wellness and healthcare monitoring device 602, such as piezoelectric sensors, accelerometers, etc., may capture signals that are processed to generate sensor data 668 that is used to determine the biomarkers 604. The sensor(s) 664 may additionally or alternatively include piezoelectric sensor(s), accelerometer(s), camera(s), temperature sensor(s), heart rate sensor(s), humidity sensor(s), inertial measurement unit (IMU) sensor(s), ECC sensor(s), gyroscope, blood glucose sensor(s), breathing sensor(s), GPS / RFID sensor(s), blood oxygen sensor(s), Hall sensor(s), passive infrared (PIR) sensor(s), etc. In an embodiment, the sensor(s) 664 may be used to identify vitals associated with the user 600, activities of the user 600, and non-activities of the user 600, as well as to characterize the vitals, activities, and / or non-activity. For example, the sensor(s) 664 may be used to determine activities of the user 600 (e.g., vocalizations, movement, eating, sleeping, etc.) and / or nonactivates of the user 600 (e.g., non-vocalizations, breathing, heart rate, blood oxygen level, etc.). Moreover, the sensor(s) may be used to determine any number of sounds generated by or emanating from the user 600, whether audio, vocals, speech, murmuring, mumbling, etc.

[0217] The wellness and healthcare monitoring device 602 may detect vocal pathological or distress biomarkers, for example, by detecting sonic or other detectable human vocal biomarkers associated with a targeted disease condition or welfare disorder. In an embodiment, the wellness and healthcare monitoring device 602 detects one or more sonic markers associated with disease or distress, characterized by a diagnostic vocal, non-vocal acoustic, or non-acoustic detected signal.

[0218] The sensor data 668 generated from the sensor(s) 664, such as an IMU, may be used to determine vocalizations of the user 600, for example, based on experienced vibrations, accelerations, inhaling / exhaling, pressure changes, and so forth. The sensor(s) 664 may be calibrated to optimize function and increase processing fidelity. For example, in an embodiment, the wellness and healthcare monitoring device 602 may self-adjust or calibrate (in real-time, or throughout repeated use sessions). Calibration may yield more sensitive or situationally adjusted results when determining the welfare, behavior, and / or pathological states 606. The calibration and sensitization may operate by combining, for example, detection and coordinate processing of amplitude for the user 600 to refine or otherwise modify subsequent detection of the welfare, behavior, and / or pathological states 606.

[0219] The sensor data 668 collected by the sensor(s) 664 may be used in combination with the audio data 660. Timestamps associated with the audio data 660 and the sensor data 668 may be associated with one another to correlate the sensor data 668 with the audio data 660. In an embodiment, to effectively detect vocalization, the sensor(s) 664 be positioned in proximity to the clavicle, thyroid, or cricoid area (or equivalent), or near vocal production mechanism of the user 600 to better detect movement or pressure changes associated with vocalization, physiological state or change, and / or respiration.

[0220] The audio data 660 and / or sensor data 668 may be analyzed to determine the biomarkers 604. In an embodiment, the biomarkers 604 may be associated with the vocalizations of the user 600, such as pitch, intonation, tone, pauses, phonation, energy changes in the harmonic spectrum, fundamental frequency, voicing, windowed and absolute sonorant-peak rates, SNR, temporal and spectral voice characteristics, frequency, spectral / cepstral cues, vocal quality / stability (e.g., shimmerjitter, harmonic to noise ratio), prosody, temporal output, and amplitude stability, and / or motor processes. In an embodiment, the biomarkers 604 may be characterized by duration, frequency, decibel level, amplitude, energy, etc. Biomarker(s) may also be associated with non-acoustic characteristic(s) of the user, such as motor processes, vitals, EEG, EKG, microsiemens, motion, fatigue, physical activity, posture, gait, location, as well as environment biomarkers such as temperature, air quality, location, date / time, etc. The biomarkers 604, as discussed above, may also be associated with movements of the user 600 (e.g., sleeping, running, walking, etc.), activities of the user (e.g., coughing, drinking, eating, etc.), the environment (e.g., weather, temperature, etc.).

[0221] In an embodiment, the wellness and healthcare monitoring device 602 may process the audio data 660 and / or the sensor data 668 to determine the biomarkers 604 for detecting the welfare, behavior, and / or pathological states 606, such as diseases, distresses, injuries, etc. The vocalizations of the user 600 may be used to deduce the welfare, behavior, and / or pathological states 606, or stated alternatively, the volitional or vegetative vocalizations of the user 600 may be compared against characteristics indicative of the welfare, behavior, and / or pathological states 606. In addition, the activities or non-activities of the user 600 may be used to determine the welfare, behavior, and / or pathological states 606, or whether the behavior of the user 600 is indicative of the welfare, behavior, and / or pathological states 606.

[0222] The reference biomarker(s), whether acoustic or non-acoustic, may be collected, sampled, generated, and / or stored as sample data, and analyzed in conjunction with audio data 660 and / or sensor data 668 detected by the wellness and healthcare monitoring device 602 to detect, diagnose, and manage a wide range of welfare, behavioral, and / or pathological states 606. Example diseases may include, butare not limited to, infectious diseases, metabolic disorders, gate disorders, respiratory diseases, gastrointestinal diseases, non-pathological conditions or states, sleeping, agitation, running, walking, psychological states, posture, etc. In an embodiment, the welfare of the user 600 may be associated with healthy, sick, diseased, etc., in discrete, graded, categorical, continuous, or probabilistic fashion. In an embodiment, the welfare of the user 600 may be used to classify or determine the behavioral and / or pathological states of the user600. Pathological states may be related to, but are not limited to, respiration, feeding, water intake, movement, digestive, gastrointestinal, disease, metabolism, gate, and so forth. These pathological states may be correspondingly utilized to identify, characterize, diagnose, treat, etc. diseases (e.g., states, factors, etc.), disabilities, distresses, welfare disturbances, pathologies, activities, congenital defects, effects of medications, environmental effects, accidents, harmful physical events, etc. associated with the user 600. Input, sensor, and device data may also be used to categorize, classify, and otherwise use data from sources adjacent to the user 600 or interacting with the user 600.

[0223] In an embodiment, the biomarkers 604 may be compared to one or more predetermined thresholds or reference character! stic(s) indicative of the welfare, behavior, and / or pathological states 606, for example, to determine whether the behavior of the user 600 is indicative of a particular behavior or pathological state. In this manner, the biomarkers 604 may be compared to reference characteristics of user 600 with particular welfare, behavior, and / or pathological states 606 and user 600 without the welfare, behavior, and / or pathological states 606. Through this comparison, a statistical likelihood may be determined whether the vocalizations and / or behavior of the user 600 is indicative of the welfare, behavior, and / or pathological states 606.

[0224] For example, the wellness and healthcare monitoring device 602 may detect one or more biomarkers associated with congenital heart disease, characterized by heart murmurs, rapid breathing, shortness of breath, fast heartbeat, etc. (e.g., congenital heart disease biomarkers). Here, the wellness and healthcare monitoring device 602 may determine characteristic(s) 662 of the user 600 for determining the biomarkers 604 and, in turn, may compare the biomarkers 604 to reference biomarkers indicative of congenital heart disease. Based on this comparison, the wellness and healthcare monitoring device 602 may determine whether the character! stic(s) 662 of the user 600 indicate congenital heart disease. However, the wellness and healthcare monitoring device 602 may be helpful to detect any number of types of pathological conditions, distress states, and a wide range of other activities and circumstances detectable from a user 600 and / or their environment. Depending upon the welfare, behavioral, and / or pathological states 606, comparable sets of reference biomarkers may be employed.'ll

[0225] In an embodiment, the wellness and healthcare monitoring device 602 may be configured to detect particular welfare, behavioral, and / or pathological states 606 of the user 600. In this sense, the wellness and healthcare monitoring device 602 may be tailored to the user 600 (e.g., with a known disease, for example) or may be used to diagnose welfare, behavioral, and / or pathological states 606. For example, the wellness and healthcare monitoring device 602 may be specifically programmed and / or optimized for the specific biomarkers 604 and / or character! stic(s) 662 that are already established or known for a behavioral and / or pathological state for a given disease, as well as the disease, the user 600, and / or the environments. For example, continuing with the above example, if a user 600 has congenital heart disease, the wellness and healthcare monitoring device 602 may be programmed to monitor for characteristics (e.g., heart rate, breathing patterns, movement, etc.) indicative of congenital heart disease, as compared to other diseases. Moreover, in an embodiment, the wellness and healthcare monitoring device 602, the microphone(s) 666, and / or the sensor(s) 664 may be calibrated to each disease, behavioral and / or pathological state of the disease, the user, the environment, location, etc. The comparison between the biomarker(s) 604 and the reference biomarkers or thresholds may determine a statistical likelihood of the welfare, behavioral, and / or pathological state of the user.

[0226] In an embodiment, the wellness and healthcare monitoring device 602 may determine fingerprint(s) 628 (or phenotype). The use of the fingerprint(s) 628 may increase an accuracy in detecting vocalizations of the user 600, discerning a type of vocalizations of the user 600, or discerning other sounds emanating within the environment. Moreover, the fingerprint(s) may be used to more accurately detect the biomarkers 604, and / or the welfare, behavior, and / or pathological states 606. For example, in an embodiment, the biomarkers 604 may be used to generate fingerprint(s) 628 (e.g., signature(s), audio signatures, acoustic fingerprint, etc.) associated with the user 600, and the fingerprint(s) 628 may be analyzed to determine whether the fingerprint(s) 628 is associated with, or includes, characteristics of the welfare, behavior, and / or pathological states 606. In an embodiment, the fingerprint(s) 628 may be compared to one or more reference fingerprints to detect the welfare, behavior, and / or pathological states 606. Capturing sounds of the user 600 allows for determining whether the user 600 includes a pathological state that warrants care, treatment, intervention, alert, etc. Each of the fingerprint(s) 628 may indicate biomarkers 604, and if a similarity between the audio fingerprint(s) 628 and a stored reference audio voiceprint exists, the welfare, behavior, and / or pathological states 606 may be detected. The fingerprint(s) 628 may be compared to any reference fingerprint(s) to determine whether any number of welfare, behavior, and / or pathological states 606 exist.

[0227] In an embodiment, the wellness and healthcare monitoring device 602 may output notification(s) 630 based at least in part on the welfare, behavior, and / or pathological states 606. The notification(s) 630 may provide feedback to the user 600, healthcare provider, and / or caregiver 608, for example, such that the caregiver 608 may administer aid, assist, treat, etc. the welfare, behavior, and / or pathological state 606. For example, if the user 600 has a disease, the caregiver 608 may administer medicine or provide other interventions. The notification(s) 630 may be output in real-time, or substantially real-time, to provide near-instantaneous feedback to the caregiver 608. In an embodiment, the notification(s) 630 may be associated with the biomarkers 604 that are detected. This may permit the caregiver 608 to make informed decisions about the care to administer, or whether to not to administer care. In an embodiment, the notification(s) 630 may be output on a mobile device 632 of the caregiver 608 and / or at the wellness and healthcare monitoring device 602.

[0228] In an embodiment, the notification(s) 630 may be audible, visual, haptic, and / or a combination thereof. The wellness and healthcare monitoring device 602 may include I / O component(s) 634 that output the notification(s) 630. For example, lighting elements (e.g., LEDs) may output light at different brightness levels, flashing at different durations, frequencies and / or specific sequences (e.g., numeric sequences of long and short flashes) in different modes, and / or by lighting in different colors. Speaker(s) may output an audible signal, such as a continuous prompt tone, sounding at different pitches or loudness in different modes, sounding at different durations, frequencies and / or specific sequences (e.g., numeric sequences of loud and soft, or high and low pitch sounds) in different modes. Other output components of the wellness and healthcare monitoring device 602 may include displays, touchscreens, etc. Example input devices may include button(s), switches, toggles, etc.

[0229] In an embodiment, the wellness and healthcare monitoring device 602 may store or have access to machine-learned (ML) model(s) 636. The ML model(s) 636 may be trained to analyze the audio data 660 and / or the sensor data 668 to determine the welfare, behavior, and / or pathological states 606 of the user 600. The ML model(s) 636 may determine the biomarkers 604 for determining the welfare, behavior, and / or pathological states 606. In an embodiment, the ML model(s) 636 may determine or generate metrics or score(s) associated with the audio data 660 and / or the sensor data 668 to determine whether the audio data 660 and / or the sensor data 668 are indicative of particular welfare, behavior, and / or pathological states 606. In an embodiment, the score(s) may relate to a probability or likelihood that the vocalizations, activity, vitals, behavior, etc. of the user 600 is associated with certain welfare, behavior,and / or pathological states 606. In other words, the score(s) output by the ML model(s) 636 may be machine-learned scores.

[0230] Machine learning generally involves processing a set of examples (called “training data”) to train a machine learning model(s). Once trained, a machine learning model(s) is a learned mechanism that can receive new data as input and estimate or predict a result as output. For example, a trained machine learning model may comprise a classifier that is tasked with classifying unknown input (e.g., an unknown audio or sensor data) as one of multiple class labels. In some cases, a trained machine learning model is configured to implement a multi-label classification task. Additionally, or alternatively, a trained machine learning model including, for example, Bayesian machine models (BMMs) relying on Bayesian networks or Dirichlet processes in complex hyperparameter environments or dynamical systems complex networks may be trained to infer a probability or a set of probabilities, for a classification task based on unknown data received as input. In the context of the present disclosure, the unknown input may be the audio data 660 and / or the sensor data 668 that is associated with the user 600, and the ML model(s) 636 may be tasked with outputting the score that indicates or otherwise relates to, a probability or residual probability of the user 600 having a particular welfare, behavior, and / or pathological state 606. Additionally, the scores may indicate or otherwise relate to a probability of the biomarkers 604 for determining whether the vocalizations, activities, vitals, and / or behavior of the user 600 is associated with the welfare, behavior, and / or pathological states 606. Other data, such as medications of the user 600, a history of the user 600, etc. may be provided as an input to the ML model(s) 636. Such information, for example, may be stored within a profile of the user 600 and others who have contributed data to the sensor(s) on the wellness and healthcare monitoring device 602.

[0231] The training data used to train ML model(s) 636 may include various data types. In general, training data for machine learning may include two or more components, features and labels. However, in an embodiment, the training data used to train the ML model(s) 636 may be unlabeled, sparsely categorized, or underspecified. Accordingly, the ML model(s) 636 may be trainable using any suitable learning technique, such as supervised learning, unsupervised learning, semi-supervised learning, reinforcement learning, and so on. The features included in the training data may be represented by a complex set of features, such as in the form of an n-dimensional feature vector of quantifiable information about an attribute of the training data. The following is a list of example features that can be included in the training data for training the ML model(s) 636 described herein. However, it is to be appreciated that the following list of features is non-exhaustive, and features used in training may include additionalfeatures not described herein, and, in some cases, some, but not all, of the features listed herein. Example features included in the acoustic training data may include, without limitation, pitch, intonation, tone, pauses, articulation, decreased energy in the harmonic spectrum, fundamental frequency, SNR, temporal and spectral vocal characteristics, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, and / or motor processes. Example features included in the behavioral training data may include, without limitation, movement, position, eye and pupil orientation and aperture, facial positioning, eating / drinking / walking / running / sleeping, passive / inactive states, personal hygiene and associated personal care, motion, relational activity, and / or volitional / vegetative fine and gross motor activity. Example features included in the environmental training data may include, without limitation, geolocation, relative location, interpersonal space, orientation in social time and / or space, psychosocial relationships, elevation, humidity, altitude, ambient temperature, particulate matter, air quality, ambient acoustic noise, heat, light, vibration, and / or olfactory influences. In an embodiment, the features included within the training data may be associated with vocalizations and / or behavior of the user 600 that contains welfare, behavior, and / or pathological states 606 and / or does not contain welfare, behavior, and / or pathological states 606.

[0232] In an embodiment, as part of the training process, weights may be applied to a set of features included in the training data, as derived from the historical data. In an embodiment, the weights that are set during the training process may apply to parameters that are internal to the ML model(s) 636 (e.g., weights for neurons in a hidden layer of a neural network). These internal parameters of the ML model(s) 636 may or may not map one-to-one with individual input features of the set of features. The weights may indicate the influence that any given feature, parameter, or characteristic has on the score that is output by the ML model(s) 636.

[0233] The ML model(s) 636 represent a single model or an ensemble of base-level machine learning models and may be implemented as any type of machine learning model. For example, suitable machine learning models for use with the techniques and systems described herein include, without limitation, neural networks, tree-based models, support vector machines (SVMs), kernel methods, random forests, splines (e g., multivariate adaptive regression splines), hidden Markov model (HMMs), Kalman filters (or enhanced Kalman filters), Bayesian networks (or Bayesian belief networks), expectation maximization, genetic algorithms, linear regression algorithms, complex networks, nonlinear regression algorithms, logistic regression-based classification models, or an ensemble thereof. An “ensemble” can comprise a collection of machine learning models whose outputs (predictions) are combined, such as by usingweighted averaging or voting. Models may be predictive or descriptive. The individual machine learning models of an ensemble can differ in their expertise, and the ensemble can operate as a committee of individual machine learning models that is collectively “smarter” than any individual machine learning model of the ensemble.

[0234] In an embodiment, and as noted above, the ML model(s) 636 may identify or determine the characteristic(s) 662, such as the biomarkers 604, for comparison against the reference characteristic(s). The ML model(s) 636 may learn to identify complex relationships between the biomarkers 604 to determine whether the vocalizations, activity, vitals, and / or behavior are associated with the welfare, behavior, and / or pathological states 606. For example, the ML model(s) 636 may leam to associate certain biomarkers 604 with whether the user has, or does not have, a certain welfare, behavior, and / or pathological state 606.

[0235] In an embodiment, the ML model(s) 636 may leam to predict that the vocalizations and / or behavior is indicative of a welfare, behavior, and / or pathological state 606 by attributing corresponding score(s). In this manner, biomarkers 604, for example, with low scores (e.g., below threshold) may not include a particular welfare, behavior, and / or pathological state 606, while the biomarkers 604 with high scores (e.g., above threshold) may include a particular welfare, behavior, and / or pathological state 606. Although the use of a threshold is described as one example way of providing labeling (i.e., replaced or maintained), other techniques are contemplated, such as clustering algorithms, or other statistical approaches that use the trust scores may be used. The ML model(s) 636 is / are retrainable with new data to adapt the ML model(s) 636.

[0236] In addition to processing the audio data 660 and / or the sensor data 668 to detect the welfare, behavior, and / or pathological states 606, the wellness and healthcare monitoring device 602 may utilize other data. For example, the wellness and healthcare monitoring device 602 may have access to data stored within a profile 638. The data may indicate demographics of the user 600, environmental factors, medications of the user 600, a history of medical diagnoses, condition, etc. of the user 600, and so forth. This data may additionally, or alternatively be used to treat, detect, manage, and / or prevent the welfare, behavior, and / or pathological states 606. For example, incorporating user-specific, treatment-related, non-vocal data may be used to more effectively treat and manage the welfare, behavior, and / or pathological states 606.

[0237] The biomarkers 604 may be stored in association with the profile 638. The profile 638 may also store the audio data 660 and / or the sensor data 668 in association with the user 600. The caregiver 608may use such data to further refine therapeutic benefits made possible by the wellness and healthcare monitoring device 602. The audio data 660 and / or sensor data 668 may also be stored in the memory 622 of the wellness and healthcare monitoring device 602 and provided (e.g., downloaded) to other devices.

[0238] The wellness and healthcare monitoring device 602 may be communicatively coupled to one or more devices, such as the mobile device 632 and / or remote computing resource(s) 640 over network(s) 642. The network(s) 642 may include any viable communication technology, such as wired and / or wireless modalities and / or technologies. The network(s) 642 may include any combination of Personal Area Networks (PANs), Local Area Networks (LANs), Campus Area Networks (CANs), Metropolitan Area Networks (MANs), extranets, intranets, the Internet, short-range wireless communication networks (e.g., ZigBee, Bluetooth, etc.) Wide Area Networks (WANs) - both centralized and / or distributed - and / or any combination, permutation, and / or aggregation thereof. The wellness and healthcare monitoring device 602 includes suitable network interface(s) 644 for communicating over the network(s) 642.

[0239] In an embodiment, the wellness and healthcare monitoring device 602 may send the audio data 660 and / or the sensor data 668 to the remote computing resource(s) 640 for processing. The remote computing resource(s) 640 may receive the audio data 660 and / or the sensor data 668 from any number of the wellness and healthcare monitoring devices 602 disposed within the environment. The remote computing resource(s) 640 may have access to the components (e.g., ML model(s) 636) or additional components for processing the audio data 660 and / or the sensor data 668. In an embodiment, the remote computing resource(s) 640 may process the audio data 660 and / or the sensor data 668, determine the biomarkers 604, determine the welfare, behavior, and / or pathological states 606, and then communicate the notification(s) 630 to the wellness and healthcare monitoring device 602, the mobile device 632, or other devices. Any level of split processing may be performed by the wellness and healthcare monitoring device 602, the mobile device 632, the remote computing resource(s) 640, and / or other devices, systems, networks, etc.

[0240] Although the techniques are described herein as being performed by the wellness and healthcare monitoring device 602, the techniques may be embodied as software or an application operating on a device (e g., mobile phone). Suitable APIs may interact and / or utilize sensor(s) of device(s), such as camera(s), microphone(s), location, temperature, etc. to carry out the operations described herein.

[0241] The remote computing resource(s) 640 may be implemented as one or more servers and may, in an embodiment, form a portion of a network-accessible computing platform implemented as a computing infrastructure of processors, storage, software, data access, etc. that is maintained and accessible via anetwork such as the Internet. The remote computing resource(s) 640 does not require end-user knowledge of the physical location and configuration of the system that delivers the services. Common expressions associated with the remote computing resource(s) 640 may include “on-demand computing”, “software as a service (SaaS)”, “platform computing”, “network-accessible platform”, “cloud services”, “data centers”, etc.

[0242] As used herein, a processor, such as the processor(s) 620 may include multiple processors and / or a processor having multiple cores. Further, the processor(s) 620 may comprise one or more cores of different types. For example, the processor(s) 620 may include application processor units, graphic processing units, and so forth. In one implementation, the processor(s) 620 may comprise a microcontroller and / or a microprocessor. The processor(s) 620 may include a graphics processing unit (GPU), a microprocessor, a digital signal processor or other processing units or components known in the art. Alternatively, or in addition, the functionally described herein can be performed, at least in part, by one or more hardware logic components. For example, and without limitation, illustrative types of hardware logic components that may be used include field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system- on-a-chip systems (SOCs), complex programmable logic devices (CPLDs), etc. Additionally, each of the processor(s) 620 may possess its own local memory, which also may store program components, program data, and / or one or more operating systems.

[0243] Memory, such as the memory 622 may include volatile and nonvolatile memory, removable and non-removable media implemented in any method or technology for storage of information, such as computer-readable instructions, data structures, program component, or other data. Such memory may include but is not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD- ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, RAID storage systems, or any other medium which can be used to store the desired information and which can be accessed by a computing device. The memory may be implemented as computer-readable storage media (“CRSM”), which may be any available physical media accessible by the processor(s) to execute instructions stored on the memory. In one basic implementation, CRSM may include random access memory (“RAM”) and Flash memory. In other implementations, CRSM may include but is not limited to, read-only memory (“ROM”), electrically erasable programmable read-only memory (“EEPROM”), or any other tangible medium that can be used to store the desired information and which can be accessed by the processor(s) 620. The memory 622 isan example of non-transitory computer-readable media. The memory 622 may store an operating system and one or more software applications, instructions, programs, and / or data to implement the methods described herein and the functions attributed to the various systems.

[0244] FIG. 7 illustrates a plurality of the wellness and healthcare monitoring devices 602 being worn by the user 600, according to examples of the present disclosure. As introduced above, the wellness and healthcare monitoring device 602 may take any shape, may be coupled or otherwise attached to the user 600 in different manners, and / or may have different functionalities.

[0245] For example, the user 600 may wear a first wellness and healthcare monitoring device 602(6), a second wellness and healthcare monitoring device 602(2), and a third healthcare monitoring device 602(3). The first wellness and healthcare monitoring device 602(6) may be worn around a neck of the user 600, for example, via a lanyard. The second wellness and healthcare monitoring device 602(2) may be worn around a wrist of the user, for example, as a watch. The third wellness and healthcare monitoring device 602(3) may be worn as a stick-on array, patch, etc. Each of the wellness and healthcare monitoring devices 602(6)-(3) may have suitable components to permit their operation, such as sensors, batteries, processor(s), etc., similar to those discussed above with regard to FIG. 6.

[0246] In an embodiment, the wellness and healthcare monitoring device(s) 602(6)-(3) may have different functionalities or different sensor(s) for determining the biomarkers 604 of the user 600. For example, the first wellness and healthcare monitoring device 602(6) may have microphone(s) to capture sounds generated by the user 600, the second wellness and healthcare monitoring device 602(2) may have a pulse oximeter sensor, a heart rate sensor, etc., and the third wellness and healthcare monitoring device may have a temperature sensor, accelerometer, etc. As such, the wellness and healthcare monitoring device(s) 602(6)-(3) may include different sensor(s) for generating the sensor data 668 for determining the biomarkers.

[0247] In an embodiment, the wellness and healthcare monitoring device(s) 602(6)-(3) may communicatively couple one another. Additionally, or alternatively, the wellness and healthcare monitoring device(s) 602(6)-(3) may communicatively couple to the mobile device 632 and / or the remote computing resource(s) 640. For example, the wellness and healthcare monitoring device(s) 602(6)-(3) may send sensor data 668 to the mobile device 632, whereby the mobile device 632 may analyze the sensor data 668 across the wellness and healthcare monitoring device(s) 602(6)-(3) to determine the biomarkers 604 and / or the welfare, behavioral, and / or pathological states 606.

[0248] Although the wellness and healthcare monitoring device(s) 602(6)-(3) are shown being worn by the user 600, the wellness and healthcare monitoring device(s) 602(6)-(3) may communicatively couple to other wellness and healthcare monitoring device(s), whether worn by the user 600 or not worn by the user 600 (e.g., stand-alone device).

[0249] FIG. 8 illustrates an example process 800 (e.g., method) related to determining whether welfares, behavioral, and / or pathological states 606 of the user 600, according to examples of the present disclosure. The process 800 described herein is illustrated as collections of blocks in logical flow diagrams, which represent a sequence of operations or acts, some or all of which may be implemented in hardware, software, or a combination thereof. In the context of software, the blocks may represent computerexecutable instructions stored on one or more computer-readable media that, when executed by one or more processors, program the processors to perform the recited operations or acts. Generally, computerexecutable instructions include routines, programs, objects, components, data structures and the like that perform particular functions or implement particular data types. The order in which the blocks are described should not be construed as a limitation, unless specifically noted. Any number of the described blocks may be combined in any order and / or in parallel to implement the process 800 or alternative processes, and not all of the blocks need be executed. For discussion purposes, the process 800 is described with reference to the environments, devices, architectures, diagrams, and systems described in the examples herein, such as, for example those described with respect to FIGS. 6 and 2, although the process 800 may be implemented in a wide variety of other environments, architectures, and systems.

[0250] At 802, the process 800 may include receiving audio data from one or more microphone(s) of a first device. For example, the wellness and healthcare monitoring device 602 may receive the audio data 660 as generated from the microphone(s) 666. In an embodiment, the audio data 660 may be at least partially processed or unprocessed (e.g., filtered). In an embodiment, rather than receiving audio data 660, the wellness and healthcare monitoring device 602 may receive audio signals from the microphone(s) 666.

[0251] At 804, the process 800 may include receiving sensor data from one or more sensor(s) of the first device. For example, the wellness and healthcare monitoring device 602 may receive the sensor data 668 as generated from the sensor(s) 664. In an embodiment, the sensor data 668 may be at least partially processed or unprocessed (e.g., filtered). In an embodiment, rather than receiving sensor data 668, the wellness and healthcare monitoring device 602 may receive sensor signals from the sensor(s) 664. The sensor(s) 664 may include accelerometer(s), piezoelectric sensor(s), IMUs, GPS, etc. The wellness andhealthcare monitoring device 602 may receive the sensor data 668 from any number of sensor(s) 664, and / or in some instances, the wellness and healthcare monitoring device 602 may receive sensor data 668 from other wellness and healthcare monitoring devices 602 that are communicatively coupled to the wellness and healthcare monitoring device 602. Any number of the sensor(s) 664 may be disposed across the wellness and healthcare monitoring devices 602 and / or the number and / or type of the sensor(s) 664 may be based at least in part on an application of the wellness and healthcare monitoring device 602, the user 600, and / or the welfare, behavioral, and / or pathological state 606 being examined, etc. For example, if the user 600 has a medical history of a certain disease, or the wellness and healthcare monitoring device 602 is being employed to treat, manage, etc. a certain disease, the types of sensor(s) 664 on the wellness and healthcare monitoring device 602 may be tailored accordingly.

[0252] At 806, the process 800 may include determining biomarker(s) associated with the user. In an embodiment, the biomarker(s) 604 may be associated with determining welfare, behavioral, and / or pathological states 606. Any number of the biomarkers 604 may be determined from the audio data 660 and / or the sensor data 668. The biomarkers 604 may be associated with sounds, vocalizations, activities, non-activities, vitals, etc. of the user 600, as well as characteristics of an environment of the user 600. As non-limiting examples, the biomarkers 604 may include a pitch of vocalizations of the user600, intonations of vocalizations of user 600, tones associated with the vocalizations of user600, changes associated with the vocalizations of user 600, weather within an environment of the user600, movement of the user 600, eating / drinking / walking / sleeping / etc. habits associated with the user600, frequency, spectral / cepstral cues, vocal quality / stability, prosody, temporal output, and amplitude stability, motor processes, and so on. In an embodiment, the biomarkers 604 may be acoustic-based or non-acoustic based. Moreover, in an embodiment, the biomarkers 604 may be based at least in part on the user 600 and / or the use of the wellness and healthcare monitoring device 602. For example, if the wellness and healthcare monitoring device 602 is being employed to treat, manage, etc. a certain disease, the wellness and healthcare monitoring device 602 may be employed to determine the presence, or absence, of certain biomarkers 604 for the disease.

[0253] In an embodiment, the ML model(s) 636 may be used to process the audio data 660 and / or the sensor data 668 for determining the biomarkers 604. For example, the ML model(s) 636 may be previously trained to determine the biomarkers 604. In an embodiment, the ML model(s) 636 may be selected from a plurality of ML model(s) 636 based at least in part on the user 600 (e.g., age, medical history, etc.) and / or the welfare, behavioral, and / or pathological states 606 the wellness and healthcaremonitoring device 602 is being employed to monitor. In this sense, the wellness and healthcare monitoring device 602 may be tailored, programmed, etc. according to the specifics of the user 600 or caretaker / healthcare provider in order more accurately detect the presence, state, condition, etc. of certain diseases.

[0254] In an embodiment, the ML model(s) 636 may receive the audio data 660 and / or the sensor data 668 as an input, and output an indication associated with whether the audio data 660 and / or the sensor data 668 are indicative of the biomarkers 604. In an embodiment, the ML model(s) 636 may output a score associated with the biomarkers 604, where the score may indicate likelihood or probability of the audio data 660 and / or the sensor data 668 being indicative of the biomarkers 604. In an embodiment, the ML model(s) 636 may output a score associated with the welfare, behavioral, and / or pathological states 606 of the user 600. Moreover, in an embodiment, the ML model(s) 636 may determine the fingerprint(s) 628 for comparison to reference fmgerprint(s) / phenotype(s) associated with the welfare, behavioral, and / or pathological states 606, for determining whether the vitals, behavior, etc. of the user 600 is indicative of the welfare, behavioral, and / or pathological states 606.

[0255] At 808, the process 800 may include determining whether the biomarker(s) satisfy predetermined threshold(s). For example, the biomarkers 604 may be compared to reference biomarker(s), or characteristic(s) that are indicative of the welfare, behavioral, and / or pathological states 606. Any number of the biomarkers 604 may be considered when determining whether the biomarkers 604 are indicative of the welfare, behavioral, and / or pathological states 606. In an embodiment, the ML model(s) 636 may be used to determine whether the biomarkers 604 are indicative of satisfying threshold(s) associated with the welfare, behavioral, and / or pathological states 606. In an embodiment, score(s) output by the ML model(s) 636 may be compared to a threshold score for determining whether the score satisfies certain thresholds associated with the welfare, behavioral, and / or pathological states 606. Moreover, as referenced above, in an embodiment, the fingerprint(s) 628 may be determined for comparison with one or more reference fingerprints to determine the welfare, behavioral, and / or pathological states 606 of the user 600. If, at 808, the process 800 determines that the biomarkers 604 do not satisfy the threshold(s), the process 800 may follow the “NO” route and proceed to 810.

[0256] At 810, the process 800 may include causing output of one or more first notification(s). For example, the I / O component(s) 634 of the wellness and healthcare monitoring device 602 may output notification(s) 630, where the notification(s) 630 may indicate that the audio data 660 and / or the sensor data 668 is not associated with a particular welfare, behavioral, and / or pathological state 606. In anembodiment, the notification(s) 630 may be output at the wellness and healthcare monitoring device 602, and / or the mobile device 632 of the caregiver 608. From 810, the process 800 may proceed to 802, whereby the process 800 may continue to receive the audio data 660 and / or the sensor data 668 for determining the welfare, behavioral, and / or pathological states 606.

[0257] Returning to 808, if the process 800 determines that the biomarker(s) 604 satisfy the threshold(s), the process 800 may follow the “YES” route and proceed to 812. At 812, the process 800 may include causing output of one or more second notification(s). For example, the I / O component(s) 634 of the wellness and healthcare monitoring device 602 may output notification(s) 630, where the notification(s) 630 may indicate the welfare, behavioral, and / or pathological states 606.

[0258] At 814, the process 800 may include sending at least a portion of the audio data and / or the sensor data to a second device. For example, the wellness and healthcare monitoring device 602 may send the audio data 660 and / or the sensor data 668 to the mobile device 632 and / or the remote computing resource(s) 640. In an embodiment, the mobile device 632 and / or the remote computing resource(s) 640 may further process the audio data 660 and / or the sensor data 668, may store / log the audio data 660 and / or the sensor data 668 to maintain a history of the user 600. The audio data 660 and / or the sensor data 668 may be used to by the mobile device 632 and / or the remote computing resource(s) 640 to retrain the ML model(s) 636, confirm the biomarker(s) 604 as determined by the wellness and healthcare monitoring device 602, and / or the welfare, behavioral, and / or pathological states 606 determined by the wellness and healthcare monitoring device 602. In addition, the mobile device 632 may output the notification(s) 630 indicative of the welfare, behavioral, and / or pathological states 606. From 814, the process 800 may loop to 802, whereby the wellness and healthcare monitoring device 602 may continue to process and analyze the audio data 660 and the sensor data 668 associated with the user 600.

[0259] While various examples and embodiments are described individually herein, the examples and embodiments may be combined, rearranged, and modified to arrive at other variations within the scope of this disclosure.

[0260] Although the subject matter has been described in language specific to structural features and / or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described. Rather, the specific features and acts are disclosed as illustrative forms of implementing the claims.

Claims

CLAIMSWHAT IS CLAIMED IS:

1. A wearable speech therapy device to be worn by a user, the wearable speech therapy device comprising: a housing including a first end and a second end spaced apart from the first end; a first microphone located closer to the first end than the second end; a second microphone located closer to the second end than the first end; a speaker; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the wearable speech therapy device to perform acts comprising: receiving, from the first microphone, first audio data associated with audio captured within an environment, receiving, from the second microphone, second audio data associated with the audio, determining, based at least in part on the first audio data and the second audio data, that the audio is associated with user speech of the user within the environment, determining, based at least in part on the audio being associated with the user speech, one or more biomarkers associated with the user speech, determining that the one or more biomarkers satisfy a threshold associated with the user speech containing hypophonia, and causing, based at least in part on the one or more biomarkers satisfying the threshold, output of a notification via the speaker.

2. The wearable speech therapy device of claim 1, the acts further comprising: receiving, from the first microphone, third audio data associated with second audio captured within the environment; receiving, from the second microphone, fourth audio data associated with the second audio; determining, based at least in part on the third audio data and the fourth audio data, that the second audio is associated with second user speech of the user;determining, based at least in part on the second audio being associated with the second user speech, one or more second biomarkers associated with the second user speech; determining that the one or more second biomarkers fail to satisfy the threshold associated with the second user speech containing hypophonia; and causing, based at least in part on the one or more second biomarkers failing to satisfy the threshold, output of a second notification via the speaker, the second notification being different than the notification.

3. The wearable speech therapy device of claim 1, further comprising at least one of a lighting element or a haptic motor, the acts further comprising causing at least one of a second notification to be output on the at least one of the lighting element or the haptic motor.

4. The wearable speech therapy device of claim 1 , further compri sing a sensor, the acts further comprising receiving, from the sensor, sensor data, wherein: determining that the audio corresponds to the user speech is based at least in part on the sensor data; and determining the one or more biomarkers is based at least in part on the sensor data.

5. The wearable speech therapy device of claim 1 , wherein when the wearable speech therapy device is worn by the user, the first microphone and the second microphone are substantially aligned with an anatomical axis of the user.

6. The wearable speech therapy device of claim 1, wherein determining the one or more biomarkers is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify hypophonia, the first audio data and the second audio data; and receiving, as an output from the ML model, an indication associated with the one or more biomarkers.

7. The wearable speech therapy device of claim 1, wherein the one or more biomarkers include at least one of a pitch of the user speech, an intonation in the user speech, a tone associated withthe user speech, a pause in the user speech, a phonation associated with the user speech, or an amplitude of the user speech.

8. A speech therapy device comprising: one or more sensors; one or more output components; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructions hat, when executed by the one or more processors, cause the speech therapy device to perform acts comprising: receiving, from the one or more sensors, data, determining that the data is indicative of speech of a user, determining, based at least in part on the data being indicative of the speech, one or more biomarkers associated with the speech, the one or more biomarkers including at least an amplitude associated with the speech, determining that the one or more biomarkers fail to satisfy a threshold, and causing, based at least in part on the one or more biomarkers failing to satisfy the threshold, output of a notification via the one or more output components.

9. The speech therapy device of claim 8, wherein the one or more sensors comprise: at least one microphone; and at least one of an internal measurement unit (IMU), an accelerometer, a gyroscope, or a piezoelectric sensor.

10. The speech therapy device of claim 8, wherein the one or more output components comprise at least one of: a lighting element; a speaker; or a haptic motor.

11. The speech therapy device of claim 8, the acts further comprising:receiving, from the one or more sensors, second data; determining that the second data is indicative of second speech of the user; determining, based at least in part on the second data being indicative of the second speech, one or more second biomarkers associated with the second speech, the one or more second biomarkers including at least a second amplitude associated with the second speech; determining that the one or more second biomarkers satisfy the threshold associated with hypophonia; and causing, based at least in part on the one or more second biomarkers satisfying the threshold, output of a second notification via the one or more output components, the second notification being different than the notification.

12. The speech therapy device of claim 11, the acts further comprising: receiving, from the one or more sensors, third data; determining that the third data is indicative of third speech of the user; determining, based at least in part on the third data being indicative of the third speech, one or more third biomarkers associated with the third speech, the one or more third biomarkers including at least a third amplitude associated with the third speech; determining that the one or more third biomarkers fail to satisfy the threshold associated with hypophonia; and causing, based at least in part on the one or more third biomarkers failing to satisfy the threshold, output of the notification via the one or more output components.

13. The speech therapy device of claim 8, wherein the one or more biomarkers further include at least one of a pitch of the speech, an intonation in the speech, a tone associated with the speech, a pause in the speech, or a phonation associated with the speech.

14. The speech therapy device of claim 8, wherein determining the one or more biomarkers is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify hypophonia data; and receiving, as an output from the ML model, an indication associated with the one or more biomarkers.

15. The speech therapy device of claim 8, further comprising a housing including a first end and a second end, wherein: the one or more sensors include a first microphone and a second microphone, the first microphone being locater closer to the first end as compared to the second microphone, the second microphone being located closer to the second end as compared to the first microphone; and when the speech therapy device is worn by the user, the first microphone is located closer to a mouth of the user as compared to the second microphone.

16. A speech therapy device configured to be worn by a user, the speech therapy device comprising: a first microphone; a second microphone; one or more output components; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructionshat, when executed by the one or more processors, cause the speech therapy device to perform acts comprising: receiving, from the first microphone, first audio data associated with a sound captured in an environment, receiving, from the second microphone, second audio data associated with the sound, determining, based at least in part on the first audio data and the second audio data, that the sound is associated with user speech of the user, determining, based at least in part on the sound being associated with the user speech, one or more characteristics associated with the user speech, determining that the one or more characteristics are indicative of hypophonia, and causing, based at least in part on the one or more characteristics being indicative of hypophonia, output of a notification via the one or more output components.

17. The speech therapy device of claim 16, wherein the one or more output components comprise at least one of a speaker, a lighting element, or a haptic motor.

18. The speech therapy device of claim 16, further comprising one or more sensors that include at least one of an accelerometer, a gyroscope, an internal measurement unit (IMU), or a piezoelectric sensor, the acts further comprising receiving, from the one or more sensors, data, wherein: determining that the sound is associated with the user speech is based at least in part on the data; and determining the one or more characteristics associated with the user speech is based at least in part on the data.

19. The speech therapy device of claim 16, the acts further comprising: receiving, from the first microphone, third audio data associated with a second sound captured inhe environment; receiving, from the second microphone, fourth audio data associated with the second sound, determining, based at least in part on the third audio data and the fourth audio data, that the second sound is associated with second user speech of the user, determining, based at least in part on the second sound being associated with the second user speech of the user, one or more second characteristics associated with the second user speech; determining that the one or more second characteristics are not indicative of hypophonia; and causing, based at least in part on the one or more second characteristics not being indicative of hypophonia, output of a second notification via the one or more output components, the second notification being different than the notification.

20. The speech therapy device of claim 16, wherein the one or more characteristics comprise at least one of a pitch of the user speech, an intonation in the user speech, a tone associated with the user speech, a pause in the user speech, a phonation associated with the user speech, an amplitude of the user speech.

21. A device to be worn by an animal, the device comprising: one or more microphones; an output component; a network interface; one or more processors; andone or more non-transitory computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the device to perform acts comprising: receiving, from the one or more microphones, audio data associated with sound captured within an environment of the animal, determining, based at least in part on the audio data, that the sound is associated with the animal, determining, based at least in part on the sound being associated with the animal, one or more biomarkers associated with the sound, determining a threshold associated with the animal having a behavioral or pathological state, determining, based at least in part on the one or more biomarkers satisfying the threshold, that the sound is indicative of the animal having the behavioral or pathological state, causing, based at least in part on the one or more biomarkers satisfying the threshold, output of a first notification on the output component, and sending data associated with a second notification to be output on an electronic device associated with a caregiver of the animal.

22. The device of claim 21, further comprising a sensor, the acts further comprising receiving, from the sensor, sensor data, wherein: determining that the sound is associated with the animal is based at least in part on the sensor data; and determining the one or more biomarkers is based at least in part on the sensor data.

23. The device of claim 21, the acts further comprising: receiving, from the one or more microphones, second audio data associated with a second sound captured within the environment; determining, based at least in part on the second audio data, that the second sound is associated with the environment; and determining one or more second biomarkers associated with the second, wherein determining that the sound is indicative of the animal having the pathological state is based at least in part on the one or more second biomarkers.

24. The device of claim 21, the acts further comprising: determining an audio signature associated with the pathological state; determining, based at least in part on the one or more biomarkers, an audio signature associated with the sound; and determining a similarity between the audio signature associated with the pathological state and the audio signature associated with the sound, wherein determining that the sound is indicative of the animal having the pathological state is based at least in part on the similarity.

25. The device of claim 21, further comprising one or more attachment mechanisms for coupling the device to the animal.

26. The device of claim 21, wherein determining the one or more biomarkers is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify the pathological state, the audio data; and receiving, as an output from the ML model, an indication associated with the one or more biomarkers.

27. The device of claim 21, wherein the one or more biomarkers include at least one of a pitch of the sound, a tone associated with the sound, a pause in the sound, a duration of the sound, or an amplitude of the sound.

28. A device comprising: one or more sensors; one or more output components; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the device to perform acts comprising: receiving, from the one or more sensors, data,determining that the data is indicative of sound generated from an animal, determining, based at least in part on the data being indicative the sound generated from the animal, one or more biomarkers associated with the sound, the one or more biomarkers including at least an amplitude associated with the sound, determining that the one or more biomarkers fail to satisfy a threshold associated with a pathological state of the animal, and causing, based at least in part on the one or more biomarkers failing to satisfy the threshold, output of a notification via the one or more output components.

29. The device of claim 28, wherein the one or more sensors comprise at least one of: at least one microphone; or at least one of an internal measurement unit (IMU), an accelerometer, a gyroscope, or a piezoelectric sensor.

30. The device of claim 28, wherein the one or more output components comprise at least one of a lighting element or a speaker.

31. The device of claim 28, the acts further comprising sending a second notification to an electronic device.

32. The device of claim 28, wherein the device is configured to be worn by the animal.

33. The device of claim 28, the acts further comprising: receiving, from the one or more sensors, second data; determining that the second data is indicative of second sound generated from the animal; determining, based at least in part on the second data being indicative the second sound generated from the animal, one or more second biomarkers associated with the second sound; determining that the one or more second biomarkers satisfy the threshold, and causing, based at least in part on the one or more second biomarkers satisfying the threshold, output of a third notification via the one or more output components.

34. The device of claim 28, wherein determining the one or more biomarkers is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify the pathological state data; and receiving, as an output from the ML model, an indication associated with the one or more iomarkers.

35. The device of claim 28, wherein the one or more sensors comprise at least one of a microphone, an accelerometer, an inertial measurement unit (IMU), a GPS, or a temperature sensor.

36. A device comprising: a first microphone; a second microphone; one or more output components; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructionshat, when executed by the one or more processors, cause the device to perform acts comprising: receiving, from the first microphone, first audio data associated with a sound captured in an environment, receiving, from the second microphone, second audio data associated with the sound, determining, based at least in part on the first audio data and the second audio data, that the sound is associated with an animal, determining, based at least in part on the sound being associated with the animal, one or more characteristics associated with the sound, determining, based at least in part on the one or more characteristics, that the sound is indicative of a pathological state, and causing, based at least in part on the one or more characteristics being indicative of the pathological state, output of a notification via the one or more output components.

37. The device of claim 36, the acts further comprising sending data associated with the one or more characteristics to an electronic device.

38. The device of claim 36, further comprising one or more sensors that include at least one of an accelerometer, a gyroscope, an internal measurement unit (IMU), or a piezoelectric sensor, the acts further comprising receiving, from the one or more sensors, sensor data, wherein: determining that the sound is associated with the animal is based at least in part on the sensor data; and determining the one or more characteristics associated with the sound is based at least in part onhe sensor data.

39. The device of claim 36, the acts further comprising determining one or more characteristics associated with an environment of the animal, and wherein determining that the sound is indicative of the pathological state is based at least in part on the one or more characteristics associated with the environment.

40. The device of claim 36, wherein determining the one or more characteristics is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify the pathological state data; and receiving, as an output from the ML model, an indication associated with the one or more characteristics.

41. A device to be worn by a user, the device comprising: one or more microphones; an output component; a network interface; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructionshat, when executed by the one or more processors, cause the device to perform acts comprising: receiving, from the one or more microphones, audio data associated with sound captured within an environment of the user, determining, based at least in part on the audio data, that the sound is associated with the user,determining, based at least in part on the sound being associated with the user, one or more biomarkers associated with the sound, determining a threshold associated with the user having a pathological or behavioral state, process, or change, determining, based at least in part on the one or more biomarkers satisfying the threshold, that the sound is indicative of the user having the pathological or behavioral state, process, or change, causing, based at least in part on the one or more biomarkers satisfying the threshold, output of a first notification on the output component, and sending data associated with a second notification to be output on an electronic device associated with a caregiver of the user.

42. The device of claim 41, further comprising a sensor, the acts further comprising receiving, from the sensor, sensor data, wherein: determining that the sound is associated with the user is based at least in part on the sensor data; and determining the one or more biomarkers is based at least in part on the sensor data.

43. The device of claim 41, the acts further comprising: receiving, from the one or more microphones, second audio data associated with a second sound captured within the environment; determining, based at least in part on the second audio data, that the second sound is associated ith the environment; and determining one or more second biomarkers associated with the second sound, wherein determining that the sound is indicative of the user having the pathological or behavioral state is based at least in part on the one or more second biomarkers.

44. The device of claim 41, the acts further comprising: determining a signature associated with the pathological or behavioral state; determining, based at least in part on the one or more biomarkers, an audio signature associated ith the sound; anddetermining a similarity between the audio signature associated with the pathological or behavioral state and the audio signature associated with the sound, wherein determining that the sound is indicative of the user having the pathological or behavioral state is based at least in part on the similarity.

45. The device of claim 41, further comprising one or more attachment mechanisms for coupling the device to the user.

46. The device of claim 41, wherein determining the one or more biomarkers is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify the pathological or behavioral state, the audio data; and receiving, as an output from the ML model, an indication associated with the one or more biomarkers.

47. The device of claim 41, wherein the one or more biomarkers include at least one of a pitch of the sound, a tone associated with the sound, a pause in the sound, a duration of the sound, or an amplitude of the sound.

48. A device comprising: one or more sensors; one or more output components; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructions that, when executed by the one or more processors, cause the device to perform acts comprising: receiving, from the one or more sensors, data, determining that the data is indicative of sound generated from a user, determining, based at least in part on the data being indicative the sound generated from the user, one or more biomarkers associated with the sound, the one or more biomarkers including at least an amplitude associated with the sound,determining that the one or more biomarkers fail to satisfy a threshold associated with a pathological or behavioral state or change of the user, and causing, based at least in part on the one or more biomarkers failing to satisfy the threshold, output of a notification via the one or more output components.

49. The device of claim 48, wherein the one or more sensors comprise at least one of: a microphone; or at least one of an internal measurement unit (IMU), an accelerometer, a gyroscope, or a piezoelectric sensor.

50. The device of claim 48, wherein the one or more output components comprise at least one of a lighting element or a speaker.

51. The device of claim 48, the acts further comprising sending a second notification to an electronic device.

52. The device of claim 48, wherein the device is configured to be worn by the user.

53. The device of claim 48, the acts further comprising: receiving, from the one or more sensors, second data; determining that the second data is indicative of second sound generated from the user; determining, based at least in part on the second data being indicative the second sound generated from the user, one or more second biomarkers associated with the second sound; determining that the one or more second biomarkers satisfy the threshold, and causing, based at least in part on the one or more second biomarkers satisfying the threshold, output of a third notification via the one or more output components.

54. The device of claim 48, wherein determining the one or more biomarkers is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify the pathological or ehavioral state; andreceiving, as an output from the ML model, an indication associated with the one or more biomarkers.

55. The device of claim 48, wherein the one or more sensors comprise at least one of a microphone, an accelerometer, an inertial measurement unit (IMU), a GPS, or a temperature sensor.

56. A device comprising: a first microphone; a second microphone; one or more output components; one or more processors; and one or more non-transitory computer-readable media storing computer-executable instructionshat, when executed by the one or more processors, cause the device to perform acts comprising: receiving, from the first microphone, first audio data associated with a sound captured in an environment, receiving, from the second microphone, second audio data associated with the sound, determining, based at least in part on the first audio data and the second audio data, that the sound is associated with a user, determining, based at least in part on the sound being associated with the user, one or more characteristics associated with the sound, determining, based at least in part on the one or more characteristics, that the sound is indicative of a pathological or behavioral state of the user, and causing, based at least in part on the one or more characteristics being indicative of the pathological or behavioral state, output of a notification via the one or more output components.

57. The device of claim 56, the acts further comprising sending data associated with the one or more characteristics to an electronic device.

58. The device of claim 56, further comprising one or more sensors that include at least one of an accelerometer, a gyroscope, an internal measurement unit (IMU), or a piezoelectric sensor, the acts further comprising receiving, from the one or more sensors, sensor data, wherein:determining that the sound is associated with the user is based at least in part on the sensor data; and determining the one or more characteristics associated with the sound is based at least in part on the sensor data.

59. The device of claim 56, the acts further comprising determining one or more characteristics associated with an environment of the user, and wherein determining that the sound is indicative of the pathological or behavioral state is based at least in part on the one or more characteristics associated with the environment.

60. The device of claim 56, wherein determining the one or more characteristics is based at least in part on: providing, as an input to a machine-learned (ML) model trained to identify the pathological or behavioral state; and receiving, as an output from the ML model, an indication associated with the one or more characteristics.

Images