Sensor system and method for characterizing health conditions

Abstract

The sensing system includes a handheld sensing device having a vibroacoustic sensor module (VSM) comprising: a voice coil component including a coil holder supporting a winding; a magnet component comprising a magnet supported by a frame, the magnet gap configured to receive at least a portion of the voice coil component in a spaced-apart, movable manner; a connector connecting the voice coil component to the magnet component, the connector being compliant and allowing relative movement of the voice coil component and the magnet component; a diaphragm configured to induce movement of the voice coil component within the magnet gap in response to incident acoustic waves; and a housing for holding the vibroacoustic sensor module having a handle end and a sensor end, the sensor end having an opening, the VSM positioned so that at least a portion of the diaphragm extends across the opening.

Description

[Technical Field]

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS This application claims the benefit of U.S. Provisional Patent Application No. 63 / 022,362, filed May 8, 2020, U.S. Provisional Patent Application No. 63 / 022,336, filed May 8, 2020, U.S. Provisional Patent Application No. 63 / 075,056, filed September 4, 2020, U.S. Provisional Patent Application No. 63 / 075,059, filed September 4, 2020, U.S. Provisional Patent Application No. 63 / 067,179, filed August 18, 2020, and U.S. Patent Application No. 17 / 096,806, filed November 12, 2020. The contents of the foregoing applications are incorporated herein by reference in their entireties.

[0002] The present invention relates generally to the field of monitoring and diagnostics, such as for identifying the health or other physical condition of a subject using active and / or passive sensing techniques. [Background technology]

[0003] Traditionally, healthcare professionals have utilized a suite of instruments to help assess a patient's biological characteristics, with each instrument specialized in a particular biometric or class of biometric. However, the suite of instruments required for a holistic and comprehensive patient assessment suffers from problems such as complexity, difficulty learning to use them properly, cost, and a relative lack of portability and data interoperability.

[0004] Furthermore, some conventional instruments have limited functionality, which contributes to an incomplete picture of a patient's health. For example, medical professionals have traditionally used tools such as stethoscopes to listen to and observe a patient's body sounds, such as sounds generated by the heart, lungs, and gastrointestinal system. However, conventional stethoscopes cannot help medical professionals observe specific cardiac, respiratory, and / or gastrointestinal information encoded in various low- and high-frequency (within and above the limits of human hearing), low-amplitude, inaudible sound signals. Furthermore, how to analyze such signals to assess patient health is currently poorly understood. Therefore, using conventional technology, low- and high-frequency (within and above the limits of human hearing), low-amplitude indicators of a patient's health are not detected or considered in conventional medical practice, resulting in a non-comprehensive diagnostic picture of the patient.

[0005] Furthermore, conventional stethoscopes require contact with the patient's skin for adequate signal detection, thereby limiting their use when signal detection must occur through clothing for a variety of reasons, such as risk of contamination, modesty, or emergency situations. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] U.S. Patent No. 8,923,956 [Patent Document 2] U.S. Patent No. 8,860,401 [Patent Document 3] U.S. Patent No. 8,264,246 [Patent Document 4] U.S. Patent No. 8,264,247 [Patent Document 5] U.S. Patent No. 8,054,061 [Patent Document 6] U.S. Patent No. 7,885,700 Summary of the Invention [Problem to be solved by the invention]

[0007] Therefore, there is a need for new and improved sensor platforms for characterizing one or more health and other physical conditions of a subject. [Means for solving the problem]

[0008] Generally, in some variations, the present technology provides a system including a sensor platform. The sensor platform may include a sensing device, such as a vibroacoustic sensor module including one or more sensors configured to detect a vibroacoustic signal; a signal processing system configured to extract a vibroacoustic signal component originating from a subject from the detected vibroacoustic signal; and at least one processor configured to characterize a physical state of the subject based at least in part on the extracted vibroacoustic signal component, e.g., using a machine learning model. In some variations, the physical state of the subject may include a health state of the subject. For example, the signal processing system may be configured to extract a bio-vibroacoustic signal component originating from a living subject, and the at least one processor may be configured to characterize a health state or other physical state of the subject based at least in part on the extracted vibroacoustic bio-signal component. Additionally or alternatively, the physical state of the subject may include another suitable physical state (e.g., a structural condition) of a living or non-living subject.

[0009] Additionally, in some variations, a method for characterizing a physical state may include detecting a periodic or aperiodic vibroacoustic signal with a vibroacoustic sensor module, the vibroacoustic sensor module including multiple sensors; extracting a vibroacoustic signal component originating from the subject from the detected vibroacoustic signal; and characterizing the physical state of the subject based at least in part on the extracted vibroacoustic signal component using a machine learning model. In some variations, the physical state of the subject may include a health state of the subject. For example, the method may include extracting a biovibroacoustic signal component from a vibroacoustic signal originating from a living subject and characterizing the health state of the subject based at least in part on the extracted biovibroacoustic signal component using a machine learning model. Additionally or alternatively, the physical state of the subject may include another suitable physical state (e.g., structural condition) of a living or non-living subject.

[0010] Further, in some variations, a vibro-acoustic sensor module may include one or more sensors configured to detect vibro-acoustic signals and one or more deflection structures interacting with one or more of the sensors, wherein the vibro-acoustic sensor module has a bandwidth of from about 0.01 Hz to at least about 160 kHz, or from about 0.01 Hz to at least about 50 kHz, with low to ultra-low amplitudes of oscillations between 0.01 and 50 μm, and a frequency response of about 10 -4 ms -2 Vibration of the skin and tracheobronchial tree with accelerations of approximately 0.1 to 50 m·s -2 There are even large-scale whole-body exercises.

[0011] Advantageously, in some variations, systems and methods are provided for remotely diagnosing or monitoring a subject's physical condition in a non-invasive manner. Direct skin contact is not required, and variations of the systems and methods can operate through clothing. In some variations, the systems and methods can operate at distances of approximately 1 mm, 2 mm, 1 cm, 10 cm, 1 meter, 2 meters, 3 meters, 4 meters, 5 meters, 6 meters, 7 meters, 8 meters, 9 meters, or 10-50 meters from the subject.

[0012] In some modifications, the systems and methods may be suitable for screening for susceptible conditions, such as those caused by coronaviruses (e.g., Covid-19). Current COVID-19 screening approaches are either simple and rapid but lack accuracy (e.g., temperature checks) or accurate but neither simple nor rapid (e.g., antibody screening). The gold standard for COVID-19 diagnosis is real-time reverse transcriptase polymerase chain reaction (RT-qPCR), but RT-PCR testing is limited to individuals with obvious symptoms in some circumstances, and there is often a significant delay between testing and reporting results, providing unknown opportunities for infection spread. COVID-19 IgG antibody, or serology, tests can quickly and inexpensively provide information about past infection. However, antibodies can take days to weeks to develop, and the duration of their effect remains unknown. Other screening approaches, such as those currently employed in many schools, daycares, hospitals, and other public settings, rely on temperature scans for fever and self-reported cough and fatigue. However, these approaches focus on nonspecific symptoms that may not appear for many days after infection. Therefore, current screening approaches are impractical and inconvenient, do not identify individuals in the early stages of infection, cannot distinguish between COVID-19 and multisystem inflammatory syndrome in children (MIS-C), and do not consider common comorbidities, including influenza and pneumonia, respiratory failure, hypertension, diabetes, and cardiopulmonary failure, which can worsen the course and outcome of the disease.

[0013] In the context of this specification, unless otherwise specified, a physical condition may refer to one or more of the following: a viral infection of a subject, a bacterial infection of a subject, a cognitive status of a subject, a reportable disease, a fracture, a laceration, an embolism, a blood clot, swelling, an obstruction, a prolapse, a hernia, a dissection, an infarction, a stenosis, a hematoma, an edema, a contusion, osteopenia, and the presence of a foreign body in a subject's body, such as an improvised explosive device (IED), a surgically implanted improvised explosive device (SIIED), and / or a body cavity bomb (BCB). Examples of viral infections include, but are not limited to, infections of the coronavirus family (e.g., COVID-19, SARS). A reportable disease is one that is considered to be of public health importance. In the United States, local, state, and national agencies (e.g., county and state health departments, the U.S. Centers for Disease Control and Prevention) require that these diseases be reported when diagnosed by a physician or laboratory.Diseases that are reportable to the CDC include anthrax, arboviral diseases (diseases caused by viruses spread by mosquitoes, sandflies, ticks, etc.) such as West Nile virus, Eastern and Western equine encephalitis, babesiosis, botulism, brucellosis, campylobacteriosis, chancroid, chickenpox, chlamydia, cholera, coccidioidomycosis, cryptosporidiosis, cyclosporiasis, dengue virus infection, diphtheria, Ebola hemorrhagic fever, ehrlichiosis, food poisoning outbreaks, giardiasis, gonorrhea, influenza, invasive diseases, hantavirus pulmonary syndrome, hemolytic uremic syndrome, post-diarrheal, hepatitis A, hepatitis B, hepatitis C, HIV infection, influenza-associated infant deaths, invasive pneumococcal disease, elevated blood lead levels, Legionnaires' disease, leprosy, leptospirosis, listeriosis, Lyme disease, and malaria. , measles, meningitis (meningococcal disease), mumps, emerging infectious disease A virus, whooping cough, pesticide-related illnesses and injuries, plague, polio, poliovirus infections, non-paralytic, psittacosis, Q fever, rabies (human and animal cases), rubella (including congenital syndromes), Salmonella paratyphi and Salmonella typhi infections, salmonellosis, severe acute respiratory syndrome-associated coronavirus disease, Shiga toxin-producing Escherichia coli (STEC), shigellosis , smallpox, syphilis including congenital syphilis, tetanus, toxic shock syndrome (non-streptococcal), trichinellosis, tuberculosis, tularemia, typhoid, vancomycin-intermediate Staphylococcus aureus (VISA), vancomycin-resistant Staphylococcus aureus (VRSA), vibriosis, viral hemorrhagic fevers (including Ebola virus, Lassa virus, among others), waterborne disease outbreaks, yellow fever, Zika virus disease and infectious diseases (including congenital).

[0014] Nearly all diseases affect livestock, poultry, fish, and other animals, adversely affecting the quality and quantity of food and other products such as hides, bones, fiber, wool, and animal power for tillage, transportation, and traction. The decline in animal production, productivity, and profitability caused by transboundary animal diseases (TADs) impacts human livelihoods. In this rapidly globalizing scenario, TADs pose a serious threat to the economy and welfare of the population and affected countries by significantly reducing production and productivity, disrupting trade, travel, and local and national economies, and threatening human health through food quality deterioration and zoonotic / infectious diseases. TADs are of great concern due to their economic importance, zoonotic nature, and the growing threat of new TADs posing a national security risk. Examples of diseases that affect animals and to which embodiments of the present technology are relevant include arbovirus, avian influenza, B virus, brucellosis, campylobacteriosis, cat scratch disease, cryptococcosis, cyanobacteria, Escherichia coli, fish tank granulomatosis, Giardiasis, hantavirus, histoplasmosis, leptospirosis, listeriosis, Lyme disease, lymphocytic choriomeningitis, MRSA, plague, psittacosis, Q fever (caused by Coxiella burnetti), rabies, raccoon roundworm, rat-bite fever, ringworm, roundworms, salmonellosis, tick-borne fever, toxoplasmosis, tularemia, valley fever (coccidioidomycosis), West Nile virus, yellow fever, Zika virus, and the like. In some embodiments, the technology can work through fur, mud, feces, or other obstacles commonly found in animal facilities.

[0015] Rapid, accurate, sensitive, and specific, non-invasive, and easily scalable mass screening tests for TADs should be developed to safely manage and avoid bottlenecks in the face of faster trade, new trade routes, globalization, intensive animal production to meet growing demand for animal protein, the effects of changing forest ecosystems, climate change, global warming, and microbial evolution, and the impact of increasing conflict and insecurity.

[0016] In the context of this specification, unless otherwise specified, "remote screening" means that the subject is not in direct contact with at least the sensor module components of the system of the present invention. Remote screening includes situations where some components of the system are remote from the subject. There is no limit to the distance of the separation. Remote screening includes "on-garment" and / or "through-garment" signal detection.

[0017] In the context of this specification, unless otherwise specified, an animal refers to an individual animal that is a mammal, a bird, or a fish. In particular, a mammal refers to human and non-human vertebrates that belong to the taxonomic class Mammalia. Non-exclusive examples of non-human mammals include companion animals and livestock. An animal in the context of this disclosure is understood to include vertebrates. The term vertebrate in this context is understood to include, for example, fish, amphibians, reptiles, birds, and mammals, including humans. As used herein, the term "animal" may refer to mammals and non-mammals, such as birds or fish. In the case of a mammal, it may be a human or non-human mammal. Non-human mammals include, but are not limited to, livestock animals and companion animals.

[0018] In the context of this specification, unless otherwise specified, a computer system may refer to, but is not limited to, an "electronic device," an "operating system," a "communication system," a "system," a "computer-based system," a "controller unit," a "control device," and / or any combination thereof appropriate to the task at hand.

[0019] In the context of this specification, unless otherwise specified, the expressions "computer-readable medium" and "memory" are intended to include media of any nature and type, non-limiting examples of which include RAM, ROM, disks (CD-ROM, DVD, floppy disk, hard disk drive, etc.), USB keys, flash memory cards, solid state drives and tape drives.

[0020] In the context of this specification, a "database" is any structured collection of data, regardless of the particular structure, database management software, or computer hardware on which the data is stored, implemented, or otherwise made available. A database may be located on the same hardware as the processes that store or use the information stored in the database, or may be located on separate hardware, such as a dedicated server or multiple servers.

[0021] In the context of this specification, unless otherwise specified, the words "first," "second," "third," etc. are used as adjectives only to enable the nouns they modify to be distinguished from one another, and not to describe a particular relationship between those nouns.

[0022] In the context of this specification, vibroacoustics refers to vibrations or sound signals that propagate through air, biological structures, solids, gases, liquids, or other fluids. This term also encompasses the term mechanoacoustics.

[0023] In the context of this specification, a sensing device may be a sensor actuator, which may be thought of as a device configured to generate a sensor signal that is a function of its electrical response to an electrical input signal and its mechanical response.

[0024] Each variation of the present technology will have at least one, but not necessarily all, of the above-mentioned objects and / or aspects. It should be understood that some embodiments of the present technology, resulting from an attempt to achieve the above-mentioned object, may not meet that object and / or may meet other objects not specifically recited herein.

[0025] Additional and / or alternative features, aspects, and advantages of embodiments of the present technology will become apparent from the following description, the accompanying drawings, and the appended claims.

[0026] The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawing(s) will be provided by the Office upon request and payment of the necessary fee. [Brief explanation of the drawings]

[0027] [Figure 1A] FIG. 1 is a schematic diagram of a sensor system for characterizing the physical state of a subject. [Figure 1B] FIG. 1 is a schematic diagram of a sensor system for characterizing the physical state of a subject. [Figure 1C] FIG. 1 is a schematic diagram of a sensor system for characterizing the physical state of a subject. [Figure 1D] FIG. 1 is a schematic diagram of a sensor system for characterizing the physical state of a subject. [Figure 1E] 1 illustrates various types of vibro-acoustic data over a range of frequencies, energy distribution, and amplitudes that are related to the sensitivity of the human ear. [Figure 1F] 1 is a schematic diagram of various output interfaces of components of an exemplary variation of a sensor system for characterizing a physical state of a subject. [Figure 1G] 1 is a schematic diagram of various output interfaces of components of an exemplary variation of a sensor system for characterizing a physical state of a subject. [Figure 1H]1 is a schematic diagram of various output interfaces of components of an exemplary variation of a sensor system for characterizing a physical state of a subject. [Figure 2] 1 is a schematic diagram illustrating an exemplary variation of a vibroacoustic sensing device for characterizing a physical state of a subject. FIG. [Figure 3A] FIG. 1 is an assembly diagram of an exemplary variation of a sensing device for characterizing a physical state of a subject. [Figure 3B] FIG. 1 is an exploded view of an exemplary variation of a sensing device for characterizing a physical state of a subject. [Figure 4A] FIG. 10 is an assembly diagram of another exemplary variation of a sensing device for characterizing a physical state of a subject. [Figure 4B] FIG. 10 is an exploded view of another exemplary variation of a sensing device for characterizing a physical state of a subject. [Figure 4C] FIG. 4C is a perspective view from the sensor end of the sensing device of FIGS. 4A and 4B. [Figure 4D] FIG. 4C is a cross-sectional view through the sensing device of FIGS. 4A and 4B with some of the internal components omitted for clarity. [Figure 4E] FIG. 4C is a perspective view from the handle end of the sensing device of FIGS. 4A and 4B. [Figure 4F] FIG. 4C is an enlarged perspective view of the sensor end of the sensing device of FIGS. 4A and 4B with a portion of the housing cut away for clarity. [Figure 4G] FIG. 4F is an enlarged plan view of the sensor end of the sensing device of FIG. 4F. [Figure 4H] FIG. 10 is a perspective view illustrating another exemplary variation of a sensing device for characterizing a physical state of a subject. [Figure 4I] FIG. 4H is an enlarged plan view of the sensor end of a sensing device that is a variation of FIG. 4G. [Figure 5A] 1 is a perspective cross-sectional view of a portion of a sensing device for characterizing a physical state of a subject. [Figure 5B]1 is a cross-sectional view of a portion of a sensing device for characterizing a physical state of a subject. [Figure 5C] 10A is a top perspective view of an exemplary variation of a vibro-acoustic sensor module having a bent arm. FIG. [Figure 5D] FIG. 10 is a bottom perspective view of an exemplary variation of a vibro-acoustic sensor module having a bent arm. [Figure 5E] FIG. 5D is a cross-sectional view of the vibro-acoustic sensor module depicted in FIGS. 5C and 5D. [Figure 5F] FIG. 5E is an exploded view of the vibro-acoustic sensor module depicted in FIGS. 5C and 5D. [Figure 5G] 5D is a top perspective view of an exemplary variation of the vibro-acoustic sensor module having a bent arm shown in FIG. 5C. [Figure 5H] 5G illustrates an exemplary variation of the flex circuit in the vibro-acoustic sensor module depicted in FIG. 5G. [Figure 5I] 5G illustrates an exemplary variation of a deflection structure having a bent arm in the vibroacoustic sensor module depicted in FIG. 5G. [Figure 6A] 1 is a schematic diagram illustrating an example variation of a vibro-acoustic sensor module having a bent arm. [Figure 6B] 6B is a schematic diagram illustrating an example variation of the flex circuit in the vibro-acoustic sensor module depicted in FIG. 6A. [Figure 7A] 10A-10C are cross-sectional views illustrating exemplary variations of the vibroacoustic sensor module having different clearance distances between the deflection structure and the surface of the subject. [Figure 7B] 10A-10C are cross-sectional views illustrating exemplary variations of the vibroacoustic sensor module having different clearance distances between the deflection structure and the surface of the subject. [Figure 8A] FIG. 10 is a perspective view of a schematic diagram of an exemplary variation of a vibro-acoustic sensor module having a bending arm. [Figure 8B] FIG. 10 is a plan view of a schematic diagram of an exemplary variation of a vibro-acoustic sensor module having a bent arm. [Figure 9A] 10A is a top perspective view of an exemplary variation of a vibro-acoustic sensor module including a bent arm. FIG. [Figure 9B] FIG. 10 is a bottom perspective view of an exemplary variation of a vibro-acoustic sensor module including a bent arm. [Figure 9C] FIG. 9C is a schematic diagram illustrating an exemplary variation of a deflection structure having a bent arm in the vibro-acoustic sensor module depicted in FIGS. 9A and 9B. [Figure 9D] 9C is a schematic diagram illustrating an example variation of the flex circuit in the vibro-acoustic sensor module depicted in FIGS. 9A and 9B. FIG. [Figure 10A] 1A and 1B are perspective views illustrating exemplary variations of vibroacoustic sensor modules including a membrane and at least one cross-axis inertial sensor (e.g., an accelerometer). [Figure 10B] 1A-1C are cross-sectional views illustrating exemplary variations of vibroacoustic sensor modules including a membrane and at least one cross-axis inertial sensor (e.g., an accelerometer). [Figure 11A] 1A and 1B are cross-sectional views illustrating exemplary variations of vibro-acoustic sensor modules including a membrane and at least one cross-axis inertial sensor (e.g., a pressure sensor). [Figure 11B] 1A and 1B are cross-sectional exploded views illustrating example variations of vibro-acoustic sensor modules including a membrane and at least one cross-axis inertial sensor (e.g., a pressure sensor). [Figure 11C] 11C is a top view of an exemplary variation of a circuit board including a pressure sensor in the vibro-acoustic sensor module depicted in FIGS. 11A and 11B. FIG. [Figure 11D] 11C is a side view of an exemplary variation of a circuit board including a pressure sensor in the vibro-acoustic sensor module depicted in FIGS. 11A and 11B. FIG. [Figure 11E] 11C is a bottom view of an exemplary variation of a circuit board including a pressure sensor in the vibro-acoustic sensor module depicted in FIGS. 11A and 11B. FIG. [Figure 12A]10A and 10B are cross-sectional views of exemplary variations of vibro-acoustic sensor modules including bent arms. [Figure 12B] 1 is a cross-sectional exploded view of an exemplary variation of a vibro-acoustic sensor module including a bent arm. [Figure 13A] 1 is a perspective view of an exemplary variation of a vibro-acoustic sensor module including an accelerometer and a microphone. FIG. [Figure 13B] 1 is a cross-sectional view of an exemplary variation of a vibro-acoustic sensor module including an accelerometer and a microphone. [Figure 13C] FIG. 1 is an exploded view of an exemplary variation of a vibro-acoustic sensor module including an accelerometer and a microphone. [Figure 14] 10A-10C illustrate an exemplary variation of a vibro-acoustic sensor module including a handle portion. [Figure 15] 1A-1C illustrate exemplary variations of a vibroacoustic sensor module including a bending arm and multiple cross-axis inertial sensors (e.g., acceleration sensors, pressure sensors). [Figure 16A] 1 is an assembly diagram of an example variation of a vibro-acoustic sensor module including a voice coil with one or more spider layers. [Figure 16B] FIG. 16B is an exploded view of a vibro-acoustic sensor module having a single layer spider, such as the vibro-acoustic sensor module of FIG. 16A. [Figure 16C] FIG. 16B is an exploded view of a vibro-acoustic sensor module having a double layer spider, such as the vibro-acoustic sensor module of FIG. 16A. [Figure 16D] FIG. 16B is a perspective view of the exemplary vibro-acoustic sensor module of FIG. 16A with the outer housing removed for clarity. [Figure 16E] FIG. 16E is an exploded view of the vibroacoustic sensor module of FIG. 16D. [Figure 17A] FIG. 16B is a cross-sectional view of the example vibroacoustic sensor module of FIG. 16A. [Figure 17B] FIG. 16C is a cross-sectional view of the example vibroacoustic sensor module of FIG. 16B. [Figure 18A]16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18B] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18C] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18D] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18E] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18F] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18G] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18H] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18I] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18J]16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18K] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18L] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18M] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18N] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18O] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18P] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18Q] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18R] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18S]16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18T] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18U] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18V] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18W] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18X] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18Y] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18Z] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18AA] 16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 18AB]16A-16E and 17A-17B. FIG. 17A illustrates an exemplary spider for use with the exemplary variations of the vibroacoustic sensor module of any of FIGS. 16A-16E and 17A-17B. [Figure 19] FIG. 10 illustrates an example circuit of an ECG sensor module of an example variation of the sensing device. [Figure 20] FIG. 10 shows a comparison of results obtained with an exemplary modification of the ECG sensor module of the present technology and conventional wet electrodes for leads II and VL. [Figure 21] FIG. 1 illustrates an exemplary circuit for a driven right leg sensor used in prior art ECG procedures. [Figure 22A] FIG. 10 illustrates a prototype circuit diagram of a driving right leg (DRL) sensor for use in an exemplary variation of the sensing device. [Figure 22B] FIG. 10 depicts an experimental setup with five DRL electrodes spaced between and around two EPIC electrodes, with optional guard electrodes, for use in an exemplary variation of the sensing device. [Figure 22C] FIG. 22F shows experimental results of the gain and phase response of the DRL feedback circuit of FIG. 22E compared with simulations using various numbers of 0.58 mm fabric separators. [Figure 22D] FIG. 22F shows experimental results of the gain and phase response of the DRL feedback circuit of FIG. 22E compared with simulations using various numbers of 0.58 mm fabric separators. [Figure 22E] 13A-13C illustrate example variations of the DRL feedback circuit of the driving right leg sensor of the present technology. [Figure 22F] FIG. 1 shows a wet electrode bioimpedance analysis application diagram. [Figure 22G] FIG. 1 is a system diagram of an exemplary sensor for bioimpedance. [Figure 22H] FIG. 1 is a circuit diagram of an exemplary electro-potential integrated circuit (EPIC) sensor. [Figure 22I]10A-10C are heat map diagrams of two simulations of an EPIC sensor in direct skin contact for incorporation into an exemplary modification of the sensing device. [Figure 22J] 10A-10C are heat map diagrams of two simulations of an EPIC sensor in direct skin contact for incorporation into an exemplary modification of the sensing device. [Figure 22K] FIG. 10 shows an experimental setup for an exemplary variation of a sensing device for testing a bioimpedance sensor in which pairs of conductive fabric strips are connected by a resistor to form a repeatable "skin" resistance. [Figure 22L] FIG. 1 shows an alternative "skin" surrogate test configuration. [Figure 22M] FIG. 1 shows an alternative "skin" surrogate test configuration. [Figure 22N] FIG. 10 shows impedance versus frequency scans obtained using two different skin substitutes and an arm. [Figure 22O] FIG. 10 shows phase versus frequency scans acquired using two different skin substitutes and an arm. [Figure 22P] FIG. 1 shows impedance versus frequency response for a conductive paper skin substitute, direct contact, and various layers of fabric. [Figure 23] 1 is a schematic diagram of an exemplary variation of an electronics system within a sensing device. [Figure 24] FIG. 10 is a schematic diagram illustrating an exemplary variation of a signal processing chain for conditioning a vibroacoustic signal from a sensing device. [Figure 25A] FIG. 25 illustrates exemplary signal processing circuitry in the analog portion of the signal processing chain of FIG. 24. [Figure 25B] FIG. 25B shows gain and phase response versus frequency for the circuit of FIG. 25A. [Figure 25C] FIG. 10 illustrates a transfer function for another exemplary signal processing circuit. [Figure 26] 25 is a schematic diagram illustrating an exemplary modification of a portion of the signal processing circuitry in the signal processing chain shown in FIG. 24. [Figure 27A] FIG. 1 is an assembly diagram of an exemplary variation of a handheld sensing device for characterizing a subject's physical state. [Figure 27B] FIG. 1 is an exploded view of an exemplary variation of a handheld sensing device for characterizing a physical state of a subject. [Figure 28A] FIG. 1 is an assembly diagram of an exemplary variation of a handheld sensing device for characterizing a subject's physical state. [Figure 28B] FIG. 1 is an exploded view of an exemplary variation of a handheld sensing device for characterizing a physical state of a subject. [Figure 29] FIG. 1 is a schematic diagram illustrating an exemplary variation of a wearable sensing device for characterizing a subject's physical state. [Figure 30] FIG. 1 is a schematic diagram illustrating an exemplary variation of a stethoscope sensing device for characterizing a subject's physical condition. [Figure 31A] 1A and 1B are front and perspective views of an exemplary variation of a sensing device embodied as a panel. [Figure 31B] FIG. 31B is a cross-sectional view of the sensing device of FIG. 31A. [Figure 31C] FIG. 31B is an exploded view of the sensing device of FIG. 31A. [Figure 31D] FIG. 10 is a front perspective view of another exemplary variation of a sensing device embodied as a panel. [Figure 31E] FIG. 31E is an enlarged view of a portion of the sensing device of FIG. 31D. [Figure 32A] FIG. 10 is a perspective view of an exemplary variation of a sensing device embodied as a panel and including a base unit. [Figure 32B] 32B illustrates an exemplary modification of the sensing device of FIG. 32A. [Figure 33A] FIG. 10 illustrates an exemplary variation of a sensing device embodied as a gateway. [Figure 33B] 33B shows partially folded and fully folded examples of the sensing device of FIG. 33A. [Figure 33C] FIG. 33B shows the sensing device of FIG. 33A, including an output display unit. [Figure 33D] FIG. 10 is a schematic diagram of emitted and received signals in an exemplary variation of a sensing device including an echo Doppler sensor module, the sensing device comprising a single emitter and receiver. [Figure 33E] 1 is a schematic diagram of emitted and received signals in an exemplary variation of a sensing device including an echo Doppler sensor module, the sensing device comprising two emitters and a single receiver. FIG. [Figure 33F] 1 is a schematic diagram of emitted and received signals in an exemplary variation of a sensing device including an echo Doppler sensor module, the sensing device comprising two receivers and a single emitter. FIG. [Figure 33G] FIG. 1 is a schematic diagram of emitted and received signals in an exemplary variation of a sensing device including an echo Doppler sensor module, the sensing device comprising two receivers and two emitters. [Figure 33H] FIG. 33E illustrates a modification of the sensing device of FIG. 33D, including an exemplary loss. [Figure 34] 1 is a flowchart summarizing an exemplary variation of a method for characterizing a physical state of a subject. [Figure 35A] 10A-10C illustrate example sensor data using an example accelerometer-based variation of the vibrometer sensor module of the present technology. [Figure 35B] 10A-10C illustrate example sensor data using an example accelerometer-based variation of the vibrometer sensor module of the present technology. [Figure 36A] FIG. 10 shows an overview of classification model development in an example of training and testing a machine learning model for classification of fatigue and non-fatigue states of a subject in an exemplary modification of the system of the present technology. [Figure 36B]10A-10C illustrate sample heart sound data extracted as segments from a vibrometer waveform using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology. [Figure 36C] FIG. 10 shows a receiver operating characteristic curve (ROC) for a classification model that best fits the average area under the ROC value of data acquired using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology. [Figure 37A] FIG. 10 illustrates vibrometer data taken in the time domain at 3 Hz captured after applying a signal to a phantom using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology. [Figure 37B] FIG. 10 illustrates vibrometer data taken in the frequency domain at 3 Hz captured after applying a signal to a phantom using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology. [Figure 37C] FIG. 10 illustrates exemplary vibrometer data taken in the time domain at 10 Hz captured after applying a signal to a phantom using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology. [Figure 37D] FIG. 10 illustrates exemplary vibrometer data taken in the 10 Hz frequency domain captured after applying a signal to a phantom using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology. [Figure 38A] 10A-10C show exemplary vibrometer data captured after applying a signal to human skin using an exemplary variation of the sensing device of the exemplary variation of the system of the present technology through direct coupling; [Figure 38B] 10A-10C show exemplary vibrometer data captured after applying a signal to human skin using an exemplary variation of the sensing device of the exemplary variation of the system of the present technology through direct coupling; [Figure 39A]FIG. 10 illustrates exemplary vibrometer data captured after applying a signal to human skin using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology through a T-shirt. [Figure 39B] FIG. 10 illustrates exemplary vibrometer data captured after applying a signal to human skin using an exemplary variation of a sensing device of an exemplary variation of a system of the present technology through a T-shirt. [Figure 40A] FIG. 10 illustrates exemplary vibrometer data captured after applying a signal to human skin using an exemplary variation of the sensing device of an exemplary variation of the system of the present technology through a wool sweater. [Figure 40B] FIG. 10 illustrates exemplary vibrometer data captured after applying a signal to human skin using an exemplary variation of the sensing device of an exemplary variation of the system of the present technology through a wool sweater. [Figure 41A] FIG. 10 illustrates program output for an exemplary program for a classification task on exemplary vibrometer data. [Figure 41B] FIG. 10 illustrates program output for an exemplary program for a classification task on exemplary vibrometer data. [Figure 41C] FIG. 10 illustrates program output for an exemplary program for a classification task on exemplary vibrometer data. [Figure 41D] FIG. 10 illustrates program output for an exemplary program for a classification task on exemplary vibrometer data. [Figure 41E] FIG. 10 illustrates program output for an exemplary program for a classification task on exemplary vibrometer data. [Figure 42A] FIG. 10 illustrates program output for a segmentation task on exemplary vibrometer data. [Figure 42B] FIG. 10 illustrates program output for a segmentation task on exemplary vibrometer data. [Figure 43]FIG. 10 illustrates an example signal from an example variation of a contextual sensor module of an example variation of a sensing device that includes a microphone and a 9-axis inertial measurement unit (including a 3-axis accelerometer, a 3-axis gyroscope, and a 3-axis magnetometer). [Figure 44] FIG. 1 illustrates an example of vibro-acoustic test data collected on a consumer drone. [Figure 45] FIG. 10 illustrates exemplary vibro-acoustic test data collected by the sensing device of the present technology from a control subject (control), an asymptomatic subject with COVID-19 infection, a symptomatic subject with COVID-19 infection, and a subject with another respiratory disease (control). [Figure 46] FIG. 1 illustrates an exemplary workflow for processing detected data including windowing and feature extraction. [Figure 47] FIG. 10 shows exemplary vibroacoustic test data collected by the sensing device of the present technology from control subjects (Control 1, Control 2, Control 3) and subjects with COVID-19 infection (S002, S003, S004), where the vibroacoustic test data was collected while the subjects were in a supine position and the data was collected from a second ICS on the right side of the subjects' bodies. [Figure 48] FIG. 10 shows exemplary vibroacoustic test data collected by the sensing device of the present technology from subjects (S002, S003, S004) suffering from COVID-19 infection at initial and follow-up visits. In this case, the vibroacoustic test data was collected while the subject was in a supine position, and the data was collected from a second ICS on the right side of the subject's body. [Figure 49] FIG. 10 shows exemplary vibroacoustic test data collected by the sensing device of the present technology from control subjects (Control 1, Control 2, Control 3) and subjects with COVID-19 infection (S001-S0015). The vibroacoustic test data was collected while the subjects were in a supine position, and the data was collected from a second ICS on the right side of the subjects' bodies. [Figure 49A] This is a continuation of Figure 49. [Figure 50] FIG. 10 illustrates exemplary vibro-acoustic test data collected by the sensing device of the present technology from a control subject and a subject with COVID-19 infection where the vibro-acoustic test data was collected while the patient was in the supine position, and from different locations on the patient's body (carotid artery, right clavicle, left clavicle, left 2nd ICS, right 2nd ICS, right 4th ICS, left 4th ICS, left 6th ICS, right 6th ICS). [Figure 50A] This is a continuation of Figure 50. [Figure 51] Figure 1 shows sensor data extracts (vibroacoustics only) for two subject cases and three control cases taken while the patient was in different positions and performing different actions (sitting, standing, supine, left lateral, cough and hand-wringing, speech) when the sensing device location was on the right second ICS on the body. [Figure 51A] This is a continuation of Figure 51. [Figure 52A] FIG. 10 illustrates vibro-acoustic test data collected by a sensing device of the present technology when a subject with no clothing obstruction is positioned 12 cm from the diaphragm of the sensing device. [Figure 52B] FIG. 10 illustrates vibro-acoustic test data collected by a sensing device of the present technology when a subject with no clothing obstruction is positioned 100 cm from the diaphragm of the sensing device. [Figure 52C] FIG. 10 illustrates vibro-acoustic test data collected by a sensing device of the present technology when a subject wearing a sweater is positioned 12 cm from the diaphragm of the sensing device. [Figure 52D] FIG. 10 illustrates vibro-acoustic test data collected by a sensing device of the present technology when a subject wearing a sweater is positioned 100 cm from the diaphragm of the sensing device. [Figure 53A]FIG. 10 illustrates vibro-acoustic test data collected by a sensing device of the present technology when a subject wearing a sweater is positioned 10 cm from and facing the diaphragm of the sensing device. [Figure 53B] FIG. 10 illustrates vibro-acoustic test data collected by a sensing device of the present technology when a subject wearing a sweater is positioned 10 cm from the diaphragm of the sensing device and facing away from the diaphragm. [Figure 53C] FIG. 10 illustrates vibro-acoustic test data collected by a sensing device of the present technology when a subject wearing a sweater is positioned 100 cm from and facing the diaphragm of the sensing device. [Figure 54] 4A-4I show vibro-acoustic test data collected by a sensing device of the present technology, such as the sensing device of FIGS. 4A-4I, from a cow, a sheep, and a goat, in accordance with some embodiments of the present technology. [Figure 55] 4A-4I show vibro-acoustic test data collected by a sensing device of the present technology, such as the sensing device of FIGS. 4A-4I, from a cow, a sheep, and a goat, in accordance with some embodiments of the present technology. [Figure 56] 4A-4I show vibro-acoustic test data collected by a sensing device of the present technology, such as the sensing device of FIGS. 4A-4I, from a cow, a sheep, and a goat, in accordance with some embodiments of the present technology. [Figure 57] 4A-4I show vibro-acoustic test data collected by a sensing device of the present technology, such as the sensing device of FIGS. 4A-4I, from a cow, a sheep, and a goat, in accordance with some embodiments of the present technology. [Figure 58] 4A-4I show vibro-acoustic test data collected by a sensing device of the present technology, such as the sensing device of FIGS. 4A-4I, from a cow, a sheep, and a goat, in accordance with some embodiments of the present technology. [Figure 59]4A-4I show vibro-acoustic test data collected by a sensing device of the present technology, such as the sensing device of FIGS. 4A-4I, from a cow, a sheep, and a goat, in accordance with some embodiments of the present technology. [Figure 60] 13A-13C are perspective views illustrating exemplary variations of sensing devices embodied as animal gates in accordance with some embodiments of the present technology. DETAILED DESCRIPTION OF THE INVENTION

[0028] Non-limiting examples of various aspects and modifications of the present invention are described herein and illustrated in the accompanying drawings.

[0029] 1. A system for characterizing physical conditions 1.a. Overview As shown in FIG. 1A, a system 100 for characterizing a physical state (e.g., health state) may include one or more sensing devices 110 having one or more sensors configured to detect one or more biological parameters (e.g., vibroacoustic signals) of a subject.

[0030] In some variations, one or more sensing devices 110 may be modular and, as part of a modular sensing platform, include interchangeable subsystems adapted for modular experimentation, optimization, manufacturing, rapid field configurability, etc. For example, sensing device 110 may include sensor modules with one or more different sensor types, electronics modules, housings, and / or other modular components that may be interchangeable for different applications and contexts. Such a modular sensing platform may provide an architecture well suited for, for example, modular suites of portable wearable devices, remote screening devices and / or point-of-care solutions for healthcare, etc.

[0031] The one or more sensing devices 110 may have any suitable form factor for detecting biological parameters of a subject in a contactless or contact manner. The sensing devices 110 may be configured and positionable in any suitable manner with respect to the subject to capture suitable parameters.

[0032] For example, sensing device 110 may comprise a handheld housing so that a user may hold and / or manipulate sensing device 110 sufficiently close to or on the subject to capture sensor data (FIG. 1A). Sensing device 110 may be configured to capture suitable parameters through the subject's clothing or in direct contact with the subject's skin.

[0033] As another example, sensing device 110 may include a wearable housing that may be attached to the subject (e.g., with an adhesive patch) (FIG. 1A) or coupled to clothing or other apparel worn by the subject.

[0034] In some variations, sensing device 110 is configured to capture data from a subject in a remote manner. In this regard, sensing device 110 may be a standalone device (FIGS. 1B and 1C) arranged to be supported by a floor, wall, ceiling, or other structural support and configured to capture data from a subject in close proximity to sensing device 110.

[0035] In other variations, the sensing device 110 may be integrated into furniture such as a bed, bedding, mattress, pillow, blanket, sofa, chair, vehicle seat, or device such as a scale, mirror, panel, kiosk, doorway, sign, fitness equipment, or the like.

[0036] The sensing device 110 may be configured to communicate (e.g., wirelessly) with one or more computing devices 102, such as a mobile computing device, a smart watch, a local data gateway, or a computer, for processing, analyzing, communicating, and / or storing sensor data and / or other suitable information. For example, as shown in FIG. 1D , the sensing device 110 may be configured to communicate with the computing device 102 and / or other suitable modules over a network 104. The computing device 102 may additionally or alternatively collect data from other devices, such as weight scales, fitness and sports equipment, blood pressure sensors, blood glucose sensors, and / or one or more suitable mobile applications that can provide supplemental environmental and social determinants of health context information.

[0037] Additionally or alternatively, the sensing device 110 may be configured to communicate directly (e.g., pairwise) with the computing device 102 and / or other devices without the network 104. In other variations, the sensing device 110 may be configured to communicate directly with the network 104.

[0038] In some variations, the sensing device 110 may be configured to communicate with a suitable module, such as a pattern evaluation module 106, which may incorporate artificial intelligence (e.g., through the application of one or more trained machine learning models) to characterize one or more physical conditions of the subject based on the sensor data. The analyzed data may be stored locally on the computing device and / or on the smartphone gateway (e.g., sensor data, analysis of the sensor data, etc.), or may be stored in a suitable data storage module 108, such as a server, through cloud storage, and / or in an electronic medical record 109 associated with the subject. Additionally, in some variations, the machine learning models used to analyze the sensor data may be continuously trained or updated over time using additional health history, sensor fusion, and / or environmental and / or social determinants of health contextual data, as described further below. For example, the sensor data may be mined via a data mining module 107 for use in training and refining predictive and / or prescriptive models across patient populations. A communications module may be provided for data communication between various modules of the system.

[0039] In some variations, one or more components of system 100 for characterizing a physical state (e.g., a health condition) may include various output interfaces, such as for communicating information related to the physical state (e.g., sensor data, analysis of the sensor data, operational status, etc.). For example, as shown in FIGS. 1A-1H , system 100 for characterizing a physical state may include sensing device 110 (e.g., a handheld device, a wearable device, a standalone device, etc.) and computing device 102. Computing device 102 may be a standalone device, separate from the sensing device (e.g., a smartphone), or integrated with sensing device 110. In some variations, computing device 102 may be implemented as network-on-chip (NoC) technology. In some variations, computing device 102 and / or the sensing device may be implemented as a wearable device. Sensing device 110 may include a charging base 120 ( FIG. 1H ). Additionally, one or more of sensing device 110, computing device 102, and charging base 120 may include at least one output interface. For example, as shown in FIG. 1F , sensing device 110 may include one or more other suitable visual indicators, such as a display and / or a light indicator (e.g., an LED indicator). As another example, computing device 102 may include a tactile interface and / or an audio output (e.g., a speaker device). As another example, charging base 120 may include a tactile interface and / or a visual indicator, such as a light indicator (e.g., an LED indicator). However, it should also be understood that in some variations, any of sensing device 110, computing device 102, and / or charging base 120 may include any suitable combination of interfaces, such as one or more of visual, auditory, tactile, etc.One or more of the sensing device 110, the computing device 102, the network, and the charging base 120 may be configured to communicate information between each other (e.g., wirelessly, via a network, etc.), as described in further detail herein.

[0040] 1.b. Vibroacoustic sensor platform In some variations, the system 100 for characterizing a physical condition is a vibroacoustic sensor platform configured to detect and process vibroacoustic signals, alone or in combination with other sensor signals (not necessarily vibroacoustic signals), to diagnose or monitor a physical condition of a subject.

[0041] In these variations, system 100 includes a sensor module consisting of one or more sensors configured to detect vibroacoustic signals, and optionally signals that are not vibroacoustic signals, such as the subject's temperature, environmental conditions, etc. The signals detected by the sensor module are generally referred to herein as "sensor data." The sensor data may originate, at least in part, from the subject and / or the subject's environment. The sensor module may be embodied in a single sensing device 110 or in multiple sensing devices 110.

[0042] The system 100 may further include a signal processing system configured to extract a biosignal component originating from the subject from the detected sensor data. In some variations, the extracted biosignal component includes a biovibration acoustic signal component originating from the subject.

[0043] System 100 may include at least one processor configured to characterize or monitor a physical state of the subject based at least in part on the extracted biosignal components, such as biovibroacoustic signal components, e.g., using a machine learning model. It should be understood that the physical state may be characterized on sensor data, such as one or more of vibroacoustic signal data, extracted biovibroacoustic signal components, signals that are not vibroacoustic signals, and biosignals extracted from non-vibroacoustic signals originating from the subject.

[0044] The at least one processor may be embodied in one or more computing devices configured to perform processing, analysis, communication, and / or storage, etc. of sensor data from the sensor module.

[0045] The sensor module may be configured to detect a wide range of vibroacoustic frequencies, which may separately and simultaneously provide useful indicators of human health, animal health, and / or structural health, including those not traditionally monitored or detected. More specifically, in some variations, the sensor module is configured to detect vibroacoustic signals below and above the threshold of human hearing.

[0046] As shown in FIG. 1E, the human hearing threshold decreases rapidly when vibration frequencies are below approximately 500 Hz. However, in a healthy subject at rest, most cardiac, respiratory, digestive, and motor-related information is inaudible to humans because this information occurs at frequencies lower than those associated with sound. Therefore, the majority of body vibrations are not detected or included in conventional diagnostic medical procedures due to the low frequency range of these vibrations and the limited bandwidth limitations of conventional instruments (e.g., conventional stethoscopes). Modifications of system 100 described herein can detect, amplify, and analyze a wide range of infrasonic, ultrasonic, and hypersonic vibration sound frequencies, thereby advantageously providing a more comprehensive and holistic view of a subject's health and condition.

[0047] Additional details of the system 100 and its components, sensing device 110, and method are further described below. Such system 100, sensing device 110, and method are described solely with respect to characterizing the physical state of a subject, which may be, for example, a human or another animal (e.g., in the context of human health care and / or veterinary care). However, it should be understood that other applications of the system 100, sensing device 110, and method may relate to characterizing non-living objects, including, but not limited to, heating, ventilation, and air conditioning (HVAC) systems, internal combustion engines, jet engines, turbines, bridges, aircraft wings, ambient infrasound, ballistics, drone and / or watercraft identification, etc. For example, the system 100, sensing device 110, and method may be applied to characterize structural health (e.g., characterizing the structural integrity of bridges, buildings, aircraft, vehicles, etc.), environmental noise pollution, rotary motor engine performance optimization, monitoring, etc.

[0048] 1.c. Exemplary Sensing Devices 1. CI handheld sensing device In some variations, system 100 may include a handheld sensing device 200, as shown generally in FIG. 2. Sensing device 200 includes a sensor module including a vibroacoustic sensor module 220 and, optionally, a context sensor module 230, and an electronics system 240 that may handle sensor data from, for example, vibroacoustic sensor module 220 and / or context sensor module 230. One or more of these components are enclosed or otherwise at least partially configured within a housing 210 of sensing device 200, which may have any suitable form factor for different applications as described further below. In some variations, housing 210 may include a display 250 for providing a user interface for the sensing device (e.g., for displaying information to a user, for enabling control of the sensing device, etc.). Some other examples of form factors of sensing device 110 or 200 are illustrated in Figures 3A and 3B, 4A-4E, 4F-4G, 5A-5I, 27A-27B, 28A-28B, 29, 30, 31A-31E, 32A-32B, and 33A-33C.

[0049] 3A and 3B illustrate another exemplary sensing device 300, which is a variation of sensing device 110. As shown in FIG. 3A, sensing device 300 may include a housing 310 within which various sensing device components may be disposed. The housing 310 shown in FIG. 3A is configured as a handheld housing 310 having a generally cuboid shape; however, it should be understood that other housing variations may have other shapes (e.g., spherical, ellipsoidal, other prismatic shapes, etc.). Other exemplary sensing device form factors are described in further detail below.

[0050] In some variations, the base 302 may be coupled to the housing 310. The base 302 may, for example, provide a surface on which the housing 310 may rest and / or provide an indication of device orientation (e.g., indicating the primary direction in which the sensor module within the housing 310 is sensing). In some variations, the base 302 may function as a charging cradle, such that the housing 310 is separable from the base 302 when the sensing device 300 is in use.

[0051] 3B , housing 310 may house a vibroacoustic sensor module 320 including one or more sensors for collecting vibroacoustic signals, an optional context sensor module 330, and an electronics system 340 including components for processing sensor data from vibroacoustic sensor module 320 and optionally context sensor module 330. In some variations, electronic components (e.g., within the vibroacoustic sensor module and / or within electronics system 340) may perform various signal processing functions to extract bioacoustic signal components from the sensor data and / or perform analysis via artificial intelligence (e.g., utilizing one or more machine learning models) to characterize a body condition based on the bioacoustic signal components.

[0052] In some variations, the housing 310 may include one or more guides or compartments for receiving and positioning various internal device components. For example, as shown in FIG. 3B , the housing 310 may include one or more internal protrusions 316 (e.g., shoulders, lips, or other guides) configured to restrain adjacent device components, such as the vibroacoustic sensor module 320 or the electronics system 340. This restraint may be provided by an interference fit, one or more fasteners, and / or the like. While FIG. 3B depicts the internal device components as generally axially aligned, it should be understood that in other variations, the components may be arranged in any suitable manner (e.g., orthogonal to one another, etc.).

[0053] The housing 310 may include one or more openings and / or other structures to facilitate communication with the sensor. For example, the housing 310 may include a sensor opening adjacent the vibroacoustic sensor module 320 that allows entry and propagation of vibroacoustic waves toward the vibroacoustic sensor within the module and / or a membrane or other receiver that interacts with the vibroacoustic sensor. Additionally or alternatively, in some variations, the sensing device 300 may include an impedance-matching diaphragm 350 arranged in series with the vibroacoustic sensor module 320 to improve sensitivity to a wide range of vibroacoustic frequencies. The diaphragm 350 may be dome-shaped, for example, and impedance matching may refer to an operating condition in which the load impedance and the internal impedance of the excitation source are matched to each other (e.g., within a tolerable impedance difference), thereby providing maximum output. Impedance mismatch may result in undesirable high attenuation and / or reflection of the source signal away from the vibroacoustic sensor module. This issue can be addressed, for example, by developing an impedance matching circuit using machined adjustable components that utilize multiple diaphragm cutout designs (also referred to as "apertures," an example of which is shown in FIG. 18) to obtain optimal dynamic range within the low infrasound range. The combined diaphragm / transducer solution can include a suitable protective material (e.g., rubber, felt liner, lightweight foam, etc.) around the diaphragm and fixed area to protect the interior of the transducer from water and particulate matter by sealing the gap. The diaphragm can comprise a passive material (including, but not limited to, polymer, polyamide, polycarbonate, polypropylene, carbon fiber, fiberglass, etc.) engineered as a structural layer with dimensions optimized for the physical system form factor. To obtain a tunable impedance matching system and structural layer, the gap between the diaphragm and voice coil can be adjustable. The structural layer can be spun and patterned to ensure shape and optimal weight, and then bonded.

[0054] In some variations, the housing 310 may include a power connection port 314 that allows for connection of an auxiliary power source to the electronics system 340 and / or other powered components. Additionally or alternatively, the housing 310 may include a data connection port (not shown) that may provide (upload or download (wired) options) data or other information to and from the electronics system 340. In some variations, power and data may be transferred through the same port (e.g., via a USB connection).

[0055] Additionally, in some variations, the housing 310 may include a user interface, such as a display 360, visible through a screen or opening in the bezel 312 of the top cover 313 or other suitable portion of the housing 310. The display 360 may include an LED screen, an LCD screen, or other suitable monitor screen. The display 360 may be configured to display information to a user (e.g., diagnostic information, sensor data, device status, etc.), such as on a graphical user interface (GUI), enable control of the sensing device (e.g., power state, operating mode, etc.), and / or the like. In some variations, the display 360 may include a touchscreen for receiving user input. In some variations, a vibro-acoustic sensor or dedicated microphone may be used to receive audio user input. Additionally or alternatively, user input may be provided to the sensing device 300 through one or more hardware user interface elements (e.g., buttons, slides, switches, touch sensors, etc.) on or coupled to the housing. Additionally or alternatively, information regarding the device and / or analysis may be communicated through other components (e.g., visual notification via an array of LEDs, audio notification via a speaker device, tactile notification via a vibration motor, etc.). For example, in some variations, acoustic data may be played on one or more speaker devices (or communicated to a peripheral device for playback). Further, in some variations, one or more device orientation and / or position sensors (e.g., an inertial measurement unit (IMU), accelerometer, gyroscope, etc.) may be used for user input and functional operation of the sensing device (e.g., shaking or rotating the sensing device may toggle device settings, wake the sensing device, or put it into sleep or standby mode, etc.).

[0056] The housing 310 may be fabricated in a variety of suitable manners, such as by injection molding, milling, 3D printing, etc. As shown in FIG. 3B , the housing 310 may include multiple connectable portions (e.g., halves of a shell-type housing split into two) that combine to form one or more internal volumes that receive internal device components. The connectable portions may be joined together by mechanical interlocking features (e.g., pins and holes sized to engage in a friction fit), fasteners (e.g., screws, adhesive, etc.), and / or any suitable manner. Alternatively, the housing 310 may be integrally formed as a single piece.

[0057] In some variations, some of the components of sensing device 300 may be intended to be reusable and / or upcycled, while other components may be disposable. For example, in some variations, less expensive components, such as housing 310 and / or subject-contacting components (e.g., impedance matching diaphragms), may be replaced, while more expensive components, such as the electronic components of sensing device 300, may be reused. In some variations, components may be sanitized (e.g., with isopropyl alcohol) instead of being replaced and discarded. In this regard, sanitizable components may be made from materials suitable for sanitization with one or more of alcohol, hydrogen peroxide, steam, ethylene dioxide, gamma radiation, ultraviolet light, etc. In some variations, sensing device 300 includes one or more sensors (e.g., proximity sensors, Hall Effect sensors, contact sensors, etc.) that detect and authenticate the attachment and removal of replaceable parts, thereby enabling system 100 to intelligently monitor usage, sterility change events, duration of use between sterility cover changes, battery levels, number of uses, and / or other usage data, etc. Such data may be used, for example, to monitor reports regarding compliance with best practices and required protocols for maintaining cleanliness.

[0058] In some variations, the housing 310 may be configured to be attachable to another type of device, such as a stethoscope. In this regard, retrofitting may be facilitated by a winding for a screw mount or other fastener.

[0059] 4A-4E illustrate sensing device 400, another exemplary variation of sensing device 110 of system 100. Sensing device 400 is also modular and includes multiple components that can be assembled together. For example, sensing device 400 includes a housing 410 connectable to a top cover 413 and a bottom cover 415. Sensing device 400 has a handle end H and a sensor end S. As before, housing 410 may be connectable to a base (not shown) for charging and / or support. Housing 410 is configured to substantially enclose one or more components, including a vibroacoustic sensor module 420, an electronics system 440, a power source 442, a diaphragm 450 (such as an impedance matching diaphragm 450), and first and optionally second capacitive electrodes 460, 470 for electrocardiogram (ECG) measurements. In some variations, the first capacitive electrode 460 includes an electro-capacitive integrated circuit (EPIC) electrode 460. In some variations, the second capacitive electrode 470 includes a drive right leg (DRL) electrode 470. In some other variations, the capacitive electrodes 460, 470 may be omitted (FIG. 4H). The diaphragm 450 is supported by the bottom cover 415 and extends from an opening in the bottom cover 415.

[0060] As can be seen in FIGS. 4F and 4G , in some variations, an outer cover 480 may also be provided to seal the connection between the bottom cover 415 and the diaphragm 450 to limit or prevent the ingress of foreign matter. The outer cover 480 may be made of rubber or any other suitable material, including materials suitable for medical applications. The outer cover 480 may be connected to the bottom cover 415 by adhesive or the like. In the variations of FIGS. 4F and 4G , the outer cover 480 stops just short of the perimeter of the bottom cover 415. However, in some other variations, the outer cover 480 may extend completely across the bottom cover 415 and also extend upward from the bottom cover 415 like a cap. This may allow the sensor end S of the sensing device 400 to be immersed in a liquid, such as a disinfectant solution. A solution may also be provided in the base, such as 302, for sterilization or cleaning. In some variations, the gap to be covered between the diaphragm 450 and the bottom cover 415 may be any length between 0.001 mm and 20 mm, such as about 2 mm, about 4 mm, about 6 mm, about 8 mm, or about 10 mm.

[0061] FIG. 4I shows one variation of the outer cover 480 of the sensing device 400 in that the outer cover 480 includes a flexure 481 that can allow the outer cover 480 to expand to accommodate diaphragm movement. The flexure is positioned radially. In this variation, the gap between the diaphragm 450 and the bottom cover 415 is approximately 2 mm, and the flexure 481 has a radius of approximately 2 mm. This can allow the diaphragm to move compliantly and linearly up to 4 mm out of plane. In one variation, the outer cover 480, including the flexure 481, can span the bottom of the device and extend up the sides of the device (not shown) to withstand fluids.

[0062] Sensing device 400 differs from sensing device 300 in its form factor as well as the types of sensors included. More specifically, sensing device 400 does not include context sensor module 330. Sensing device 400 may include EPIC electrodes 460 and DRL electrodes 470 for collecting ECG data. EPIC and DRL electrodes 460, 470 are described in further detail below with reference to FIGS. 19-22.

[0063] In some variations, the sensing device 400 has a size and shape that allows it to be handheld and operated. Optimizing the size and shape of the sensing device 400 involves many considerations, including one or more of the spacing of the EPIC electrodes 460 and DRL electrodes 470 from each other and / or from the vibroacoustic sensor module 420 to avoid interference, and the size of the diaphragm 450. Thus, in one variation, four EPIC electrodes 460 and four DRL electrodes 470 are provided, spaced apart from each other and positioned around the diaphragm 450. The diaphragm 450 has a diameter of approximately 39 mm, and each EPIC electrode has a length of approximately 10 mm. It will be understood that the arrangement and size of the electrodes 460, 470 may differ from that illustrated. Sensing devices with alternative dimensions are within the scope of this disclosure.

[0064] In some variations, one or more of the DRL electrodes 470 may include a guard electrode. Indeed, in some variations, the function of the electrode 470 may be configurable as either a guard electrode or a DRL electrode depending on whether the use is skin contact or non-direct skin contact. In some other variations, the bottom cover 415 is made of a conductive material and configured as a guard electrode.

[0065] 4H illustrates sensing device 400′, a variation of sensing device 400 of FIGS. A-E, differing only in that the EPIC electrode 460 and driving right leg (DRL) electrode 470 for sensing ECG data are omitted. Sensing device 400′ is configured to sense only vibroacoustic signals.

[0066] 1.c.ii. Exemplary Sensors for the Vibrometer Sensor Module In some variations, a vibroacoustic sensor module, such as vibroacoustic sensor modules 220, 320, 420, 2720, 2820, 2920, and 3140, may include one or more sensors configured to detect vibroacoustic signals. The vibroacoustic sensor module may also include one or more deflection structures in conjunction with the sensor components of the sensor. For example, the one or more sensors may be selected and / or configured to interact with the one or more deflection structures to measure various characteristics of the deflection structure's movement (e.g., in response to vibroacoustic waves). Such movement, measurable by the one or more sensors, may be analyzed to assess the subject's physical condition. As described further below, in some variations, the one or more sensors may be printable circuits, flexible, skin-friendly printable circuits embodied as electronic tattoos, and / or disposed on a suitably flexible circuit board or other structure that does not significantly interfere with the interaction of the sensor and the deflection structure, thereby avoiding a reduction in the sensitivity and / or bandwidth of the vibroacoustic sensor module. Exemplary variations of sensor arrangements and deflection structures are described in further detail below.

[0067] In some variations, the vibroacoustic sensor module may have a bandwidth suitable for detecting vibroacoustic signals in the infrasound range, such as a bandwidth in the range of about 0.01 Hz to at least about 20 Hz. Additionally, in some variations, the vibroacoustic sensor module may have a wider bandwidth covering a wider frequency range from infrasound to ultrasound, such as a bandwidth in the range of about 0.01 Hz to at least 160 kHz. In some variations, the bio-vibroacoustic signal component extracted from the detected vibroacoustic signal may have a bandwidth in the range of about 0.01 Hz to 0.1 Hz.

[0068] For example, in some variations, the vibroacoustic sensor module may have an overall bandwidth within the range of about 0.01 Hz to at least about 50 kHz, about 0.01 Hz to at least about 60 kHz, about 0.01 Hz to at least about 70 kHz, about 0.01 Hz to at least about 80 kHz, about 0.01 Hz to at least about 90 kHz, about 0.01 Hz to at least about 100 kHz, about 0.01 Hz to at least about 110 kHz, about 0.01 Hz to at least about 120 kHz, about 0.01 Hz to at least about 130 kHz, about 0.01 Hz to about 140 kHz, about 0.01 Hz to at least about 150 kHz, about 0.01 Hz to about 160 kHz, or about 0.01 Hz to greater than about 150 kHz.

[0069] The vibroacoustic sensor module may, in some variations, comprise a single sensor that provides one or more of the above-mentioned bandwidths of the detected vibroacoustic signal.

[0070] In some other variations, a vibro-acoustic sensor module may include a set of multiple vibro-acoustic sensors, each having a respective bandwidth that forms a segment of the overall vibro-acoustic sensor module bandwidth. At least some of these multiple sensors may have respective bandwidths that at least partially overlap. Thus, various sensor module bandwidths may be achieved based on the selection of specific sensors that collectively contribute to a particular vibro-acoustic sensor module bandwidth. In other words, a bandwidth extension and linearization approach (bandwidth predistortion) may utilize modular sensor fusion and response feedback information, such as to compensate for the bandwidth limitations of any particular single sensor with an overlapping combination of sensors to cover a wider bandwidth with optimal performance.

[0071] For example, the vibro-acoustic sensor module may include one or more sensors for measuring vibro-acoustic signals. The one or more sensors may be selected from passive and active sensors for acquiring vibro-acoustic data, such as one or more of a microphone, a voice coil, an accelerometer, a pressure sensor, a piezoelectric transducer element, a Doppler sensor, and the like. For example, the vibro-acoustic sensor module may include a voice coil-based sensor. In another example, the vibro-acoustic sensor module may include a voice coil-based sensor and an echo-doppler-based ultrasonic sensor. The vibro-acoustic sensor module may include one or more microphones, such as a dynamic microphone, a large diaphragm condenser microphone, a small diaphragm condenser microphone, and / or a ribbon microphone. Additionally or alternatively, the vibro-acoustic sensor module may include a linear position transducer. Such sensors may be configured to detect and measure vibro-acoustic signals by interacting with a suitable deflection structure that moves in response to the vibro-acoustic signals. In some variations, the vibroacoustic sensor module may combine multiple microelectromechanical systems technology cross-axis inertial sensors capable of detecting vibroacoustic signals in the range of about 20 Hz to about 20 kHz.

[0072] Additionally or alternatively, the vibroacoustic sensor module may include a MEMS cross-axis inertial sensor fusion capable of detecting vibroacoustic signals in a range of about 1 Hz (or lower frequencies) to several kHz (e.g., frequencies between about 1 Hz and about 2 kHz). Furthermore, the vibroacoustic sensor module may additionally or alternatively include a MEMS cross-axis inertial sensor capable of detecting vibroacoustic signals in a range of about 0.01 Hz to several hundred Hz (e.g., about 0.05 Hz to about 10 kHz). Additionally or alternatively, other suitable vibroacoustic sensors may be included in the vibroacoustic sensor module, such as a voice coil transducer, a piezoelectric transducer, etc. In some variations, transmission of the vibroacoustic waves may occur through an intermediate medium such as air and / or across a deflecting structure.

[0073] Additionally, a suite of sensors within a vibroacoustic sensor module may be configured (alone or in combination with a context sensor module, described in more detail below) to more fully capture longitudinal and lateral vibrations, as well as environmental context and disturbances. In some variations, the environmental context signal may be useful for contextualizing the collected associated biophysiological data. Additionally or alternatively, in some variations, environmental disturbance data (e.g., ambient noise) may be used for noise reduction from the associated bioacoustic vibration signal components. Such noise reduction may be performed, for example, on-device as active noise cancellation or as a post-processing stage. In some variations, the deflection structure of the vibroacoustic sensor module may generally have a nominal or static configuration in which the deflection structure is arranged and configured in a plane, and the deflection structure may deflect or flex in response to out-of-plane forces. In these variations, the deflection structure may be configured to have low stiffness (or resistance) to out-of-plane movement, good compliance with skin movement, but high stiffness or resistance to in-plane movement, and low crosstalk between in-plane axes. Thus, the deflection structure is highly sensitive to acoustic waves directed at the deflection structure (ie, acoustic waves with vector components orthogonal to the deflection structure), yet is robust to noise contributed by other forces.

[0074] Additionally, in some variations, the deflection structure of the vibroacoustic sensor module may have a relatively small mass in the moving parts of the deflection structure to reduce inertia (and further improve sensitivity to out-of-plane forces). In some variations, the deflection structure may be designed with low or no hysteresis so that out-of-plane motion is highly linear.

[0075] Additionally or alternatively, the deflection structure may be designed to have low material fatigue over time so as to be predictable and consistent over long-term use of the sensing device.

[0076] The deflection structure may, in some variations, comprise a harder material, such as a hard plastic, and may be formed by 3D printing, milling, injection molding, or any suitable manner. For example, the deflection structure may comprise materials including, but not limited to, polyamide, polycarbonate, polypropylene, carbon fiber, fiberglass, and / or other suitable materials.

[0077] Bending arm type vibration and acoustic sensor module In some variations, the vibroacoustic sensor module may include one or more deflection structures including at least one flexure arm. For example, FIGS. 5A-5I illustrate an exemplary variation of a vibroacoustic sensor module 500 having multiple flexure arms. As shown in the cross-sectional views of FIGS. 5A and 5B, the vibroacoustic sensor module 500 may be disposed within a housing 510 of a vibroacoustic sensing device (e.g., similar to the sensing device 300 shown in FIGS. 3A and 3B or the sensing device 400 shown in FIGS. 4A and 4B), such as being constrained within a slot or other cavity in the housing 510. In some variations, the vibroacoustic sensor module 500 may include one or more flexure arms configured to deflect in a direction generally perpendicular to the sensing side (S) of the housing 510.

[0078] As shown in FIGS. 5C-5F, the vibroacoustic sensor module 500 may include a deflection structure 520 and a flex circuit 530 including a cross-axis inertial sensor, such as an accelerometer 532 (e.g., a MEMS accelerometer). The deflection structure 520 may include two or more flexure arms 524 supported at their outer ends by the frame 520 and at their inner end grooves by a central hub 526 that functions as a signal pickup. In some variations, the flexure arms may be radially distributed around the central hub 526. The flex circuit 530 may include at least one accelerometer 532 attached to the central hub 526 at its inner end, which may move with the central hub 526 in response to impinging vibroacoustic waves. In some variations, an impedance-matching receiver 540 may also be coupled to the central hub 526 to increase the sensitivity of the deflection structure 520 to impinging vibroacoustic waves. As shown in FIG. 5E, the receiver 540 may be, for example, a dome-shaped diaphragm. Thus, the vibrating acoustic waves may cause movement of the receiver 540 and the deflecting structure 520 that is detected and measured by the accelerometer 532 .

[0079] 5G-5I illustrate additional details of the vibroacoustic sensor module 500. As shown in FIGS. 5G and 5I, the deflection structure 520 may include two flexure arms 524 coupled to a generally planar central hub 526. The flexure arms 524 are radially distributed around the central hub 526 and, in some variations, may be radially equidistant from each other or asymmetrically spaced apart (i.e., tuned to different resonant frequencies and therefore different frequency bands) (i.e., the two flexure arms 524 may be arranged approximately 180 degrees apart around the central hub). The flexure arms 524 may be undulating (e.g., arched or sinusoidal), which helps optimize out-of-plane flexibility and in-plane stiffness. The cross-axis inertial sensor (on the flex circuit 530) may rest on a cantilever beam to enable effective coupling with minimal mechanical constraints from the rest of the system. The pre-buckled, out-of-plane, arc-like geometry allows the flexed arms 424 (e.g., serpentine interconnects) to adopt a traction-free architecture that accommodates multiple modes of tensile deformation. Finite element modeling (FEM) can highlight the significant mechanical advantages of these non-coplanar interconnects through out-of-plane linearity and stiffness compliance for in-plane motion compared to traditional planar serpentine layouts. Thus, out-of-plane forces (e.g., vibratory acoustic waves) can deflect the flexed arms 524 and bias the central hub to move out-of-plane, while resulting in little or no lateral in-plane movement of the flexed arms 524. In some variations, as shown in FIG. 5I, the central hub can be at least partially hollowed out with one or more cutouts to reduce the hub's mass and inertia (which may help, for example, improve sensitivity to out-of-plane forces).

[0080] As mentioned above, the flex circuit 530, like the bending arm 524, may be flexible and able to accommodate out-of-plane movement. The flex circuit 430 may provide the vibrometer sensor module with an overall less agile stiffness that is advantageously more isotropic. In some variations, as shown in FIG. 5H , the flex circuit board 530 may have a generally spiral shape in a plane, winding outward from the inner end to the outer end. As mentioned above, the inner end of the flexible circuit board 530 may be coupled to the central hub 526 such that the accelerometer 532 on the inner end of the flexible circuit board 430 moves in conjunction with the central hub 526 (e.g., in response to vibrational acoustic waves). At its outer end, the flexible circuit board 530 may be configured to attach to a flex circuit anchor 523 (e.g., on or coupled to the frame of the deflectable member 520) shown in FIG. 5I. The outer end of the flexible circuit board 530 may also include a cable connector 534 electrically connected to conductive traces (not shown) that traverse the flexible circuit board 530 so that signals can be transmitted to and from the accelerometer 532 via the cable connector 534 and the conductive traces.

[0081] However, the flexible circuit board 530 may have any suitable shape that is sufficiently flexible and capable of accommodating out-of-plane movement. For example, FIG. 6A shows an exemplary variation of a vibroacoustic sensor module 600 that includes a deflection structure 620 having two bending arms that meet at a central hub and a flexible circuit board 630 coupled to the deflection structure 620. Similar to the flexible circuit board 530 described above with reference to FIGS. 5A-5I, the flexible circuit board 630 includes a cable connector 634 at the outer end of the flexible circuit board 630 and an accelerometer 632 at the inner end of the flexible circuit board 630, with conductive traces extending between the accelerometer 632 and the cable connector 634 for transmission of signals to and from the accelerometer 632. However, as shown in FIG. 6B, unlike the spiral-shaped flexible circuit board 530, the flexible circuit board 630 may have a generally zigzag shape.

[0082] In some variations, the flexure arm may have a flexure thickness of less than about 1 mm so as to be sufficiently resilient while still being sufficiently sensitive to vibroacoustic signals. For example, the flexure arm may be formed from an SLA type photopolymer, which has good stiffness and low internal damping.

[0083] In some variations, the central hubs of the deflection structures (e.g., deflection structures 520 and 620) may have different thicknesses to provide different amounts of clearance for the bending arms. For example, FIG. 7A shows a vibro-acoustic sensor module 700 that includes a deflection structure 720 and a flexible circuit board 730. Similar to those described above, the deflection structure 720 may include a central hub 726 that is configured to move laterally, as indicated by the arrow, when the bending arms move in an out-of-plane direction.

[0084] 7B shows a vibro-acoustic sensing module similar to vibro-acoustic sensing module 700, except that vibro-acoustic sensing module 700′ has a thicker central hub 726′ that extends farther from the rest of the deflection structure. Central hubs 726 and 726′ may be applied to contact surface (C) (e.g., the skin of a subject) when the sensing device is in use, but the thicker central hub 726′ increases the distance between the bending arms of deflection structure 720 and contact surface (C) more than central hub 726 does.

[0085] While FIGS. 5A-5I illustrate an exemplary variation of the deflection structure 520 having two flexure arms, it should be understood that other variations of the deflection structure can include other suitable numbers of flexure arms. The flexure arms can be distributed radially around a central hub, and in some variations, can be equally radially or asymmetrically distributed around the central hub to tune to different resonant frequencies and thereby different frequency bands. For example, FIGS. 8A and 8B depict a vibroacoustic sensor module 800 including a deflection structure 820 having three flexure arms 824. The three flexure arms can be wavy (e.g., arched or sinusoidal) and equally radially arranged around the central hub of the deflection structure (i.e., the three flexure arms can be spaced approximately 120 degrees apart around the central hub). The vibroacoustic sensor module 800 can further include a flexible circuit board 830. Although flexible circuit board 830 is shown in Figures 8A and 8B as having a zigzag shape similar to flexible circuit board 620 shown in Figure 6B, in other variations, flexible circuit board 830 may have a spiral shape similar to that shown in Figure 5H, or other suitable shape.

[0086] Additionally, a flex circuit may not be necessary if the sensor is powered by induction or piezoelectricity and can return data to the electronic system through induction, radio frequency, magnetic flux change, or optical communication.

[0087] Additionally, in some variations, the deflection structure may have four or more bending arms. For example, in some variations, the deflection structure may include four bending arms along the edge of the sensor mass (e.g., based on FEM for design optimization). In still other variations, the deflection structure may include five, six, or more than six bending arms. In addition to such mechanical deflection structure designs, more sophisticated flexible dielectric nanocomposites may be incorporated into the deflection structure, including, but not limited to, graphene, reduced graphene oxide / titanium dioxide (rGO / TiO) nanocomposites, polyvinyl alcohol-modified polyvinylidene fluoride-graphene oxide, polyvinyl fluoride (PVF) or -(CHCHF)n-, and / or any combination of poly(vinylidene fluoride) or polyvinylidene fluoride or polyvinylidene difluoride (PVDF) incorporated with reduced graphene oxide (rGO) and poly(vinyl alcohol)-modified rGO (rGO-PVA). Nanocomposites have inherently unique properties and may be conveniently fabricated into nanostructures of different morphologies, such as by substrate spraying and doping, from atomically thin monolayers to nanoribbons. In yet other variations, the deflection structure may have five, six, or more than six bending arms.

[0088] While the above description exclusively describes a deflection structure in which an accelerometer is coupled to a central hub connecting two or more flexure arms, in some variations, the deflection structure may include multiple accelerometers, each coupled to a respective flexure arm. For example, as shown in FIGS. 9A-9C , an exemplary variation of a vibroacoustic sensor module 900 may include a deflection structure 920 having multiple flexure arms 924 arranged radially around a central hub 926 and an accelerometer 932 disposed on the outer or distal ends of the flexure arms 924. In some variations, the deflection structure 920 may include a handle portion 950 and be handheld (and / or may be used to secure to another portion of a sensing device). The flexure arms 924 may be arranged in an undulating (e.g., sinusoidal or arcuate) and / or uniformly radial configuration around the central hub 926, similar to those described above. 9D , a flexible circuit board 930 may be coupled to each flex arm 924 and shaped in a similar manner to its respective flex arm 924. Additionally, in some variations, the vibro-acoustic sensor module 900 may include a receiver (e.g., a dome-shaped diaphragm) coupled to a central hub 926 of the deflection structure. Similar to the vibro-acoustic sensor modules described above, the central hub 926 may move out of plane in response to a vibro-acoustic signal, which may result in a corresponding out-of-plane movement of the accelerometer mount 928 acted upon by the flex arm 924 that is detectable by the accelerometer 932 at the outer end of the flex arm 924.

[0089] Although the vibroacoustic sensor module 900 is shown in FIGS. 9A and 9B as including a deflection structure having three flexure arms, it should be understood that the deflection structure may include more than three (e.g., four, five, six, or more than six) flexure arms arranged about a central hub, with each flexure arm including a respective accelerometer arranged thereon.

[0090] Membrane-type vibroacoustic sensor module In some variations, the deflection structure comprises a membrane, and one or more sensors may be configured to interact with the membrane to detect out-of-plane movement of the membrane in response to a vibroacoustic signal. For example, as shown in FIGS. 10A and 10B , a vibroacoustic sensor module 1000 may comprise a deflection structure 1020 having a membrane 1024 extending across a frame 1022 and an accelerometer 1032 coupled to a central region of the membrane 1024 via a flexible circuit board 1030. The membrane may comprise a thin sheet or diaphragm of flexible material, such as an elastomeric material (e.g., latex, nitrile, etc.), stretched taut over the frame 1022. In some variations, the membrane may be coupled to the frame 1022 by one or more suitable fasteners, such as a clamp ring 1025 that radially compresses and secures the membrane 1024 to the frame 1022, by other suitable fasteners (e.g., epoxy), or in any other suitable manner. In some variations, FEM design optimization models suggest an optimal membrane thickness range of 10 to 5,000 micrometers and a diameter ranging from about 2.5 millimeters to about 75 millimeters. One goal of the membrane is to enable high-fidelity data collection while providing a comfortable, non-irritating data-harvesting interface. The membrane material can be configured to deform naturally with low-amplitude body movements. For example, in some variations, the membrane design can incorporate deformable, non-coplanar interconnects, a base strain-insulating layer, a soft encapsulation overlayer, and / or a hollow cavity configuration. Together, these features can provide a low-modulus, elastic mechanical structure for ultrasensitive signal pickup.

[0091] Similar to the flex-arm-based variations of the vibroacoustic sensor module described above, the vibroacoustic sensor module 1000 may include a flexible circuit board 1030 that is flexible and capable of accommodating out-of-plane forces. For example, while FIGS. 10A and 10B depict the flexible circuit board 1030 having a zigzag shape similar to that shown in FIG. 6B , it should be understood that the flexible circuit 1030 may alternatively have a spiral shape (as shown in FIG. 5H ) or any suitable shape. The flexible circuit board 1030 may have an inner end having an accelerometer 1032 and an outer end having a cable connector 1034, with conductive traces extending between the accelerometer 1032 and the cable connector for transmitting signals to and from the accelerometer 1032. In some variations, at least the inner end of the flexible circuit board 1030 may be bonded to the membrane with epoxy or any other suitable fastener, or in any suitable manner. Thus, movement of the membrane 1024 may be tracked and measured by the accelerometer 1032, and corresponding vibro-acoustic sensor signals from the accelerometer 1032 may be analyzed to detect one or more physical states of the subject, etc.

[0092] Additionally or alternatively, in some variations, the vibroacoustic sensor module may include a membrane-based deflection structure that interacts with or with one or more sensors across a cavity. For example, as shown in Figures 11A and 11B, a vibroacoustic sensor module 1100 may include a deflection structure 1120 having a cavity 1126 sealed by a membrane 1124 extending across a frame 1122. In some variations, the deflection structure 1120 includes a handle portion 1150 and may be handheld (and / or may be used to secure to another portion of a sensing device).

[0093] The membrane 1124 may be fabricated and attached in a manner similar to that described above with respect to FIGS. 10A and 10B. Additionally, FIGS. 11C-11E depict an exemplary rigid circuit board 1130 including one or more sensors 1132 and a cable connector 1134. Signals to and from the sensors 1032 may be transmitted via conductive traces extending between the sensors 1132 and the cable connector 1134. As shown in FIGS. 11A and 11B, the active circuit board 1130 may be positioned within the vibroacoustic sensor module 1100 such that the one or more sensors 1132 are disposed within or adjacent to the cavity (or are otherwise configured to be in fluid communication with the cavity, such as through an opening). For example, as shown in FIG. 11A, the sensor 1132 may be positioned opposite the membrane 1124. In such an arrangement, deflection of the membrane 1124 in response to the vibrational acoustic waves can cause changes in the cavity that are detectable and measurable by one or more cross-axis inertial sensors 1132. For example, in some variations, the one or more sensors 1132 can include a pressure sensor (e.g., a MEMS pressure sensor) that detects pressure variations in the cavity induced by deflection of the membrane 1124. For example, the pressure sensor can be read at a high rate (e.g., greater than about 500 Hz), and the pressure data can be used to form a low-frequency signal waveform (e.g., from about 0.01 Hz to about 1 kHz, or from about 0.01 Hz to about 0.5 kHz). As another example, in some variations, the one or more sensors 1132 can include a microphone (e.g., a MEMS microphone) that traverses the cavity and detects vibrational acoustic waves induced by deflection of the membrane 1124. The microphone can be configured to sense different bandwidths, such as between about 5 Hz and about 2 kHz-10 kHz.

[0094] Additionally, in some variations, the one or more sensors 1132 may include both a pressure sensor and a microphone, and / or any other suitable sensor (e.g., a voice coil transducer, a piezoelectric transducer, etc.).

[0095] In some variations, the vibroacoustic sensor module may include damping features to help isolate the sensing components from hand movements that may introduce noise and / or errors into the acquired vibroacoustic signal. For example, Figures 12A and 12B show one variation of a vibroacoustic sensor module 1200 similar to that described above with reference to Figures 10A and 10B in that the vibroacoustic sensor module 1200 may include a deflection structure 1220 having a cavity 1226 sealed by a membrane 1224 spanning a frame 1222. The deflection structure 1220 may include an annular handle portion 1250 indirectly coupled to the central frame 1212 via a flexible layer 1228 and / or other suitable flexible structure that helps isolate the sensing components (e.g., sensor 1232, membrane 1224, etc.) from fluctuations in hand pressure and other movements. The flexible layer 1228 may include, for example, a damping material such as a foam or elastomeric material and / or include damping structures such as radial ribs that help damp and / or decouple movement from the handle 1250 from the rest of the deflection structure. Damping can also be implemented in other "active" ways using one or more controlled mechanisms or materials such as actuators (e.g., micro- and / or servo motors), piezoelectric or electrosensitive polymers, etc.

[0096] Combination type vibration and acoustic sensor module Additionally, in some variations, the vibroacoustic sensor module may combine aspects of any of the variations described above, such as integrating multiple sensors for vibroacoustic detection within the same sensor module. For example, as shown in FIGS. 13A-13C , an exemplary variation of a vibroacoustic sensor module 1300 may include various components that enable integration of an accelerometer and a microphone for wider sensing bandwidth. For example, the vibroacoustic sensor module 1300 may include a deflection structure 1320 including a membrane 1324 that interacts with both an accelerometer 1342 coupled to the membrane 1324 via a rigid or semi-rigid coupling disk 1344. The membrane 1324 may be tensioned on and attached to a support ring 1326, which may be attached to a frame 1322. Furthermore, the coupling disk 1324 may be coupled to the membrane 1324 with, for example, epoxy or other suitable types of fasteners, and simultaneously coupled to a portion of a flexible circuit board 1340 that includes the accelerometer 1342. Similar to that described above with respect to the bent arm sensor module, movement of the membrane 1324 can be detected, measured and converted into a vibro-acoustic signal by an attached accelerometer 1342 .

[0097] Additionally, similar to that described above with respect to membrane-type sensor modules, the membrane 1324 may extend over a cavity and interact with a microphone sensor 1332 through the cavity. For example, the vibroacoustic sensor module 1300 may include a rigid circuit board 1330 having a microphone sensor 1332 configured to detect vibroacoustic signals transmitted across the membrane 1324 and the cavity. Additionally or alternatively, the rigid circuit board 1330 may include a pressure sensor and / or other suitable sensors (e.g., voice coil transducer, piezoelectric transducer, etc.).

[0098] In some variations, the combination sensor module may include a handle portion to allow for manual manipulation of the sensor module. For example, as shown in FIG. 14, vibroacoustic sensor module 1400 may be similar to vibroacoustic sensor module 1400, except that vibroacoustic sensor module 1400 includes an accelerometer and a microphone for detecting and measuring vibroacoustic waves and may include a handle portion 1450 having a flange-like gripping feature. However, other variations may include a handle portion having any suitable shape.

[0099] In some variations, the vibro-acoustic sensor module may include various components that enable the integration of an acceleration sensor and a pressure sensor. For example, as shown in FIG. 15, one exemplary variation of a vibro-acoustic sensor module 1500 may include one or more accelerometers 1532 (e.g., similar to the vibro-acoustic sensor module 900 described above with reference to FIGS. 9A-9D ) disposed on each bending arm 1524, and a membrane whose deflection is measurable by the pressure sensor 1532 (e.g., similar to that described above with reference to FIGS. 11A and 11B ).

[0100] Additionally, it should be understood that other variations of the vibrometer sensor module may include combinations of different aspects of the variations described above, such as to accommodate different types of sensors capable of providing vibroacoustic signals (e.g., for various bandwidths or frequency ranges) and / or for use with different sensing device form factors and applications.

[0101] Voice Coil Transducer As described above, in some variations, a vibroacoustic sensor module may include a voice coil transducer, alone or in combination with any of the sensors described above and / or other suitable sensors. For example, FIGS. 16A-16C show an exemplary variation of a voice coil transducer 1600. The voice coil transducer 1600 includes a frame 1610 (also referred to as a surround pot) having a cylindrical body 1620 with a bore 1630 and a flange 1640 extending radially outward from the cylindrical body 1620. The frame 1610 may be made of steel. A core 1650 of soft iron or other magnetic material is attached to the cylindrical body 1620 and aligns with the bore 1630 of the cylindrical body 1620. As can be seen, the iron core 1650 extends around the bore 1630 of the cylindrical body 1620 and also extends across the end 1660 of the cylindrical body 1620. The iron core 1650 has an open end. The magnet 1670 is positioned within the bore 1630 and is surrounded by and spaced apart from the iron core 1650, defining a magnet gap 1680. A voice coil 1690, including one or more layers of wire windings 1692 supported by a coil holder 1693, is suspended and centered with respect to the magnet gap 1680 by one or more spiders 1695. The wire windings 1692 may be made from a conductive material such as copper or aluminum. The periphery of the spider is attached to the frame 1610, and the center is attached to the voice coil 1690. The voice coil 1690 penetrates at least partially into the magnet gap 1680 through the open end of the iron core 1650. One or more spiders 1695 allow relative movement between the voice coil 1690 and the magnet 1670 while minimizing or avoiding torsional and in-plane movement. A diaphragm (not shown in FIG. 16) is provided that may be attached to the voice coil transducer 1600. This diaphragm may be the same as diaphragm 450. In the steady state, when no pressure is applied to the diaphragm, the voice coil 1690 may be positioned such that it is not fully received within the magnet gap (off-center with respect to optimal placement within the magnet gap).In use, the voice coil 1690 can be forced into and centered within the magnet gap when pressure is applied to the diaphragm under normal use.

[0102] A dust cap 1697 may be provided over the open end to prevent the ingress of foreign matter. The voice coil transducer 1600 of FIGS. 16A-16C may be incorporated into any of the sensing devices described herein, such as the sensing device 400 of FIGS. 4A-4F. Attaching the diaphragm to a portion of the voice coil transducer 1600 (such as the cylindrical body 1620) may be done by any suitable attachment means, such as adhesive. Alternatively, the diaphragm and voice coil 1690 may be fabricated as a single unit. In any of these variations, an outer cover (not shown) may be provided over the diaphragm to seal any openings between the diaphragm and the housing 1610. The outer cover may be made of an elastomeric material, such as rubber.

[0103] In use, sensing device 400 or sensing device 300 may be used to detect acoustic signals from a subject by coupling diaphragm 450 or the outer covering of sensing device 400 to the subject's skin or clothing, or by positioning the subject and sensing device 400 in close proximity to one another. Acoustic wave-induced movement causes diaphragm 450 to move, which in turn induces movement of voice coil 1690 within the magnet gap, resulting in an induced electrical signal.

[0104] In some voice coil transducer variations, the transducer configuration is arranged to pick up more quadrature signals than in-plane signals, thereby improving sensitivity. For example, one or more spiders are designed to have out-of-plane compliance and be stiff in-plane. The same applies to the diaphragm, whose material and stiffness may be selected to improve out-of-plane compliance. The diaphragm may have a convex configuration (e.g., a dome shape) that further helps to reject non-quadrature signals by deflecting them. Furthermore, signal processing can further derive any non-quadrature signals, for example, by using a three-axis accelerometer. This particularly allows for further rejection of non-quadrature signals or even for the non-quadrature signals passing through the sensor to derive the angle of origin of the incident acoustic wave.

[0105] It will be appreciated that different applications of sensing devices may require different sensitivities and face different noise-to-signal ratio challenges. For example, clothing-contact applications may require higher sensitivity and a higher signal-to-noise ratio than direct skin-contact applications. Similarly, non-contact applications may require higher sensitivity and a higher signal-to-noise ratio than contact applications.

[0106] Thus, to provide sensing devices with sensitivity and signal-to-noise ratios suitable for different form factors (e.g., contact or non-contact applications), developers have discovered that modulation of several variables can optimize a voice coil transducer for a particular intended application, these variables being magnet strength, magnet volume, voice coil height, wire thickness, number of windings, number of winding layers, winding material (e.g., copper vs. aluminum), and spider configuration. This is further described in Example 6.

[0107] In some variations, the voice coil 1690 is configured to have an impedance of greater than about 10 ohms, greater than about 20 ohms, greater than about 30 ohms, greater than about 40 ohms, greater than about 50 ohms, greater than about 60 ohms, greater than about 70 ohms, greater than about 80 ohms, greater than about 90 ohms, greater than about 100 ohms, greater than about 110 ohms, greater than about 120 ohms, greater than about 130 ohms, greater than about 150 ohms, or about 150 ohms. This is higher than conventional heavy magnet voice coil transducers, which have impedances of about 4 to 8 ohms. This is achieved by adjusting one or more of the number of windings, wire diameter, and winding layers in the voice coil. Many permutations of these parameters are possible and have been tested by the developers, as presented in Example 6. In one such variation, the voice coil is made of thin wire and configured to have an impedance of about 150 ohms and associated reduced power requirements by increasing the number of windings.

[0108] The developers have also discovered that adaptations to the configuration of the spider 1695 contribute to increased sensitivity and signal-to-noise ratio. More specifically, it has been determined through experimentation and simulation that making the spider 1695 more compliant, such as by incorporating apertures in the spider, increases sensitivity. The apertures also allow air to flow freely. These are described in further detail below with respect to Figures 17 and 18.

[0109] The use of voice coil-based transducers for these applications, including but not limited to those involving body contact and / or capturing sounds below the threshold of hearing, is counterintuitive. Voice coils are commonly used in audio speaker systems and are optimized for converting electrical energy into acoustic energy. To achieve useful sound pressure levels, these audio speaker voice coils must handle large amounts of power, ranging from 10 to 500 watts. The design considerations employed for this make them unsuitable for use in microphones or general sensing applications. Because power can be described by the equation P = IV = V² / R, low-resistance voice coils allow for high power handling at relatively low voltages, making them compatible with the power semiconductors typically used in audio amplifiers. In fact, most audio equipment manufacturers cite the amplifier's ability to drive low-impedance speaker loads as an advantage. While high-turn, high-impedance coils are efficient in terms of the force generated per unit current, the voltage required to drive such currents requires bulky insulation and hinders thermal management. Ferrofluid cooling is a possible solution, but the viscosity of such fluids reduces sensitivity. Of course, this is not an issue when high-power amplifiers are available. Therefore, low-impedance speakers, such as 8 Ω and 4 Ω, are relatively common. These feature heavy voice coils and magnet structures that accommodate the heavy windings these coils have. Noise can also be a factor, with temperature-induced thermal noise increasing as the impedance of the conductor / resistor increases.

[0110] Furthermore, to maintain reasonable efficiency at low frequencies around 20 Hz, woofers and subwoofers typically use very heavy cones, so voice coil mass is not a significant issue. In contrast, tweeters require light voice coils to achieve reasonable efficiency with air-diaphragm impedance matching using small diaphragms with wide frequency bandwidths, and are therefore very inefficient when operating at low frequencies. Tweeters typically have very light and delicate diaphragms, making them unsuitable for direct-contact microphones.

[0111] Also, due to the needs of two-way and three-way transducer speaker designs, a crossover circuit is usually required to achieve a wide frequency response for audio speakers operating between 20 Hz and 20 kHz.

[0112] However, conventional microphones typically operate under entirely different conditions, requiring low sound pressure levels to be picked up with minimal noise. As such, they are typically constructed with low-mass diaphragms, and the best microphones typically require an external power source because they operate as variable capacitors rather than as true voice coil / magnetic gap transducers. Again, just like tweeters, the delicate diaphragms of sensitive microphone designs are too fragile to be suitable for direct-contact microphones. They also suffer from low dynamic range and high natural resonant frequencies due to their method of operation.

[0113] Thus, the discovery that an adapted voice coil transducer could be used as a biosignal microphone was a surprising development by the inventors. In some variations of the present technology, it has been discovered that by adapting at least the voice coil and spider configuration of a conventional heavy magnet audio speaker, it is possible to achieve a microphone with higher sensitivity, wider frequency range detection capability, a dynamic and adjustable frequency range, and high signal-to-noise ratio characteristics. In some variations, the single voice coil transducer of the current technology can provide a microphone frequency response from below about 1 Hz to above about 150 kHz, or from about 0.01 Hz to about 160 kHz.

[0114] Furthermore, the use of such a vibroacoustic sensor module also enabled the size of the vibroacoustic sensor module to be reduced to the minimum practical size for handheld applications. This combination of modifications enabled the voice coil to generate relatively higher voltages in response to vibroacoustic signals than would be possible using the voice coil of a typical audio speaker. As a result, sensing of these voltages could be achieved with a low-noise J-FET-based amplifier, achieving a desired combination of frequency response, dynamic range, spurious signal rejection, and signal-to-noise ratio, for example.

[0115] In some variations of the present technology, the voice coil transducer 1600 comprises a single layer of spider 1695 (FIG. 16C). In some other variations of the present technology, the voice coil transducer 1600 comprises a double layer of spider 1695 (FIG. 16D). Multiple layers of spider 1695 are also possible, including but not limited to three, four, or five layers.

[0116] Several configurations of spider 1695 are illustrated in FIGS. 17 and 18A-18AB. As can be seen, instead of a monolithic, continuous wave configuration as known from conventional spiders in conventional voice coils, in some variations of the current technology, the spider 1695 has a discontinuous surface. The spider 1695 may include at least two deflection structures 1700 spaced apart from one another to allow airflow therebetween. In some configurations, the deflection structures 1700 include two or more spaced apart arms 1710 extending radially from a central portion 1720 of the spider 1695. In the variation illustrated in FIGS. 17A and 17B and 18B, the deflection structure 1700 includes four arms 1719 extending radially from the central portion 1720. The four arms 1710 increase in width as they extend outward. Each of the arms 1710 has a wave configuration. The opening 1730 between each of the arms 1710 is larger than the area of ​​each deflection arm.

[0117] FIGS. 18A-18AB show another variation of a spider 1800 for a voice coil transducer, such as voice coil transducer 1600. The spider comprises a deflection structure 1800 including one or more arms 1810 extending from a central portion 1820 and defining an opening 1830 therebetween. The one or more arms 1810 may be straight or curved. The one or more arms 1810 may have a width that varies along its length or is constant along its length. The one or more arms 1810 may be configured to extend helically from the central portion 1820 to a periphery 1840 of the spider 1800. The periphery 1840 of the spider 1800 may be provided with a solid ring, which is omitted from FIGS. 18A-18AB for clarity but is shown in FIG. 16E. In some variations, a single arm 1810 may be provided that is configured to extend spirally from a central portion 1820 of the spider 1800 to the periphery 1840 of the spider 1800. In these cases, the rotation of the spiral arm 1810 defines an aperture 1830. The spider 1800 may be defined as including a segmented form that includes portions that are solid (the arms 1810) and portions that have apertures 1830 defined therebetween. The arms 1810 may be the same or different (e.g., FIG. 18C). In variations in which multiple layers of spiders 1800 are provided within the voice coil transducer, the spiders 1800 of each layer may be the same or different.

[0118] The configuration selected for a given application of the sensing device will depend on the amount of compliance required for that given application. For example, a lower compliance voice coil configuration may be selected for contact applications over non-contact applications. In contact applications, the spider may be coupled to the voice coil in a manner that offsets the voice coil from the magnet gap when no pressure is applied to the diaphragm, and when the expected pressure is applied to the diaphragm, the voice coil is pushed into the magnet gap for optimal positioning and acoustic signal detection.

[0119] In some variations, the compliance of the diaphragm may be in the range of about 0.4 to 3.2 mm / N. The compliance ranges may be described as low, medium, and high as follows: 0.4mm / N: Low compliance -> fs centered around 80~100Hz, 1.3mm / N: Medium compliance -> fs centered around 130Hz, 3.2mm / N: High compliance -> fs centered around 170Hz.

[0120] In some variations, two or more voice coil sensors may be included in a sensing device (e.g., a vibro-acoustic sensor module) that may better enable triangulation of faint body sounds detected by the voice coil sensors and / or cancellation and / or filtering of noise, such as environmental disturbances. The sensor fusion data of the two or more voice coil sensors may be used to generate a low-resolution acoustic intensity image.

[0121] In some variations, the voice coil transducer may be optimized for vibro-acoustic detection, such as by using unconventional voice coil materials and / or winding techniques. For example, in some variations, the voice coil material may include aluminum instead of the traditional copper. Although aluminum has a low specific conductivity, its low mass may improve the overall sensitivity of the voice coil transducer. Additionally or alternatively, the voice coil may include more than two layers or winding levels (e.g., three, four, five, or more layers or levels) to improve sensitivity. In some variations, the wire windings may include silver, gold, or alloys for desired properties. Any suitable material may be used for the wire windings for the desired functionality. In some other variations, the windings may be printed on the diaphragm, for example, using conductive ink.

[0122] Advantageously, some variations of the present technology may be used as a standalone stethoscope, or as an add-on to a conventional acoustic stethoscope, or as an attachment to a smartphone or phonocardiograph device to detect infrasound-ultrasound vibroacoustic signals from a subject. The sensing device may have any suitable form factor for contact or non-contact vibroacoustic detection. The vibroacoustic sensor module of some variations of the present technology has advantages over conventional acoustic and electrical stethoscopes used to detect acoustic signals related to a subject.

[0123] First, the technology may be deployed in contactless applications such as remote monitoring, whereas conventional acoustic stethoscopes require contact with the subject's skin for proper sound detection.

[0124] Second, acoustic signals can be detected over a wide range with a good signal-to-noise ratio. In contrast, conventional acoustic stethoscopes convert the movement of the stethoscope's diaphragm into air pressure, which is then transmitted directly to the listener's ear via a tube. This results in poor sound volume and clarity due to poor acoustic energy transmission. Therefore, the listener hears the direct vibration of the diaphragm via the air tube.

[0125] Current technology has advantages over traditional electronic stethoscope transducers, which tend to be one of two types: (1) a microphone mounted behind the stethoscope diaphragm, or (2) a piezoelectric sensor mounted on or physically connected to the diaphragm.

[0126] A microphone mounted behind the stethoscope diaphragm picks up the sound pressure generated by the stethoscope diaphragm and converts it into an electrical signal. Because the microphone itself has a diaphragm, the acoustic transmission path includes or consists of the stethoscope diaphragm, the air inside the stethoscope housing, and finally the microphone diaphragm. The presence of two diaphragms and the air path between them can result in excessive ambient noise pickup by the microphone and in inefficient transmission of acoustic energy. This inefficient transmission of acoustic energy is a common problem with electronic stethoscopes, as described below. Existing electronic stethoscopes use additional technologies, such as adaptive noise cancellation for the microphone and various mechanical sound isolators, to counter this fundamentally inferior sensing technology. However, these only compensate for the inherent deficiencies of acoustoelectric transducers.

[0127] Piezoelectric sensors operate on a slightly different principle than simply sensing sound pressure on a diaphragm. They generate electrical energy through the deformation of a crystalline material. In some cases, the movement of the diaphragm deforms a piezoelectric sensor crystal mechanically coupled to the diaphragm, resulting in the generation of an electrical signal. The problem with this sensor is that the transduction mechanism can cause signal distortion compared to sensing pure diaphragm movement. The resulting sound is therefore somewhat different in timbre and distorted compared to an acoustic stethoscope.

[0128] Capacitive acoustic sensors are commonly used in high-performance microphones and hydrophones. Capacitive microphones perform acoustic-to-electrical transduction by utilizing the variable capacitance caused by a vibrating capacitive plate. Capacitive microphones placed behind a stethoscope diaphragm suffer from the same ambient noise and energy transfer problems as other microphones mounted behind a stethoscope diaphragm.

[0129] Acoustoelectric transducers operate on the principle of capacitance-to-electrical transduction, directly detecting the movement of a diaphragm and converting it into an electrical signal that is a measure of the diaphragm movement. Further amplification or processing of the electrical signal facilitates the production of an amplified sound with characteristics very similar to that of an acoustic stethoscope, but with increased amplification while maintaining low distortion.

[0130] This is a significant improvement over the more indirect diaphragm sound sensing produced by the microphone or piezoelectric approaches described above. Because the diaphragm movement is sensed directly, the sensor is less susceptible to external noise and the signal is a more accurate measurement of diaphragm movement. In an acoustic stethoscope, the diaphragm movement creates sound pressure waves that are sensed by the listener's ear. In an acoustoelectric sensor, that same diaphragm movement generates an electrical signal in a direct manner. The signal is used to drive an acoustic output transducer, such as an earphone or headphone, which sets up the same sound pressure waves that strike the listener's ear.

[0131] While acoustoelectric transducers overcome many of the inherent problems faced by previous stethoscope designs, they also add a significant amount of white noise to the signal. White noise is sound that contains equal amounts of all frequencies within the human hearing range (typically 20 Hz to 20 kHz). Most people perceive this sound as containing more high-frequency components than low-frequency components, but this is not the case. This perception occurs because each successive octave has twice the frequency of the preceding octave. For example, from 100 Hz to 200 Hz, there are 100 discrete frequencies. The next octave (200 Hz to 400 Hz) has 200 frequencies.

[0132] As a result, it is difficult for a listener to distinguish body sounds from white noise. For higher intensity body sounds (i.e., louder sounds), the listener can hear the body sounds well, but lower intensity sounds get lost in the background white noise. This is not the case in some variations of the present technology. Echo sensor-based vibroacoustic module

[0133] Variations of the system 10 or sensing device 110 may include one or more echo sensor-based vibro-acoustic modules, such as, but not limited to, one or more of continuous wave Doppler (CWD), pulsed wave Doppler (PWD), and time-of-flight based echo sensors.

[0134] Continuous Wave Doppler (CWD): A continuous ultrasound signal is emitted by a source oscillator, reflected by the subject, and returned to the receiver. Vibrations on the subject change the frequency / phase of the emitted ultrasound signal, allowing the original vibration signal to be retrieved. This gives a maximum sampling frequency of the subject under investigation.

[0135] Pulsed-Wave Doppler (PWD): Short ultrasound bursts are transmitted and the receiver waits for a response. This technique can resolve subject vibrations similar to CWD, but introduces a subject sampling frequency due to the burst interval. The Nyquist frequency of the corresponding sampling frequency is (pulses / second) / 2. Therefore, with one pulse per millisecond, the maximum resolvable subject vibration frequency is 500 Hz. However, PWD can resolve vibrations at a specific depth or distance from the emitter / sensor. This is achieved by considering the time-of-flight information of the pulse and rejecting signals outside the desired distance. Therefore, PWD can reject signals outside the target distance, signals generated by other sources, or emitted pulses traveling beyond the subject and reflected off walls.

[0136] Time of Flight: A simpler version compared to PWD is the pulsed ultrasound signal where only the time of flight is considered.

[0137] Advantageously, these echo-based modules can enable measurement of vibrations (such as vibroacoustic signals from a subject) as well as distance or velocity. The echo-based modules are non-contact, non-invasive, and non-harmful to the subject. The vibroacoustic signals can be detected from a distance of about 1 cm to about 10 m in some variations. The detection distance can be about 10 meters, about 9 meters, about 8 meters, about 7 meters, about 6 meters, about 5 meters, about 4 meters, about 3 meters, about 2 meters, or about 1 meter. Signal detection can be performed through the subject's clothing or other garments. Furthermore, signal detection can be obtained over a wide frequency band.

[0138] Echo-based acoustic systems are generally active systems that include an emitter component and a receiver component, relying on the receiver component to detect signals from a subject in response to radiated signals from an emitter incident on the subject. Thus, in some variations, radiated signals in the ultrasonic range are used, preferably above 25 kHz to maintain some headroom at the audible end of the spectrum (since radiated signals in the audible range are not preferred). At the higher end, the maximum value can be as high as 100 kHz, depending on the absorption of ultrasonic signals in air and the sampling rate of the ADC. At 50 kHz, acoustic absorption in air is approximately 1-2 dB / m, at 100 kHz, approximately 2-5 dB / m, at 500 kHz, approximately 40-60 dB / m, and at 1 MHz, approximately 150-200 dB / m. Technical challenges at higher frequencies relate to the ability to capture signals with sufficient quality, such as the availability of high-speed analog-to-digital converters.

[0139] The number of emitter and receiver components of the echo sensor module is not limited. Different combinations can be used, as will be described in more detail with reference to Figures 33D-33H. For example, a single emitter and a single receiver, or two emitter and a single receiver, or a single emitter and two receiver, or two emitter and two receiver components can be provided.

[0140] The emitter component can be any type of emitter configured to emit an ultrasonic signal. The emitter should probably be as unidirectional as possible. One example is the Pro-Wave Electronics 400ET / R250 Air Ultrasonic Ceramic Transducer.

[0141] The receiver component may be any receiver type configured to detect emitted ultrasound signals from the subject. In some variations, the receiver component may be a microphone capable of capturing ultrasound signals with a sufficient signal-to-noise ratio. This may include any type of microphone, such as a condenser microphone, a dynamic microphone, or a MEMS microphone. In some variations, ultrasound-enabled MEMS microphones are preferred due to their compactness. In other variations, the receiver component is a dedicated ultrasound receiver tuned to the frequency. In some variations, the receiver component is as unidirectional as possible. Examples of receiver components include the Pro-Wave Electronics 400ST / R100 Transducer, the Pro-Wave Electronics 400ST / R160 Transducer, or the Invensense ICS-41352.

[0142] 1.c.iii. Other Sensor Examples According to some variations, system 100 may include sensor modules other than the vibro-acoustic sensor module described above. One or more other sensor modules may be incorporated within the housing of the sensing device or may be separate and connected thereto.

[0143] Bioelectric-based sensor module In some variations of the system 10, such as the sensing device 400, a bioelectric-based sensor module is also provided. In some variations, the bioelectric-based sensor module is configured to detect electrical impulses on the subject's skin. These may be representative of electrical impulses in the nerves of the subject's cardiac tissue. The bioelectric-based sensor module can therefore function as an ECG module and is therefore referred to herein as an ECG sensor module. The ECG sensor module can indicate physical conditions such as trauma to the cardiac nerve tissue network, damage to the cardiac tissue, such as from a previous heart attack or infection, severe nutritional imbalances, or stress due to excessive physiological or psychological pressure.

[0144] For example, as shown in FIG. 4B, in some variations, the sensing device 400 of the sensing module comprises a capacitive sensor electrode, such as an electric potential integrated circuit (EPIC) electrode 460 and a driving right leg (DRL) electrode 470.

[0145] Figure 19 illustrates the basic circuit of the ECG sensor module, the operation of which can be explained as follows.

[0146] Dipoles are the elementary units of cardiac activity. Each dipole consists of a positive (+) and a negative (-) charge generated by the action of ion channels. As activation spreads, the sources add together and act as a continuous layer of sources. Simply put, an electric dipole consists of two particles with equal and opposite charges separated by a short distance. The charged particles in the heart are sodium (Na - ), potassium (K+ ), calcium (Ca 2+ ), phosphate (PO4 3- ), proteins, and other ions. The separation distance is the distance across the cardiac muscle cell membrane. Negatively charged particles remain inside the cell because they are too large to pass through the small membrane channels, but positive ions move back and forth through specific channels and "ion pumps," creating polarization and depolarization across the membrane.

[0147] When enough dipoles are present together, they produce a measurable voltage. Resting cardiac cells within the heart are normally at a potential of -70 mV. This means that a charge imbalance naturally exists within the heart at rest. This imbalance, called cell polarization, attracts positive ions to the cell's interior. When cardiac cells are activated by an external stimulus, channels in the cell membrane activate, allowing excess positive ions outside the cell to surge inside. This process, called depolarization, makes the cell less negatively charged and is associated with the "activation" of cardiac cells. When millions of these cells activate together, the heart contracts, pumping blood to the rest of the body. The combined activation of these cells generates enough voltage to be measured at the skin surface by an electrocardiogram (ECG). The resulting intracardiac electrogram (EGM) extends beyond the area of ​​the dipole signal by as much as five times, reducing resolution and clarity.

[0148] A variation of the ECG sensor module of the present technology includes one or more bioelectric sensors for measuring electric fields and electrical impulses.

[0149] In the case of tight coupling (Cext>>Cin), this is usually defined by Equation 1:

[0150]

number

[0151] where: a = equivalent shared electrode / target area d = distance between target and sensor ε0 = permittivity of free space ε r = relative permittivity of the dielectric in which the sensor is operating

[0152] For remote coupling (Cext>>Cin), we have the limiting case (self-capacitance) shown in Equation 2 below. C ext -8ε0ε r r formula 2 where r is the diameter of the sensor plate.

[0153] Analysis of this circuit shows that it has a classic single-pole transfer function, as shown in Equation 3.

[0154]

number

[0155] The corner frequency (Fc1) can be expressed by the following equation 4:

[0156]

number

[0157] The input capacitance of the ECG sensor module is 10 -17 The input resistance can be as low as 10 15 Ω, thus keeping interaction with the target field to an absolute minimum and ensuring that all currents are small displacement currents only. in and R in Controlling the value of gives the effective value, allowing control of both the gain plateau and corner frequency (Fc1 transitions to Fc2). The response of the ECG sensor module was optimized via a staged design and a positive feedback loop.

[0158] The ECG sensor module can be used as an alternative to traditional wet electrode ECG pads because it does not require gel or other contact promoters. Once the ECG sensor module is attached to (or brought close to) the patient, the ECG signal can be restored. The sensor can perform simple electrocardiogram "monitoring" or even more rigorous clinical screening and prediagnostic measurements. In infectious disease status diagnostic applications, such as for COVID-19, the ECG sensor module can be used as an alternative to traditional 12-lead ECGs, with electrodes attached to the limbs and torso to provide a clearer picture of how the patient's heart is working. A four-lead array of potential sensors can be used in the ECG sensor module, recreating each and every 12-lead ECG trace required with resolution comparable to or better than that achieved using traditional systems. Figure 20 shows a comparison of results using a modified ECG sensor module with results using traditional wet electrodes on leads II and VL. The top panel shows the control, while the bottom panel shows the experimental.

[0159] Next, reference will be made to the DRL electrodes and their operation with reference to FIGS. 4B-4E, 21 and 22. FIG.

[0160] Driven Right Leg (DRL) is a technology used in conventional ECG systems to reduce noise pickup by ECG sensors. DRL is a differential electronic technique for improving spurious signal rejection and signal-to-noise ratios in bioelectric signal acquisition. Typically, this technology is used during ECG procedures and involves attaching at least one skin-contacting electrode to the subject's lower leg, as shown in FIG. 21 . However, existing technologies cannot be used through clothing and typically require the patient to be sitting or lying down to apply at least the skin-contacting electrode, making them impractical for use with ambulatory patients. Therefore, developers have developed current ECG sensor modules that, in some variations, do not require attachment to the subject's leg while improving spurious signal rejection and signal-to-noise ratios. In some variations, the ECG sensor module includes at least one EPIC sensor and at least one DRL sensor.

[0161] According to some variations on current technology, a prototype circuit diagram for at least one DRL sensor is shown in Figure 22A. Figure 22B shows an experimental setup with five DRL electrodes spaced between and around two EPIC electrodes, along with optional guard electrodes. The guard electrodes provide a ground, which is important when there is direct skin contact. They provide a signal reference to the system ground when there is direct skin contact; otherwise, the ECG signal would be undetectable or very noisy. As a three-dimensional structure, the guard electrodes can also function as a Faraday cage and provide EMI shielding. When there is no direct skin contact (such as in clothing applications), it is the DRL that provides the reference through capacitive coupling and an active feedback signal. The guard electrode then loses its grounding purpose but still provides EMI shielding. This may enable switching from DRL to non-DRL function. In the experimental setup, the two EPIC electrodes were spaced approximately 46 mm apart. The width of the DRL sensor between the two EPIC sensors was approximately 21 mm. Each DRL sensor surrounds each EPIC sensor by 15mm 2 The outer dimensions were:

[0162] The gain and phase response of the full circuit model of the DRL feedback circuit shown in Figure 22E was compared to SPICE ("Simulation Program with Integrated Circuit Emphasis") simulations using various numbers of 0.58 mm fabric separators, and the results are shown in Figures 22C and 22D, showing satisfactory results with four layers of fabric between the sensor and the skin, even at 200 kHz. The DRL feedback circuit incorporates a TWIN-T filter, a bioelectric sensor model, a 60 Hz noise source, and a simulated ECG signal.

[0163] Without being bound by any theory, EPIC capacitive sensors generally require some kind of ground reference to be able to pick up body electrical signals without being dominated by noise. Such noise is often dominated by 50 / 60 Hz power line interference. With contact EPIC setups, meaning the ECG sensor is in direct contact with the skin, this issue is less relevant because a conductive ground electrode can provide the necessary signal reference to the body. However, when attempting to use EPIC sensors in a non-contact setup, the missing conductive path to the body poses a problem. This can be solved by using DRL feedback to the body, either through direct conductive or non-contact capacitive electrodes. Because EPIC technology focuses on capacitive (non-galvanic) measurement of body electrical signals, it makes the most sense to also feed back the DRL signal capacitively. This allows for DRL electrodes placed directly on the stethoscope near the EPIC sensor to be integrated with the ability to measure electrical signals without forming a conductive path to the skin through clothing or other obstructions.

[0164] DRL works by feeding back an amplified signal that is the sum or other suitable combination (i.e., weighted sum, difference, sum of only a subset of sensors, etc.) of two or more EPIC sensor signals. The amplification requires proper tuning to cancel noise picked up by the EPIC sensors, but can be implemented as an automatic algorithm. Also, there can be many combinations of which and how many EPIC sensor signals to combine to generate a DRL signal for each available DRL electrode. This process can be automated as well.

[0165] In some variations, the EPIC and DLR electrodes required to perform an ECG may advantageously be spaced within 5 cm or less of each other. Furthermore, in some variations, the relative placement of the electrodes advantageously provides increased spurious signal rejection and a higher signal-to-noise ratio in the signal compared to where the DRL electrodes are placed further away from each other, as is typical in existing ECG tests. Bioimpedance-based sensor module

[0166] Bioimpedance-based sensor modules can be used to detect skin potentials. Electrodermal activity (EDA) is a marker of sympathetic nervous network activity. Electrodermal activity is generated by sweat glands and the overlying epidermis and mediated by the supraspinal column, including the orbitofrontal cortex, posterior hypothalamus, dorsal thalamic nucleus, and ventral bilateral reticular formation. This response occurs spontaneously and can be triggered by stimuli such as breathing, coughing, loud noise, startle, mental stress, and electrical stimulation. It is referred to as the sympathetic skin response or peripheral autonomic surface potential. EDA consists of two components: (1) a phasic component of the skin conductance response (SCR) observed after activation of sweat gland motor fibers, and (2) the skin conductance level (SCL), which corresponds to the subject's inherent baseline skin conductance. The SCR is a powerful component used as a marker of the sympathetic nervous network. Its frequency range is 0.05 to 1.5 Hz. An EDA active recording electrode is placed on the palmar or plantar surface, and an indifferent electrode is placed on the palmar surface. With low-pass filter settings of 0.1–2 Hz and high-pass filter settings of 1–5 kHz, upper limb latencies ranged from 1.3 to 1.5 seconds, and lower limb latencies ranged from 1.8 to 2.1 seconds.

[0167] While ECG and related techniques measure bioelectrical signals originating within a subject's body, bioimpedance measurements, such as galvanic skin response, examine a subject's resistance or response to an applied electromotive force. Galvanic skin response measurements have traditionally been direct current measurements requiring silver / silver chloride electrodes. The collective term for DC and AC measurements is electrodermal activity (EDA), and for AC-only measurements, bioimpedance. An advantage of AC measurements over DC measurements is that dry electrodes may be used. In some variations of the present system 10, non-contact electrodes are possible because only an AC potential needs to be applied to the body to induce an AC current in it.

[0168] The issue of electrode impedance is addressed by employing a four-wire measurement topology that largely eliminates the effect of electrode impedance on the measurement. Two drive electrodes pass the signal through the body, and two sense electrodes measure the differential voltage. Impedance is calculated using Ohm's law: Impedance = Sense Voltage Drive Current. A capacitor blocks DC current from flowing. R ACCESSX represents the resistance of the wire and electrode in conventional contact measurement. R LIMIT is the safe current limit, intended to stay within safety requirements in the event of a hardware or software failure.

[0169] A typical wet electrode bioimpedance analysis application diagram is shown in Figure 22F.

[0170] An example of a sensor for bioimpedance is the AD5940 (Analog Devices, Inc., Norwood, Massachusetts, USA) (Figure 22G). This chip can perform AC measurements from sub-Hz up to 200 kHz. However, the application notes for this device do not allow for the use of dry electrodes, or non-contact electrodes, or measurements through clothing.

[0171] Disclosed herein are devices, methods, and systems for measuring electrodermal activity (EDA) and body impedance analysis (BIA) using electro-particle integrated circuit (EPIC) sensors (such as those from Plessey Semiconductors Ltd., Roborough, Plymouth, Devon) that allow for contactless, distance, and through-clothing measurements. Some EPIC sensors used in the present systems and devices may include one or more of those described in U.S. Patent Nos. 6,213,999; ... and 6,213,999, the contents of which are incorporated herein by reference. A schematic diagram is shown in FIG. 22H.

[0172] In some variations, the AD5940 functions as described, with the touch sensing electrodes replaced by an EPIC sensor. DRL electrodes may also be included, similar to the ECG, to prevent or minimize sensor saturation in 50 / 60 Hz environments. Capacitive coupling to the DRL electrodes and to ground can affect measurements. Keeping the sensing electrodes separate from the DRL electrodes minimizes this source of error. The output of the AD5940 is amplified to couple sufficient bioimpedance drive current into the body through a small capacitance. The EPIC sensor pads its output down with a gain of 10 to prevent overloading the AD5940 input. The DRL filter and amplifier are the same as those described herein.

[0173] In some cases, non-contact DRL electrodes may be used, while in other cases, a transconductance amplifier such as the LM13700 OTA, Texas Instruments, Inc., Dallas, Texas, may be used to mitigate variations in drive electrode capacitance.

[0174] Example of bioimpedance analysis using EPIC sensors The drive electrode of the AD5940 was in direct skin contact. Electrode simulations were performed and heatmap displays were generated. Figures 22I and 22J show the heatmap results of two simulations. The highest voltage (red) is the voltage at the bioimpedance drive electrode. The lowest voltage (green) is 0 volts at the bioimpedance current sense input electrode. The voltage difference between the EPIC sensors was extracted in the simulations. In these figures, the difference between the active and floating DRL electrodes is shown. Grounding the DRL electrodes significantly reduces the sensed voltage. Therefore, in initial experiments, these electrodes were left floating.

[0175] Figure 22K shows the test setup, consisting of an AD5940 evaluation board, a pair of EPIC sensors, DRL circuitry, and a (floating) battery power supply for the sensors. The electrodes were machined from a 0.8 mm thick PCB, with copper milled between the electrodes and an opening milled for the EPIC sensors. Various test configurations were created to verify system performance. In Figure 22K, a pair of conductive cloth squares are connected by a resistor to form a repeatable "skin" resistance.

[0176] An alternative "skin," shown in Figures 22L and 22M, uses conductive paper (Eisco Labs PH0918DFM). This paper is used as a teaching aid to visualize the electric field; current is sent through the paper and voltage is measured along its surface. Gradual layers of Kapton tape were placed on the paper over the sensor area of ​​the EPIC. These layers were used to compare measurements with various distances to the sensor. Figure 22L shows the test configuration used for the conductive paper and mesh "skin," in which the skin was firmly clamped to the sensor through compliant foam. The paper was clamped directly to the driving electrode. As a result of the initial measurements, two changes were made to the system.

[0177] The EPIC sensor was a PS25014A5 model with a low cutoff frequency of 30 Hz (not a limiting factor for bioimpedance). The DRL electrode was connected to a bioimpedance-driven electrode, which increased the contact area and improved the consistency of the measurements.

[0178] Figures 22N and 22O show impedance frequency scans of the three "skins." 1.Conductive paper 2.Conductive mesh 3. Forearm

[0179] Paper had the highest resistance of approximately 15 kΩ, mesh had a resistance of 1340 Ω, and forearm was much lower, in the range of 220-30 Ω.

[0180] The phase plot shown in Figure 22O shows that the actual "skin"—the forearm—has a large reactive component. The impedance of the forearm is lower than expected. This can be attributed to the large area for the driven electrodes. The measured impedance of the 2 kΩ mesh is about 20% lower. This reflects the lower-than-expected gain of the EPIC sensor.

[0181] The effect of spacing between the skin and the electrodes was tested using conductive paper (these tests were performed before discovering how low the impedance of the forearm was). These results are shown in Figure 22P. The larger the spacing, the lower the apparent impedance. This is a result of the EPIC sensor measuring lower signal levels at larger spacings.

[0182] After consistent measurements were achieved on the synthetic "skin," testing on the forearm showed that the typical impedance was much lower. Typical body fat bioimpedance measurements are performed with widely spaced electrodes, resulting in high impedance. Lower impedances result in lower signal levels, but repeatable measurements are well within the realm of possibility.

[0183] Even in the absence of a DRL signal, the AD5940 appeared undisturbed by 60 Hz pickup. Between measurements, the EPIC sensor exhibited high levels of 60 Hz pickup. However, once measurements began, the 60 Hz noise level dropped dramatically. This can be attributed to the drive electrodes presenting a low impedance when active. The measurement method used by the AD5940 is also tolerant to 60 Hz interference.

[0184] We have developed a software-defined, multimodal sensor and data fusion platform for improving data capture and low-power / weak biosignals from diverse sensing modalities that models the intertwining of the brain, five senses, central nervous system, and autonomic nervous system. This involves using vibroacoustic sensors, potential sensors, volatile organic compound (VOC) electronic noses, thermal and optical sensors, and cameras in the same way that humans intuitively understand their surroundings using their noses, eyes, ears, and touch. Our goal is to develop a sensory modular platform in which the entire signal chain, from electromechanical actuation at the cellular / tissue level to mechanoacoustic transduction and vibroacoustic biosignal data (various audible and inaudible sounds from the human body), can be used to assess health. For example, a cough can signal many things depending on its duration, intensity, frequency, etc. There are over 64 human illnesses and diseases caused by the movement of air and / or bodily fluids. We have successfully developed a method for the automatic recognition of respiratory diseases such as COVID-19, pneumonia, asthma, cystic fibrosis, and chronic obstructive pulmonary disease (COPD). Additionally, vibroacoustic features such as articulation, effort, and acoustic roughness, as well as vowel pronunciation and other speaking styles, may provide clues to an individual's health status.

[0185] Context Sensor Module Little is known about what happens in real life, how lifestyles and daily circumstances affect vital signs, how quality of life is affected by illnesses and medical conditions, and the extent to which treatment and care recommendations are actually followed. Developers have determined that placing health data and care within the context of everyday life can, in some variations, add important insights for richer, more personalized interpretations of biosignals, vital signs, and well-being. In some variations, system 100 can further comprise one or more sensors that provide environmental and / or other contextual data (e.g., social determinants of health). This can be used to calibrate and / or better interpret vibroacoustic data acquired using either the vibroacoustic sensor module or other sensor modules. Such data (e.g., environmental and / or social determinants of health) can help, for example, to contextualize data for more accurate machine learning and / or AI data analysis. For example, as shown in FIG. 3B, in some variations, sensing device 300 can include a contextual sensor module 330. Similar to the vibroacoustic sensor module 320, the context sensor module 330 may be interchangeable and modular, such as for use in a variety of different sensing device form factors, including an integrated handheld sensing device such as that shown in FIG. 3B. The context sensor module 330 may be in communication with one or more processors (e.g., within the electronics system 340) such that sensor data from the context sensor module 330 may be taken into account when analyzing the vibroacoustic data and / or other suitable data.

[0186] The context sensor module 330 may include one or more suitable sensors, such as an environmental sensor, for measuring one or more ambient characteristics and / or one or more characteristics of the sensing device relative to the environment. For example, the context sensor module 330 may include an ambient light sensor, an ambient humidity sensor, an ambient pressure sensor, an ambient temperature sensor, an air quality sensor (e.g., for detecting volatile organic compounds (VOCs)), an altitude sensor (e.g., a relative pressure sensor), a GPS, and / or other suitable sensors to characterize the environment in which the sensing device is operating. Additionally or alternatively, the context sensor module 330 may include an inertial measurement unit (IMU), a separate gyroscope and / or accelerometer, and / or other suitable sensors to characterize the sensing device relative to the environment.

[0187] Acoustic Cardiography (ACG) Sensor Module In some variations, the system 100 may further include one or more sensor modules for detecting cardiac vibrations as blood moves through the various chambers, valves, and great vessels using an acoustic cardiography sensor module. The ACG sensor module can record these vibrations in four cardiac configurations, providing a "graph signature." While the opening and closing of the heart valves contribute to the graph, the contraction and strength of the heart muscle also contribute to the graph. The result is a dynamic picture of the heart's movement. If the heart is working efficiently and without stress, the graph will be smooth and clear. If the heart is working inefficiently, there will be clear patterns associated with each type of contributing dysfunction. An ACG is not the same as an ECG, a common diagnostic test. An electrocardiogram (ECG) records electrical impulses as they appear on the skin as they pass through the nerves in the heart tissue. An ECG exclusively indicates whether the heart's nervous tissue network is affected by some kind of trauma, injury (e.g., from a previous heart attack or infection), severe nutritional imbalance, or stress from excessive pressure. Only effects on the nervous system are detected. It doesn't tell us how well muscles or valves are functioning. Plus, ECGs are used exclusively to diagnose disease. Because ECGs look at cardiac function as well as electrical function, they act as a window into the entire nervous system and muscle metabolism. Using the heart allows us to see in "real time" how nerves and muscles are working together. This interaction results in unique, objective insight into the health of the heart and the human body.

[0188] Passive Acoustic Cerebral Graphic (ACG) Sensor Module In some variations, the system 100 may further include one or more passive acoustic cerebrograhography sensor modules for detecting blood circulation within brain tissue. This blood circulation is influenced by blood circulation within the brain's vasculature. Each heartbeat causes blood to circulate within the skull in a recurring pattern according to the resulting oscillations. The effectiveness of these oscillations, in turn, depends on the size, shape, structure, and vasculature of the brain. Thus, each heartbeat induces minute movements within brain tissue and even cerebrospinal fluid, causing slight changes in intracranial pressure. These changes can be monitored and measured within the skull. The one or more passive acoustic cerebrograhography sensor modules may include passive sensors, such as accelerometers, to correctly identify these signals. Sometimes, highly sensitive microphones may be used.

[0189] Active Acoustic Cerebral Grafting (ACG) Sensor Module In some variations, the system 100 may further include one or more active acoustic cerebral grafting (ACG) sensor modules. Active ACG sensor modules can be used to detect multi-frequency ultrasound signals and classify adverse changes at the cellular or molecular level. In addition to all of the benefits offered by passive ACG sensor modules, active ACG sensor modules can also perform spectral analysis of received acoustic signals. These spectral analyses can reveal changes in the brain's vasculature as well as changes in cellular and molecular structure. Active ACG sensor modules can also be used to perform transcranial Doppler examinations, optionally with a color display. These ultrasound examination procedures can measure blood flow velocity in the brain's blood vessels. These can diagnose embolism, stenosis, and vasoconstriction, which occur, for example, in the aftermath of subarachnoid hemorrhage. Ballistocardiography (BCG) Sensor Module

[0190] In some variations, the system 100 may further include one or more ballistocardiography sensor modules (BCGs) for detecting impulse forces caused by the heart. The downward movement of blood through the descending aorta creates an upward bounce, moving the body upward with each heartbeat. As various portions of the aorta expand and contract, the body continues to descend and ascend in a repeating pattern. Ballistocardiography is a technique that creates a graphical representation of the repetitive movements of the human body caused by the rapid pumping of blood into the great vessels with each heartbeat. This is a vital sign in the 1-20 Hz frequency range that is caused by the mechanical movement of the heart and can be recorded noninvasively from the body's surface. Major cardiac dysfunctions can be identified by observing and analyzing the BCG signal. BCGs can also be monitored noncontact using a camera-based system. One example of the use of BCGs is a ballistocardiogram scale, which measures the bounce of a person's body as they step on the scale. The BCG scale can display not only weight but also a person's heart rate.

[0191] Electromyogram (EMG) Sensor Module In some variations, the system 100 may further include one or more electromyography (EMG) sensor modules for detecting electrical activity generated by skeletal muscles. The EMG sensor module may include an electromyograph for generating recordings called electromyograms. Electromyographs detect electrical potentials generated by muscle cells when they are electrically or neurologically activated. These signals may be analyzed to detect medical abnormalities, activation levels, or recruitment sequences, or to analyze the biomechanics of human or animal movements. EMG may also be used in gesture recognition. Electro-oculography (EOG) sensor module

[0192] In some variations, the system 100 may further include one or more electro-oculography (EOG) sensor modules for measuring the corneal-retinal orienting potentials present between the front and back surfaces of the human eye. The resulting signal is called an electro-oculogram. Its primary applications are ophthalmic diagnostics and recording of eye movements. Unlike electroretinography, EOG does not measure responses to individual visual stimuli. To measure eye movements, a pair of electrodes is typically placed above and below the eye or on either side of the eye. When the eye moves from a central position toward one of the two electrodes, this electrode "sees" the positive side of the retina, and the opposite electrode "sees" the negative side of the retina. This results in a potential difference between the electrodes. Assuming a constant resting potential, the recorded potential is a measure of eye position.

[0193] Electroolfactography (EOG) sensor module In some variations, system 100 may further include one or more electroolfactography (EOG) sensor modules for detecting the subject's sense of smell. The EOG sensor modules can detect changing electrical potentials in the olfactory epithelium in a manner similar to how other forms of electrograms (such as ECG, EEG, and EMG) measure and record other bioelectrical activity. Electroolfactography is closely related to electrotactile recording, which is the electrical recording of insect antennal olfaction.

[0194] Electroencephalography (EEG) Sensor Module In some variations, system 100 may further include one or more electroencephalography (EEG) sensor modules for electrophysiological detection of the brain's electrical activity, or a vibroacoustic sensor module placed on a skull-mounted anechoic chamber to "listen" to the brain and capture subtle pressure and pressure gradient changes related to the sound processing circuitry. EEG is typically noninvasive, with electrodes placed along the scalp, although invasive electrodes are sometimes used, as in electrocorticography. EEG measures voltage fluctuations resulting from ionic currents within the brain's neurons. Clinically, it refers to recording the brain's spontaneous electrical activity over a period of time, such as recorded from multiple electrodes attached to the scalp. Diagnostic applications focus on either event-related potentials or the spectral content of EEG. The former examines potential fluctuations time-locked to events such as "stimulus onset" or "button press." The latter analyzes the types of neural oscillations (commonly referred to as "brain waves") that can be observed in EEG signals in the frequency domain. EEG can be used to diagnose epilepsy, which can cause abnormalities in EEG readings. It can also be used to diagnose sleep disorders, depth of anesthesia, coma, brain damage, and brain death. EEG, along with magnetic resonance imaging (MRI) and computed tomography (CT), can be used to diagnose tumors, stroke, and other focal brain disorders. Advantageously, EEG is an available, mobile technology and offers millisecond-range temporal resolution not possible with CT, PET, or MRI. Derivative techniques of EEG include evoked potentials (EPs), which involve averaging EEG activity time-locked to the presentation of some type of stimulus (visual, somatosensory, or auditory). Event-related potentials (ERPs) refer to EEG responses time-locked and averaged to more complex stimulus processing.

[0195] Ultra-wideband (UWB) sensor module In some variations, the system 100 may further include one or more ultra-wideband sensor modules (UWB, also known as ultra-wideband or ultraband). UWB is a wireless technology that can use very low energy levels for short-range, high-bandwidth communication across a large portion of the radio spectrum. UWB has traditionally had applications in non-cooperative radar imaging. Recent uses target sensor data collection, high-precision location, and tracking applications. A significant difference between traditional wireless communications and UWB is that traditional systems transmit information by varying the power level, frequency, and / or phase of a sinusoidal wave. UWB transmits information by generating radio energy at specific time intervals, occupying a wide bandwidth, thereby enabling pulse-position modulation or time modulation. Information can also be modulated onto the UWB signal (pulse) by encoding the polarity of the pulse and its amplitude and / or using orthogonal pulses. UWB pulses can be transmitted sporadically at a relatively low pulse rate to support time or position modulation, but can also be transmitted at rates up to the inverse of the UWB pulse bandwidth. Pulsed UWB systems, using a continuous stream of UWB pulses (Continuous Pulse UWB or C-UWB), have been demonstrated with channel pulse rates exceeding 1.3 gigapulses per second, supporting forward error correction encoded data rates exceeding 675 Mbit / s.

[0196] A valuable aspect of UWB technology is that UWB wireless systems can determine the "time of flight" of transmissions at various frequencies. This helps overcome multipath propagation, since at least some of those frequencies have line-of-sight trajectories. In cooperative symmetric two-way positioning techniques, distance can be measured with high resolution and precision by compensating for local clock drift and stochastic inaccuracies.

[0197] Another feature of pulse-based UWB is that the pulses are so short (less than 60 cm for a 500 MHz wide pulse, and less than 23 cm for a 1.3 GHz wide pulse) that most signal reflections do not overlap with the original pulse, eliminating multipath fading for narrowband signals. However, there is still multipath propagation and inter-pulse interference for fast pulse systems, which must be mitigated by coding techniques.

[0198] Ultra-wideband is also used in "see-through-the-wall" high-precision radar imaging techniques, high-precision localization and tracking (using distance measurements between radios), and high-precision time-of-arrival based positioning approaches, which have a spatial capacity of approximately 1013 bits / s / m 2 UWB radar has been proposed as an active sensor component for automatic target recognition applications, designed to detect humans or objects that have fallen onto subway tracks.

[0199] Ultra-wideband pulse Doppler radar can also be used to monitor human vital signs, such as heart rate and respiration signals, as well as human gait analysis and fall detection. Advantageously, UWB consumes less power and has a higher-resolution range profile than continuous-wave radar. However, its low signal-to-noise ratio makes it vulnerable to errors.

[0200] In the United States, the Federal Communications Commission (FCC) defines ultra-wideband as a radio technology with a bandwidth exceeding 500 MHz or 20% of the arithmetic center frequency, whichever is less. The February 14, 2002, FCC Report and Order authorized unlicensed use of UWB in the 3.1 to 10.6 GHz frequency band. The FCC power spectral density emission limit for UWB transmitters is -41.3 dBm / MHz. This limit also applies to unintentional radiators in the UWB band ("Part 15" limit). However, UWB transmitter emission limits may be significantly lower (down to -75 dBm / MHz) in other segments of the spectrum. Deliberations at the International Telecommunications Union Radiocommunication Sector (ITU-R) resulted in the publication of a Report and Recommendation on UWB in November 2005. The UK regulatory authority, Ofcom, issued a similar decision on August 9, 2007. More than 48 devices have been certified under the FCC's UWB rules, the majority of which are radar, imaging, or location systems.

[0201] Concerns have been raised about interference between narrowband and UWB signals that share the same frequency band. Previously, the only radio technology using pulses was the spark transmitter, which was banned by international agreement because it interfered with medium-wave receivers. However, UWB uses less power. This subject was widely discussed in the proceedings that led to the adoption of FCC regulations in the United States, and in the ITU-R UWB-related meetings that led to reports and recommendations on UWB technology. Commonly used electrical appliances radiate impulse noise (e.g., hair dryers), and proponents have successfully argued that widespread deployment of low-power wideband transmitters would not result in excessively high noise levels.

[0202] Psychocardiography (SCG) Sensor Module In some variations, system 100 may further include one or more psychocardiographic (SCG) sensor modules for noninvasively measuring cardiac vibrations transmitted by the heart to the chest wall during movement. SCG was first introduced around the mid-20th century. Several promising clinical applications have been proposed. These include observing changes in SCG signals due to ischemia and using SCG in cardiac stress monitoring. The origins of SCG date back to the 19th century, when scientists reported observing heartbeats while standing on a scale. Although SCG has not generally been deployed in clinical settings, several promising applications have been proposed. For example, SCG has been proposed to be valuable in assessing the timing of various events in the cardiac cycle. These events could be used to assess, for example, myocardial contractility. It has also been proposed that SCG can provide sufficient information to calculate heart rate variability estimates. A more complex application of cardiac cycle timing and SCG waveform amplitude is calculating respiratory information from SCG.

[0203] Intracardiac electrogram (EGM) sensor module In some variations, system 100 may further include one or more intracardiac electrogram (EGM) sensor modules for noninvasively measuring cardiac electrical activity generated by the heart during movement. This provides a record of changes in electrical potentials of specific cardiac trajectories as measured by electrodes placed within the heart via a cardiac catheter. This is used for trajectories that cannot be assessed by body surface electrodes, such as the His bundle or other areas within the cardiac conduction system.

[0204] Pulse Plethysmograph (PPG) Sensor Module In some variations, system 100 may further include one or more pulse plethysmograph (PPG) sensor modules for noninvasively measuring vascular congestion dynamics. The modules may use a single wavelength of light, or multiple wavelengths of light, including far-infrared, near-infrared, visible light, or ultraviolet light. In the case of ultraviolet light, the wavelengths used are between approximately 315 nm and 400 nm, and the modules are intended to irradiate the subject with less than 8 milliwatt-hours per square centimeter per day during their operation.

[0205] Galvanic Skin Response (GSR) Sensor Module In some variations, system 100 may further include one or more galvanic skin response (GSR) sensor modules, which may utilize either wet (gel), dry, or non-contact electrodes as described herein.

[0206] Volatile Organic Compound (VOC) Sensor Module In some variations, the system 100 may further include one or more VOC sensor modules for detecting volatile organic compounds (VOCs) or semi-VOCs in the subject's breath. The potential for breath analysis is enormous, with applications in many fields, including, but not limited to, disease diagnosis and monitoring. Some VOCs are linked to biological processes within the human body. For example, dimethyl sulfide is present in breath as a result of hepatic halitosis, and acetone is excreted via the lungs during diabetic ketoacidosis. Typically, VOC or semi-VOC excretions can be measured using plasmon surface resonance, mass spectrometry, enzyme-based, semiconductor-based, or imprinted polymer-based detectors.

[0207] Voice Intonation (VTI) Sensor Module In some variations, system 100 may further include one or more voice intonation (VTI) sensor modules. VTI analysis can indicate a range of mental and physical conditions that cause a subject to stutter, stretch sounds, or speak nasally. These may cause a subject's voice to creak or jitter briefly, even if the shortness of the slurred speech is undetectable by the human ear. Furthermore, changes in voice tone may indicate upper or lower respiratory tract or even cardiovascular conditions. Developers have discovered that VTI analysis can be used for early diagnosis of several respiratory illnesses from Covid-19 infection (see Examples 7 and 8).

[0208] Capacitive Sensor Module In some variations, the system 100 may further include one or more capacitive / non-contact sensor modules. Such sensor modules may include non-contact electrodes. These electrodes were developed because, without an impedance-adapting material, skin-electrode contact can become unstable over time. This difficulty was addressed by avoiding physical contact with the scalp through a non-conductive material (i.e., a small dielectric between the skin and the electrode itself). However, there is an abnormal increase in electrode impedance (>200 Mohm), which is quantifiable and stabilizes over time.

[0209] Certain types of dry electrodes are known as capacitive or insulated electrodes. These electrodes act as simple capacitors in series with the skin, so they do not require ohmic contact with the body and the signal is therefore capacitively coupled. The received signal can be connected to an operational amplifier and then to standard instrumentation.

[0210] The use of dielectric materials that make good contact with the skin results in a fairly large coupling capacitance, from 300 pF to a few nanofarads, and as a result, a system with low noise and adequate frequency response is easily achievable using standard high-impedance FET (field-effect transistor) amplifiers.

[0211] While wet and dry electrodes require physical contact with the skin to function, capacitive electrodes can be used without contact through an insulating layer such as hair, clothing, or air. Although these non-contact electrodes are generally described as simple capacitive electrodes, in reality they also have a small resistive element, since insulators also have a significant resistance.

[0212] Capacitive sensor modules can be used to measure cardiac signals, such as heart rate, and monitor respiratory rate of a subject via direct skin contact or through one- and two-layer clothing without dielectric gel and a ground electrode. High-impedance potential sensors can also be used to measure respiratory and cardiac signals.

[0213] Capacitive Plate Sensor Module In some variations, the system 100 may further include one or more capacitance plate sensor modules. Surprisingly, the developers discovered that changes in the body's dielectric properties associated with differences in hydration, electrolytes, and sweat levels can also be used to probe the body's resistive properties. In this variation, the sensing device may include two parallel capacitance plates positionable on either side of the body or body part to be probed. A specific time-varying potential is applied to the plates, and the instantaneous current required to maintain the specific potential is measured and used as input to a machine learning system that correlates physiological conditions to the data. As the dielectric properties of the body or body part change with resistance, the change is reflected in the current required to maintain the potential profile. In some variations, target physical conditions may be screened using such capacitance plates by allowing a subject to be probed while standing on the capacitance plates.

[0214] In some embodiments, a system for screening a target condition in a subject may be provided, the system comprising a device having a first portion incorporating a first sensor, the first sensor configured to capture physiological data from the subject, the first sensor including two capacitive plates positionable on either side of the subject's body or body portion. In some embodiments, the first portion is configured to be supported by a support surface such as a wall, floor, or ceiling during use. In some embodiments, the device further comprises a second portion, a top unit, extending perpendicular to the two capacitive plates. In some embodiments, the top unit is configured to face the subject's head during use. In some embodiments, the surface of at least one of the capacitive plates is mirrored. In some embodiments, there is further provided a computing system communicatively coupled to the first sensor and the second sensor and configured to perform a method, the method including obtaining physiological data captured from one or both of the first sensor and the second sensor, and providing the obtained physiological data to an MLA, the MLA being configured to determine a likelihood that the subject has a target state based on the obtained physiological data.

[0215] Machine Vision Sensor Module In some variations, system 100 may further include one or more machine vision sensor modules equipped with one or more optical sensors, such as cameras, for capturing the movement of a subject, or portions of a subject, while the subject is standing or moving (e.g., walking, running, playing sports, balancing, etc.). In this manner, physiological conditions affecting kinematic behavior, such as balance and gait patterns, trembling, swaying, or guarding, can be detected and correlated with other data obtained from other sensors in the device, such as center of gravity positioning. Machine vision amplifies skin movement to enable accurate measurement of physiological parameters such as blood pressure, heart rate, and respiratory rate. For example, heart rate / respiratory rate, heart rate / respiratory rate variability, and heart rate / respiratory rate duration can be estimated from measurements of subtle head movements caused by blood pumping to the head, hemoglobin information from observed skin color, and the periodicity of light reflected from arteries and skin near the face. Aspects of pulmonary health can be assessed from movement patterns of the chest, nostrils, and ribs.

[0216] A wide range of motion analysis systems enable motion capture in a variety of settings, which can be broadly categorized into direct (body-fixed devices, e.g., accelerometers) and indirect (vision-based, e.g., video or photoelectric) techniques. Direct methods allow kinematic information to be acquired in diverse environments. For example, inertial sensors have been used as tools to provide insight into the execution of various movements (walking, discus throwing, equestrianism, and swimming). Sensor drift, which affects the accuracy of inertial sensor data, can be reduced during processing but has not yet been fully resolved, and capture periods are limited. In addition, it has been recognized that motion analysis systems for biomechanical applications should meet the following criteria: they should be able to collect accurate kinematic information, ideally in a timely manner, without burdening the performer or affecting their natural movement. As such, indirect techniques can be distinguished as more appropriate in many settings compared to direct techniques, as data is collected remotely from participants with minimal interference to their movements. Indirect methods were previously the only viable approach to biomechanical analysis performed during sports competitions. Over the past few decades, indirect vision-based methods available to biomechanical researchers have dramatically advanced toward more accurate and automated systems. However, tools that fully satisfy the aforementioned key attributes of motion analysis systems have yet to be developed. Therefore, these analyses can be used in coaching and physical therapy for dance, running, tennis, golf, archery, shooting biomechanics, and other sports and physical activities. Other applications include ergonomic training for occupations that expose people to risk of repetitive stress injuries and other physical stressors related to movement and posture. This data can also be used in the design of furniture, self-training, tools, and equipment.

[0217] The machine vision module may include one or more digital camera sensors for imaging, for example, one or more of the subject's pupil dilation, scleral erythema, skin color changes, facial flushing, and / or erratic movements. Other optical sensors operating with coherent light may be used or may use time-of-flight computation. In some variations, the machine vision module includes a 3D camera, such as Orrbec's Astra Embedded S.

[0218] Thermal Sensor Module In some variations, the system 100 may further include one or more thermal sensor modules, including an infrared sensor, a thermometer, or the like. The thermal sensor module may be integrated with the sensing device or may be separate from it. The thermal sensor may be used to perform temperature measurements of one or more of the subject's lacrimal lake and / or external tear duct. In some variations, the thermal sensor module may include, but is not limited to, an infrared thermometer, a 3V, single sensor (not an array), gradient-compensated, medical-grade thermopile with ±0.2 to ±0.3 degrees Kelvin / Celsius, and a 5-degree field of view (FOV).

[0219] Strain Gauge Sensor Module In some variations, the system may include strain gauge sensors that may be used to measure the subject's weight. In other variations, these sensors may be used to obtain a vibratory cardiogram or ballistocardiogram. These sensors may be, without limitation, resistive or piezoelectric strain gauges.

[0220] Sensor module combination Any combination of the sensor modules described above may be used in variations of the present system 100. The combination of sensor modules may be housed in a single device or multiple devices. The relative positioning of the combination of sensor modules is selected to ensure that data is captured from the subject along the most appropriate plane. In some variations, two sensor modules are positioned orthogonally to each other to capture data from the subject along different planes. For example, the combination of sensor modules may include a capacitive plate sensor module positioned substantially horizontally and configured for the subject to stand on, and a vibroacoustic sensor module positioned substantially horizontally and configured to capture vibroacoustic signals from the subject.

[0221] Sensor Data In typical applications, the individually implemented modular systems described above require the attachment of "markers" or "beacons" to the subject to enable accurate signal chain tracking and surface motion amplification. By using sensor fusion, the current technology provides a method for tracking limb and body movements without requiring the attachment of separate displacement sensors or beacons to the subject.

[0222] In some variations, the sensor modules used with system 100 may each capture data as a concatenated raw amplitude train or as a combined short-time Fourier transform spectrum. In some variations, data from the sensor modules is captured from subjects in less than 15 seconds per subject, preferably less than 10 seconds per subject.

[0223] In some embodiments, the collected or monitored data may include one or more of optical, electromagnetic, and vibroacoustic monitoring parameters, which may be referred to as biometric data or biofield data.

[0224] The biometric data may be collected by any suitable device or system, such as one or more of the sensors described herein.

[0225] In some embodiments, biometric data is collected or monitored by one or more sensors, such as, but not limited to, a vibro-acoustic sensor, an electric potential sensor, a volatile organic compound sensor, and a broadband terahertz sensor (used to generate and detect electromagnetic waves at terahertz frequencies), a microphone, a volatile organic compound sensor, an altitude sensor, an ambient temperature sensor, a barometric pressure sensor, and an air quality sensor.

[0226] In some embodiments, the biometric data may include a single data parameter that is collected and / or monitored. In some other embodiments, the biometric data includes two or more data parameters that are collected and / or monitored.

[0227] In some embodiments, biometric data is collected and / or monitored during one or both of the baseline and baseline update phases, which can correct for physiological drift.

[0228] During the baseline phase, biometric data may be collected and / or monitored over 1 to 5 days, 1 to 4 days, 1 to 3 days, 1 to 2 days, 2 to 5 days, 3 to 5 days, 4 to 5 days, 1 to 3 days, or 2 to 3 days. Data collection may be continuous or in data segments.

[0229] During the update phase, biometric data may be collected and / or monitored between 1 and 25 seconds, 5 and 25 seconds, 10 and 25 seconds, 15 and 25 seconds, 20 and 25 seconds, 1 and 20 seconds, 1 and 15 seconds, 1 and 10 seconds, 5 and 10 seconds, and 5 and 15 seconds.

[0230] In some embodiments, the biometric data is acquired in data segments of about 15 seconds to about 20 seconds in length, or about 10 seconds to about 25 seconds in length, or any other data segment length that meets data quality and data quantity requirements.

[0231] In some embodiments, data is collected between about 2 and about 4 days for continuous health characterization and baseline.

[0232] In some embodiments, updates using baselined data require a shorter confirmation data read or top-up of about 5 to about 10 seconds.

[0233] In some embodiments, the method includes obtaining biometric data of a subject at a first time point and storing the data in a database (a "pre-screening step"). The method further includes obtaining additional biometric data at a second time point and using the stored data for monitoring or diagnostic purposes.

[0234] The pre-screening process may be carried out over a period of about 1 to 5 days.

[0235] In some embodiments, the baseline data is ephemeral (can be deleted, overwritten, or lose validity).

[0236] In some embodiments, the method may include collecting and / or monitoring data at a sampling rate of about 0.01 Hz to about 20 THz, greater than about 10 THz, about 10 THz to about 100 THz, or about 0.01 Hz to about 100 THz. In some embodiments, this may result in improved data quality and quality implications. These sampling rates may be considered "high resolution" compared to conventional data sampling. In the terahertz range, data may be captured passively or actively.

[0237] In some embodiments, data from the sensors is captured from the subjects in less than 15 seconds per subject, preferably less than 10 seconds per subject.

[0238] In some embodiments, high-resolution biometric / biofield / biosignature data can be processed using cross-frequency coupling and cross-phase signal coupling methods to abstract the biosignature data.

[0239] In relation to current technology, cross-frequency coupling (CFC) methods include analysis of mean vector length or modulation index, phase synchrony, envelope-to-signal correlation, amplitude spectrum analysis, analysis of coherence between amplitude and signal, analysis of coherence between power and signal time course, and eigendecomposition of multi-channel covariance matrices. These methods analyze how frequency changes in one signal affect the frequency of a second signal.

[0240] Cross-frequency coupling (CFC) also includes the analysis of phase-amplitude coupling (PAC), a form of cross-frequency coupling in which the amplitude of a first frequency signal is modulated by the phase of a second frequency signal. These also include the analysis of phase-phase coupling (PPC), in which the phase of one frequency signal affects the phase of a second frequency signal, and amplitude-amplitude coupling (AAC), in which the amplitude of the first signal affects the amplitude of the second frequency signal.

[0241] In yet another embodiment, the processing includes performing steps of sampling the data stream from one or more sensors, sampling signal acceptance / rejection, sampling signal enhancement, sampling signal normalization, feature extraction, cross-frequency coupling and signal aggregation and abstraction, and comparison of the biometric sample with all stored samples in a database.

[0242] 1.d. Electronic Systems In some variations, the sensing device may further include an electronics system. In the variation illustrated in FIG. 3B, the sensing device 300 includes an electronics system 340 including various electronics components for supporting the operation of the sensing device 300. In the variation illustrated in FIG. 4B, the sensing device 400 includes an electronics system 440 including various electronics components for supporting the operation of the sensing device 400. For example, at least a portion of the electronics system 340, 440 may comprise a circuit board modularly arranged within the housing 310, 410 of the vibrometer sensing device 300, 400. The electronics system 340, 440 may be configured to perform signal conditioning, data analysis, power management, communications, and / or other suitable functions of the device. The electronics system 340, 440 may be in communication with other components of the sensing device, such as the vibratory and acoustic sensor module 320, 420, the context sensor module 330, 430, the power source 342, 442 (e.g., a battery), and / or the display 330. Thus, the electronics system 340, 440 may, in some variations, function as a microcontroller unit module for the sensing device 300, 400.

[0243] 23 , the sensing device electronics system 2300 may include at least one processor 2310, at least one memory device 2320, suitable signal processing circuitry 2330, at least one communications module 2340, and / or at least one power management module 2360. One or more of these components or modules may be arranged on one or more electronic circuit boards (e.g., PCBs), which may then be mounted within the housing of the sensing device. In some variations, the electronics system may also include a microphone and / or speaker to enable additional functionality such as voice or data recording (e.g., enabling reading of medical notes for an electronic health record).

[0244] The processor 2310 (e.g., a CPU) and / or the memory device 2320 (which may include one or more computer-readable storage media) may cooperate to provide a controller for operating the system. For example, the processor 2310 may be configured to set and / or adjust sampling frequencies for any of the various sensors in the vibroacoustic sensor modules 320, 420 and / or context sensor modules 330, 430. As another example, the processor 2310 may receive sensor data (e.g., before and / or after a sensor signal condition), and the sensor data may be stored in one or more memory devices 2320. In some variations, some or all of the data stored in the memory device 2320 may be encrypted using a suitable encryption protocol (e.g., for HIPAA-compliant security). In some variations, the processor 2310 and the memory device 2320 may be implemented on a single chip, while in other variations, they may be implemented on separate chips.

[0245] The communications module 2340 may be configured to communicate sensor data, analytical data, and / or other information to an external computing device. Additionally or alternatively, the communications module 2340 may communicate with external sources for microcontroller programming and software updates. The external computing device may be, for example, a mobile computing device (e.g., a mobile phone, a tablet, a smartwatch), a laptop, a desktop, a medical device, or other suitable computing device. The external computing device may run an application for presenting the sensor data (and / or its analytical results) to a user through a user interface.

[0246] Additionally or alternatively, the communications module 2340 may be configured to communicate data to one or more network devices, such as a hub, a server, a cloud network, etc., paired with the system 100. In some variations, the communications module 2340 may be configured to communicate information in an encrypted manner. In some variations, the communications module 2340 may be separate from the processor as a separate device, while in variations, at least a portion of the communications module may be integrated with the processor 2340 (e.g., the processor may include encryption hardware such as an Advanced Encryption Standard (AES) hardware accelerator (e.g., 128 / 256-bit key) or HASH (e.g., SHA-256)). Additional aspects of the communications scheme are described in further detail below with respect to the signal processing system.

[0247] The communications module 2340 may communicate via a wired connection (e.g., including a physical connection such as a cable having a suitable connection interface, such as USB, mini-USB, etc.) and / or a wireless network (e.g., through NFC, Bluetooth, WiFi, RFID, or any type of digital network not connected by a cable). For example, the devices may communicate with each other directly in a pair-wise connection (a 1:1 relationship), or in a hub-spoke or broadcast connection (a "1:many" or 1:m relationship). As another example, the devices may communicate with each other through a mesh networking connection (e.g., a "many-to-many," or m:m relationship), such as Bluetooth mesh networking. The wireless communications may use any of a number of communication standards, protocols, and technologies, including, but not limited to, Global System for Mobile Communications (GSM), Enhanced Data GSM Environment (EDGE), High Speed ​​Downstream Packet Access (HSDPA), High Speed ​​Upstream Packet Access (HSUPA), Evolution, Data-Only (EV-DO), HSPA, HSPA+, Dual-Cell HSPA (DC-HSPA), Long Term Evolution (LTE), Near Field Communication (NFC), Wideband Code Division Multiple Access (W-CDMA), Code Division Multiple Access (CDMA), Time Division Multiple Access (TDMA), Bluetooth®, Wireless Fidelity (WiFi®) (e.g., IEEE® 802.11a, IEEE® 802.11b, IEEE® 802.11g, IEEE® 802.11n, and the like), or any other suitable communication protocol. Some wireless network deployments may combine networks from multiple cellular networks (e.g., 3G, 4G, 5G) and / or may use a mix of cellular, WiFi, and satellite communications.

[0248] In some variations, the communications module 2340 (e.g., used for input and function operations and / or haptic feedback) may include multiple data communication streams or channels to help ensure wideband data transfer (e.g., over 20 kHz with a minimum-delay codec). Such multiple data communication streams are an improvement over typical wireless data transmission codecs. For example, most wireless data communication codecs (e.g., G.711) use bandpass filters to encode only the optimal range of human speech, from 300 Hz to 3,400 Hz (commonly referred to as narrowband codecs). As another example, some wireless data transmission codecs (e.g., G.722) encode the range from 300 Hz to 7,000 Hz (commonly referred to as wideband codecs). However, most of the energy is concentrated below 1,000 Hz, with virtually no audible sound above 5,000 Hz, while there is a measurable amount of energy above the 3,400 Hz cutoff frequency of most codecs. The data throughput requirements for both G.711 and G.722 are the same because the modulation used in G.722 is a modified version of PCM called adaptive differential pulse code modulation (ADPCM). This type of complexity adds latency when the codec and processing power are constant. As such, G.711 introduces latency of just under a millisecond, while G.722 can introduce delays of tens of milliseconds, which is unacceptably long for vibroacoustics.

[0249] The power management module 2360 may be configured to manage power from one or more power sources and distribute the power appropriately to the processor, communications module, sensors, and / or any other electrical components. For example, the power source may include one or more batteries (e.g., lithium-ion batteries) disposed within the housing of the sensing device as described above. In some variations, the power source may be rechargeable, such as through wireless charging methods (e.g., inductive charging, RF coupling, etc.) or by harvesting kinetic and / or thermal energy such as that generated through movement (e.g., including harvesting thermal energy from the body as the user walks while wearing the garment, or by using energy collection, amplification, and storage cells that collect and convert light into an electrical signal and / or cells that directly convert temperature or a temperature gradient into electricity). In some variations, the power management module 2360 may be connected to the power source through a suitable charge controller.

[0250] In some variations, the power management module 2300 may comprise electronic components for converting power to a predetermined voltage output suitable for other components in the sensing device (e.g., a processor and / or memory device, a signal processing system, etc.). For example, the power management module 2300 may comprise a buck-boost converter for outputting 3.3V and 5V, and an on-board universal serial bus (e.g., USB-C) that can be used to charge the module and / or power supply with an external charger (e.g., a mobile charger, a power outlet, etc.).

[0251] 1.e. Signal Processing System Various analog and digital processes may condition the sensor data to extract useful signals from noise and communicate appropriate data to one or more external host devices (e.g., a computing device such as a mobile device, one or more storage devices, medical equipment, etc.). At least a portion of the signal processing chain may appear on a circuit board within one or more sensor modules (e.g., a flexible or rigid circuit board within a vibroacoustic sensor module, a context sensor module, and / or any other sensor module). Additionally or alternatively, at least a portion of the signal processing system may appear external to the sensor module (e.g., the electronics system 340, 440 or the microcontroller unit module).

[0252] In some variations, a signal processing chain for handling data such as vibroacoustic sensor data, ECG data, background condition data, thermal data, optical data, etc., may be configured to provide a low-noise (high signal-to-noise ratio (SNR)) output signal, provide sufficient amplification to allow proper digitization of the analog signal, and function in a manner that maintains a sufficiently high overall signal fidelity. The signal processing chain may also be configured to (i) overcome signal attenuation and signal strength loss as the signal propagates over one or more media, and / or (ii) move the digitized data through the various components of the sensing device sufficiently quickly to avoid significant signal and / or data loss. In some variations, the signal processing chain may include a programmable gain stage that adjusts the gain in real time during operation of the sensing device to optimize the signal range for the analog-to-digital converter. While the frequency and bandwidth requirements of the signal processing chain may vary depending on the particular application, in some variations, the signal processing chain may have sufficient bandwidth to sample frequencies up to about 160 kHz or up to about 320 kHz and have a low-frequency response of about 0.1 Hz or less.

[0253] High-precision signal control can be important for minimizing signal and / or data loss in biofield and other vibroacoustic active / passive sensing. However, the difficulty of obtaining model parameters is one of the major obstacles to obtaining high-precision tracking control of biofield signals using model-dependent methods. Vibroacoustic systems with uncertain parameters can prevent signal and / or data loss by having highly accurate system output information. In some variations, an adaptive output feedback control scheme may be implemented by an in-line servo system with uncertain parameters and unmeasurable states instantiated in a controller and a parameter adaptation algorithm, thereby ensuring that biofield signal tracking errors are uniformly bounded. This method may be combined with conventional proportional-integral-derivative (PID) control methods with optimal parameters (e.g., obtained using a genetic algorithm), sliding mode control based on exponential reaching laws, and / or adaptive control methods and adaptive backstepping sliding mode control, thereby achieving high tracking accuracy. Vibroacoustic systems may also have better interference prevention capabilities with respect to signal load changes.

[0254] Vibroacoustic signal control, which may be referred to as active vibroacoustic control, may be achieved in some variations using multiple servo motors, actuators, and sensors with fully coupled feedforward or feedback controllers. For example, in some variations, feedback may be achieved using multiple small cross-axis inertial sensors (e.g., accelerometers) with either co-located force or piezoelectric actuators placed below each sensor. Co-located actuator / sensor pairs and distributed (local) feedback may be optimized over a bandwidth of interest to ensure stability of multiple local feedback loops. For example, a control system may include an array of actuator / sensor pairs (e.g., an n×n array of such actuator / sensor pairs, such as 4×4 or greater), which may include n 2 They can be connected together in local feedback control loops. Using force actuators, significant frequency-averaged reductions up to 1 kHz in both kinetic energy (e.g., 20-100 dB) and transmitted sound power (e.g., 10-60 dB) can be obtained with appropriate feedback gains in each loop.

[0255] 24 illustrates an exemplary signal processing chain for manipulating vibro-acoustic sensor data from one or more sensors 2402 in a vibro-acoustic sensor module. As shown in FIG. 24, a first set 2410 of analog signal processing steps may be performed by analog circuitry on one or more printed circuit boards (PCBs), a second set 2420 of analog signal processing steps may be performed by analog circuitry within a microcontroller unit module, and a third set 2430 of digital signal processing steps may be performed by digital circuitry within a microcontroller unit module. However, it should be understood that these steps may be performed by analog and / or digital components located in any suitable location on or within a sensing device (e.g., an electronic system).

[0256] The analog portions of the signal chains 2410, 2420 may be optimized for low noise and high SNR signal acquisition of the vibroacoustic sensor signals. In addition to low-noise components, the PCB itself may also be designed and optimized for low-noise operation. For example, the PCB may include multiple layers (e.g., four layers) including an entire layer dedicated as a ground plane to prevent ground loops, provide a low-resistance ground, and act as a shield between signal lines on the remaining layers. In some variations, the signal chain may include at least second- to fourth-order low-pass filtering to help prevent aliasing, as described further below.

[0257] The raw signal from sensor 2402, in some variations, may typically be in the range of 100 μV to 1 mV, but may be up to 10 mV for high SNR sensors in the sensing device. First stage amplifier 2412 is a specific low-noise input-optimized amplifier and may have a gain between about 50 and about 200, depending on the specific sensor. First stage 2412 may also include a first-order low-pass filter with a cutoff frequency at about 15 kHz to 20 kHz. The signal may then be fed to a first- or second-order active filter stage 2413 with a cutoff frequency of about 15 kHz to 20 kHz. In some variations, this active filter stage may be a second-order filter implemented using a Sallen-Key topology. The signal from the first stage may be fed to a second stage amplifier 2414, which in some variations has a gain between about 1 and about 10-100 with another low pass filter having a cutoff frequency from about 0.01 Hz to about 120 kHz (e.g., from about 0.01 Hz to at least about 50 kHz, from about 0.01 Hz to at least about 60 kHz, from about 0.01 Hz to at least about 70 kHz, from about 0.01 Hz to at least about 80 kHz, from about 0.01 Hz to at least about 90 kHz, from about 0.01 Hz to at least about 100 kHz, from about 0.01 Hz to at least about 110 kHz, from about 0.01 Hz to at least about 120 kHz, from about 0.01 Hz to at least about 130 kHz, from about 0.01 Hz to at least about 140 kHz, from about 0.01 Hz to at least about 150 kHz, from about 0.01 Hz to at least about 160 kHz, from about 0.01 Hz to greater than 160 kHz). In some variations, the signal may additionally be fed through another low frequency, high pass filter or AC coupling (1615) with a cutoff frequency of about 0.01 Hz.

[0258] The filters described above (at 2412, 2413, 2414, and 2415) may be combined to form a second- to fourth-order filter with an overall cutoff frequency between about 10 kHz and about 20 kHz, which acts as an anti-aliasing filter for a downstream sigma-delta ADC in the signal processing chain. At 20 kHz, the second-order filter has an attenuation of about -15 dB, which can be easily corrected in the digital domain. In some variations, the internal ADC sampling rate is about 3 MHz, so sufficient attenuation is required at the Nyquist frequency of 1.5 MHz. The second-order filter achieves greater than -100 dB at this frequency.

[0259] The signal from the second stage amplifier and low pass filter (2414) may be boosted by an offset voltage (2416) of between about 0.5V and about 5V in some variations. In some variations, the offset voltage may be between 0.5V and about 2V, or between about 1V and about 2V. The offset may depend on the actual configured ADC range (e.g., for an ADC range of 0V to 1V, the offset may be about 0.5V). This offset may accommodate, for example, an ADC that supports only positive voltages for conversion. This stage may also incorporate an AC coupling capacitor with a cutoff frequency of about 0.01Hz to 0.1Hz (e.g., about 0.05Hz), which may facilitate low frequency response and / or DC offset blocking.

[0260] Figure 25A illustrates a transfer function for the exemplary signal processing circuit of Figure 24. Figure 25B illustrates a transfer function of an analog circuit that performs the first set of analog signal processing steps 2410 using the parameters described above. Figure 25C illustrates a transfer function of another analog circuit.

[0261] Following the offset 2416, the signal may enter a second set of analog signal processing steps 2420. As shown in Figure 24, the first stage in the MCU may comprise a programmable gain amplifier (PGA), which represents the third gain stage in the signal processing chain and provides flexibility for on-the-fly adjustment of gain to optimize for the ADC range. The general equation for amplification is shown in Equation 5 below: V out =V Ref +(V In -V Ref )×Gain Equation 5

[0262] Therefore, to amplify the dynamic signal, the offset previously added in circuit 2416 is V Ref As shown in the circuit diagram of FIG. 26, a digital-to-analog converter (DAC) 2620 controls the V Ref pin, which allows V Ref On-the-fly adjustment of the value of DAC 2620 can be achieved. However, DAC 2620 may add significant noise to the amplifier beyond its acceptable range. To address this issue, a capacitor 2630 can be added between the DAC signal and ground. Because the DAC signal is generally static, this addition poses no concerns about dynamic behavior except for brief periods when the voltage changes.

[0263] In addition to allowing on-the-fly adjustment of gain, the PGA may advantageously mitigate offset voltage variations caused by upstream circuitry. These variations, for example, are due to tolerances in electronic components and can add up to ±20% dispersion of the offset signal (within 1616). The PGA may also be used to compensate for this offset, moving it to half the ADC reference voltage (e.g., a voltage of 2.048V). Attempting to move the offset to exactly half this voltage, i.e., 1.024V, would result in the appropriate V in Equation 6: Ref This will be determined.

[0264]

number

[0265] Note that Equation 6 is only valid for gains of 2 or greater. Therefore, in this example, the minimum gain to use is 2 rather than 1. This new V Ref Prior to the calculation of the offset, the true offset present on each PCB must be known, which can be determined through calculation. For example, determining the true offset can be achieved by determining the offset value from an external circuit via recording the signal for a short period of time and forming an average value. This can be done either directly on the MCU or via a connected target device, and the correction can be sent back to the MCU.

[0266] Next in the signal chain is a track-and-hold component (2424) that holds the voltage constant after the trigger signal. Ideally, the voltage is constant for the duration required for the ADC to sample the complete value, which is important for accurate results. The trigger fires at the end of the previous ADC sample period, and the ADC immediately proceeds with the next sample.

[0267] The track-and-hold voltage may be held constant for a suitable predetermined time, such as approximately 5-50 μs, before being released again. This release allows the track-and-hold to follow the current voltage for another period of time (e.g., approximately 2 μs) before the next trigger. When sampling multiple sensor signals, the track-and-hold component helps ensure proper time synchronization between the sensor signals, with synchronization accuracy within the nanosecond range.

[0268] The final component in the analog signal processing chain is a sigma-delta ADC (2426). In some variations, while operating internally at approximately 3 MHz, the ADC uses oversampling to achieve a 16-bit, approximately 48-96 kHz signal within a range of approximately 0 V to approximately 2.048 V. In some variations, accuracy can be further increased by referencing the ADC ground to the circuit board ground as described above.

[0269] The analog signals entering the ADC may then be sampled into the digital domain. After the analog signals are sampled into the digital domain, they may be moved to a buffer location in memory via high-speed DMA transfers (2432), etc. The signals may be further buffered in a ring buffer before being assembled into packages (2434) and transmitted to one or more external host devices (2436). These packages may include a header and a payload. The header may have a package / frame start ID and other supplemental data, which may help keep the data in the correct order after transmission and aid in detecting lost data points. The payload is sensor fusion data (e.g., vibro-acoustic data, contextual data, etc.). In some variations, the multimode data points are transmitted in each package via USB or wireless transmission (e.g., BLE, Bluetooth classic, Wi-Fi, etc.) to one or more external host devices 2404 (e.g., the cloud, mobile computing devices, etc.) for analysis, etc. However, it should be understood that in other variations, other package sizes and / or other communication modalities may additionally or alternatively be incorporated. Additionally or alternatively, analysis of the vibro-acoustic data may be analyzed locally on the sensing device (e.g., using appropriate machine learning models) before the data and / or analysis results may likewise be communicated to one or more external host devices.

[0270] 1.f. Encoding Module Software-defined biotelephony is a software-intensive approach that balances the needs of biosensing, communications, and digital health computing with the needs of the various networks its users may operate over. The tradeoffs involve the degree of flexibility created a priori (e.g., through new protocols) versus the limited degree of autonomy allowed by existing radios and receivers.

[0271] Encoding audible / inaudible, visible / invisible, felt / unfelt, contact / non-contact, proximity / remote capture data without loss of information so that the encoded signal contains and transmits the same breadth, depth, quality and robustness of information whether wireless or wired is not possible with existing CODECS. Some variations of the current technological advances of sensor module combination and sensor data fusion also require improvements in methods and systems for data collection and data communication. The proposed technology will support a wide range of frequencies from 0.01 Hz to 10 11 It is intended to sense, process, and transmit vibroacoustic, electrical, magnetic, and electromagnetic signals with characteristic bandwidths exceeding 100 Hz (Figure 1E), and therefore, appropriate hardware and associated software applications, usually referred to as CODECS (compressors / decompressors), are required to enable the operation of the current technology.

[0272] There are two general categories of factors that affect the fused, encoded wideband data stream output by the encoder of a data stream (e.g., audio) codec: details about the source (e.g., audio) format, and the codec and its configuration during the encoding process. For each factor that affects the encoded data stream, there is a simple rule that is almost always true: the fidelity of a digital data stream (e.g., audio) is determined by the granularity and precision of the samples taken to convert it into a data stream; the more data used to represent the digital version of the audio, the more accurately the sampled sound will match the source material.

[0273] Variations on the current technology address both problems by providing highly efficient software-based, ultra-high bandwidth encoding, transmission, and decoding devices, methods, and systems. Variations on the current technology provide transceiver systems and methods configured to receive and decode multi-mode signals from multi-sensor data streams. The proposed signal transceiver system utilizes "audio beacon" data streams. The disclosed signal receiver provides accurate signal decoding of low-level signals, even in the presence of significant noise, and the software technology stack consumes very low power. Also provided are systems including the receivers, as well as methods of using the same.

[0274] Effect of source audio format on encoded audio output There are several features that can be used to balance reproduction quality and file size. Because standard encoded wireless data streams (e.g., audio, i.e., audible vibrations) inherently use fewer bits to represent each sample, the source audio format may actually have less of an impact on the encoded audio size than might be expected. However, many factors still affect the quality and size of the encoded audio. Table 1 shows the major factors of the source audio file formats considered and optimized in this technology and their acceptable impact on the encoded audio.

[0275] [Table 1]

[0276] First, channel count only affects the signal's directionality or spatial positioning. Depending on the content, file size may be multiplied by the number of encoded channels, or some methods may utilize inter-channel redundancy to reduce file size without significant signal degradation. Second, signal noise or hiss tends to degrade audio quality both directly (by masking foreground audio details) and indirectly (by making the data stream waveform more complex and therefore more difficult to reduce in size while maintaining accuracy). Thus, hiss, static, or background noise increases the complexity of the data stream, generally reducing the amount of compression possible. Third, the higher the sample rate, the more samples available per second, and the higher the fidelity of the resulting encoded data stream. However, this comes at the cost of the size of the encoded file or the bit rate of the encoded stream.

[0277] Finally, the sample size affects the detail available in each sample, but this depends on the codec, which typically has an internal sample format that may or may not be the same as the original sample size. However, the more detail in the source, the larger the encoded file will be, never smaller.

[0278] The effects detailed above can be modified by decisions made when encoding the data stream: for example, if the encoder is configured to reduce the sample rate, the effect of the sample rate on the output file will be correspondingly smaller.

[0279] Current technology codecs, in some variations, employ software-defined hardware and firmware algorithms to take fusion source structural and physiological health data streams and compress them to occupy substantially less space in memory or network bandwidth without sacrificing information or data quality. In some variations of encoder configurations, the encoder may be tuned using parameters that select particular algorithms, tune those algorithms, and specify the number of passes to apply during encoding.

[0280] Current technology, Infrasound-to-Terahertz over wireless / IP, differs from traditional audio CODECs in that it has evolved and optimized several aspects for low-frequency, low-amplitude biometric data transmission: smart scalable switching (easy to add more ports and only what is needed); breaking down distance barriers; improved input to output ratios; separate fused and multiplexed data stream standards that extend beyond the local facility; convergence of low-frequency, low-amplitude biometric data and wireless communications; and new options for local, edge, and cloud processing.

[0281] Hardwired, circuit-based switching is essentially a point-to-point technology. A matrix switcher for wideband data streams is simultaneously an intelligent "destination" and "source." All transmitter and receiver combinations are resolved within the matrix switch, allowing any source to be matched, assigned, and optimally used with any destination depending on the number of signal transmission and signal reception ports available on the data stream matrix switch. For example, an 8x8 matrix switch allows eight sources to be used with eight destinations.

[0282] Current technology devices and systems can perform local, edge, and back-end handshaking operations. For example, instead of only making any input available on any output, any input can be shown on any—even multiple—outputs. Thus, biometric data transmission from a wideband biometric device and data visualization can be source routed from a device transformer box to a data visualization matrix switcher, which can then be wired to multiple devices that can simultaneously show the biometric data stream in real time.

[0283] With wireless / IP (and packet-based switching), the number of sources connected to a wireless / IP switch is unlimited. When physical ports are exhausted, multiple wireless / IP switches can be connected to expand. The number of ports can be scaled to meet needs even more conveniently. It is possible to keep adding sources and destinations without a substantial overhaul of the core data stream matrix switcher becoming a major limiting factor.

[0284] The ratio of inputs to outputs can also be tweaked: it is possible to have many inputs but few outputs, or few inputs but many outputs, or many of both in widely differing amounts.

[0285] In some variations, current technology, Infrasound-to-Terahertz over wireless / IP, significantly increases flexibility by overcoming limitations on the number of sources and destinations, and even distance limitations.

[0286] In some variations, current technology Infrasound-to-Terahertz over wireless / IP devices use standards-based packetization for transmission over wireless / IP networks and compatibility with wireless / IP switches, while some use proprietary packetization schemes that work with wireless / IP networks and standard wireless / IP switches but not with other products on the market.

[0287] Generally, standards-based schemes offer the potential for interoperability between products from different vendors.

[0288] In some variations, the standards-based and proprietary packetization schemas of current technology do not alone determine interoperability, nor do they determine whether a product is more or less secure. It is the current variations of tightly coupled Infrasound-to-Terahertz over wireless / IP encoders and decoders that provide data safety and security. One reason for this tight coupling is to provide guaranteed specifications and performance. This also allows for a very controlled, out-of-the-box ease of setup and a user-friendly experience.

[0289] For some aspects of current technology Infrasound-to-Terahertz over wireless / IP products, encryption technologies exist that address multiple components of data stream system design.

[0290] Current technology devices provide encryption on the command and control signaling to the encoder and decoder devices. This provides security against hacking the operation of the box—including turning streaming on and off or switching the source being displayed. Another security aspect is the ability to encrypt the data stream itself. This ensures that if the data stream is intercepted, it cannot simply be decoded and viewed.

[0291] In other embodiments, current technology products offer support for third-party devices that use digital key exchange or encryption. Leveraging lessons learned from consumer examples, the overwhelming majority of AV customers are interested in the simplest case—High-Bandwidth Digital Content Protection (HDCP). The purpose of HDCP is to protect digital copyright-protected content as it travels between devices. For example, a cable or satellite receiver box or a media player with an HDMI output may play HD or 4K content that is protected content. Such content is locked and can be viewed by HDCP-enabled products after proper authentication. Similarly, current technology data streams have distributed ledger (e.g., blockchain) provenance, recording, security, and limitations on how protected content can be extended, augmented, modified, or viewed.

[0292] Piggybacking data transport protocols and other CODECs to transport Infrasound-to-Terahertz over wireless / IP Telephone touch-tone protocols are perhaps the most ubiquitous voice data transmission heard every day. Here, multi-frequency tones are used to dial numbers in the voice frequency band. Similarly, infrasound, ultrasound, and terahertz data streams can be overlaid on audible sound to achieve faithful broadband data stream communication. Structural, physiological, and gray and standard health data collected by current technology sensor fusion and data fusion platforms should be encoded onto a non-audible, near-ultrasound layer placed on top of normal audible sound, and / or onto a non-audible, near-infrasound layer placed just below normal audible sound, to cover broadband data transmission. The overlay of near-infrasound and / or near-ultrasound layers transforms any acoustic stethoscope, any smartphone, any microphone and speaker, IoT device, etc. into a data transmission device that can then be used for health data transfer, health insurance payment information transfer, user authentication, and smart city applications such as digital locks and mass transit turnstiles.

[0293] A universal, interoperable Infrasound-to-Terahertz over wireless / IP platform takes advantage of the lack of a common general-purpose protocol for connecting the Internet of Things, leveraging the immense power, performance, and intelligence of smartphones, smart speakers, IoT surveillance, and the proliferation of any / anywhere digital solutions around us to do just that.

[0294] Current technology weaves infrasound-terahertz data streams into common means of audible sound such as VOIP, streaming music, public announcements, etc. Infrasound and ultrasound are types of vibrations / sounds that are not detectable by the human ear but can be picked up by specialized equipment.

[0295] A smartphone or microphone can then be used to receive the generated audio pulses and decode the data. The speaker of a nearby smartphone can then transmit the cataloged data stream frequency tag to any receiver, such as one embedded in system 100, upon turning on the technology. Indeed, any specific action can be triggered, including a purchase or communication with an automated call center.

[0296] In some variations, patients entering urgent care can be identified, greeted, registered, and processed via target / beacon functionality using a combination of legacy stored data streams and real-time collected infrasound-terahertz data streams using their own smart device and on-site speaker. Infrasound-to-terahertz over wireless / IP solutions communicate even when smartphones are in airplane mode and data is turned off for use in high-security environments and transportation verticals. Traditional networks, such as wireless, can become overloaded when critical information—such as public safety issues—needs to be communicated. This intelligent system manages and optimizes crowding issues that can occur in large crowd locations, such as stadiums. Additionally, utilizing an “audio” smartphone / speaker beacon consumes less battery power than Bluetooth.

[0297] In some other variations, the WLAN (Wireless Local Area Network) unlicensed frequency bands used in three different regions - industrial (902-928 MHz), scientific (2400-2483.5 MHz), and medical (5725-5850 MHz) - devised in 1985 and administered by the Federal Communications Commission (FCC) provide the ability for Infrasound-to-Terahertz over wireless / IP solutions to utilize multiple longitudinal degrees of freedom across cooperative spectrum wide frequency networks.

[0298] 1.g. Artificial Intelligence Module Without the right algorithms to refine the data, the true value of high-resolution sensor data fusion remains hidden. Popular approaches, such as neural nets, model correlations rather than causality and do not support extrapolation from the data. In response, developers are developing new structural machine learning (SML) platforms—natural feedforward and feedback platforms—that can achieve faster and more accurate data exploration and utilization. Automatic representation synthesis tools build generalizable, evolving models to distill sensor data into human-interpretable forms, unlocking the power of fused data in intelligent, agile, networked, and autonomous sensing and utilization systems.

[0299] In this regard, in some variations, system 10 includes an artificial intelligence module configured to optimize analytical software, including a data-driven feedback loop, using machine learning and other forms of adaptation (e.g., Bayesian probabilistic adaptation) for analyzing vibroacoustic and / or other sensor data. Training such machine learning models to analyze data from sensing devices may begin with human-derived prior knowledge, or “soft knowledge” artifacts. These “soft knowledge” artifacts are advantageously generally much more expressive than off-the-shelf ML models such as neural nets or decision trees. Furthermore, in contrast to mainstream machine learning scenarios that have clearly distinguished between training and testing phases, analytical software for analyzing data from sensing devices may involve learning and optimizing the software inline. In other words, the concept here is to embed “inline learning” algorithms within artificial intelligence (AI) software systems, enabling the AI ​​systems to adaptively learn as they process new data. Such in-line (and real-time) adaptation typically results in software AI systems that perform better with respect to various functional and non-functional characteristics or metrics because, at a minimum, (i) the AI ​​system can correct for suboptimal biases introduced by human designers and (ii) can respond quickly to changes in characteristic operating conditions (mostly fluctuations in the data being processed).

[0300] Specifically, with regard to analyzing vibroacoustic data, the vibroacoustic biofield harvested from a patient can be saved as an audio (.wav) file. Custom cross-frequency coupling techniques can also be used in combination with averaging wavelets, such as the Daubechies and Haar wavelet approaches, to analyze infrasound data as static images within a set time frame. The Haar wavelet is the first and simplest orthogonal wavelet basis. Because the Daubechies wavelet averages more data points, it is smoother than the Haar wavelet and may be more suitable for some applications. Audio scenes typically have complex content, including background noise mixed with a rich foreground of audible and inaudible vibrations and their background context. In general, both background noise and foreground sounds can be used to characterize a "diagnostic scene" for use in characterizing the subject. Other data, such as contextual data, can be converted into a visual 2D representation and attached to the static infrasound image to create a new image. This new image can then be analyzed as a whole to improve algorithm performance.

[0301] However, foreground sounds typically occur in arbitrary order, making it difficult to uncover hidden sequential patterns. Therefore, the ability to recognize and “unmask” the surrounding diagnostic environment by separating and identifying audible and inaudible vibration signals against a background context has potential for many diagnostic applications. One approach to achieving this is to move from traditional classification techniques to modern deep neural networks (DNNs) and random convolutional neural networks (CNNs). However, despite their best performance, variations of these networks may be inadequate at modeling sequences in some applications. Therefore, in some variations, an AI system may incorporate a combined deep, symbolic, hybrid recurrent and convolutional neural network (R / CNN). Furthermore, in some variations, a separate DNN may generate and propose “crisp” (symbolic) programs, and feedback from the execution of such programs may be used to tune / train the DNNs and / or CNNs in a hybrid symbolic-subsymbolic approach.

[0302] In some variations, the sensitivity of the sensing platform may be enhanced by using biophysiologically accurate simulated patient entities for training purposes of machine learning algorithms. For example, such simulated entities may be uploaded and modified in a training environment using high-precision clinical data (e.g., heart rate, pulse rate, respiratory rate, heart rate variability, respiratory rate variability, pulse delay, core body temperature, upper and lower respiratory temperature gradients, etc.) collected from well-characterized clinical patients to create large, realistic training datasets.

[0303] In some variations, the machine learning module is configured to (i) design a Covid-19 biosignature during the training phase using variations of the sensing devices and systems described herein, and / or (ii) apply a Covid-19 biosignature using variations of the sensing devices and systems described herein.

[0304] A novel aspect of the method performed by the machine learning module involves posing the machine learning problem (here, the design of a COVID-19 biosignature) as a program synthesis task. To this end, a domain-specific language (DSL) was designed to express various designs of biosignatures as programs in that language. In some variations, input to the DSL includes the raw time series (detected frequency signals) as well as various types of features extracted from that series, such as FFT spectra, STFT spectrograms, MFCCs, vibration scale features, peak placements, etc. These correspond to specific data types in the custom DSL. The DSL is equipped with functions (instructions) that can process specific types of inputs and variables. DSL functions are based on domain-specific knowledge. For example, DSL functions that we currently use on a daily basis include convolution, peak detection, parameterizable low-pass and high-pass filters, arithmetic operations on time series, etc.

[0305] Importantly, these building blocks are defined at a much higher level of abstraction than the typical vocabulary of SOTA ML techniques; for example, deep learning models are essentially always nested compositions of dot products with nonlinearities. Second, they are based on the body of available knowledge that has proven useful in signal processing and analysis over the past few decades. Third, the DSL's grammar only allows operations that make sense in the context of signature identification and can be used to convey expert knowledge about the problem.

[0306] To express the model as a program, we can draw on a wealth of theoretical and practical knowledge about programming language design and semantics. For data representation, we can rely on the formalized approach of type systems, which allow us to reason in a principled and coherent way about data pieces, their relationships, and their operations. To this end, we rely on the basic formalism of algebraic data types, which allows us to systematically create new data types by aggregating and composing existing ones. In some variants (e.g., so-called dependent types), we can "propagate" data characteristics through functions and constrain their output types. The actual processing of the data can then be conveniently expressed using recursive schemes, which provide a universal framework for aggregating and decomposing information for any variable-sized data structure (e.g., time series). Last but not least, DSLs are designed in a way that suits the structural characteristics of the problem.

[0307] In the case of application of the systems, methods, and sensing devices of the present technology to detect Covid-19 infection in subjects (see, e.g., Example 8), structure emerges from a matched case-control setup, with observations spanning multiple “dimensions” related to auscultation point placement, patient position, age, gender, etc. These observable factorization variables can be explicitly incorporated within the DSL. Unlike deep learning approaches, this facilitates distinguishing between correlation and causation by filtering out other confounding variables. By leveraging the structure imparted by the MCC setup, this allows for the capture and exploitation of various structures at multiple levels, among other things, within recordings (by aggregating multiple alternative feature extraction techniques), and between recordings.

[0308] The above mechanisms "regularize" the program synthesis process, making it more likely to find a plausible solution (program) for a given problem. In particular, this avoids overfitting to the available training data and makes it more likely to generalize effectively. This makes it possible to synthesize robust signatures, classifiers, and regression models from a limited number of training examples.

[0309] 1.h. Other features of the system In some variations, full-spectrum vibroacoustic sensing may be selectable by enabling or disabling vibroacoustic sensing functionality based on a software-as-a-service subscription model. One or more sensing devices may transition between a basic or minimal functional state, for example, in which no vibroacoustic sensing is performed. In variations in which the sensing device is a customized medical instrument (e.g., a stethoscope) as described below, the basic functional state may be a mode in which the voice coil is active via a standard air tube, similar to a traditional acoustic stethoscope. Functionality may then be scaled to one or more intermediate or maximum functional levels. At intermediate levels, some functionality is partially enabled or provided at a degraded level, while at maximum functional level, all vibroacoustic sensing features are functional. Furthermore, in some variations, these functional modes may additionally or alternatively be intentionally selected by a user as different operating modes of the sensing device (e.g., a user may disable sensing of inaudible frequencies if such signals are not of interest in a particular application). Additionally or alternatively, in some variations, the control mechanism may be hardwired into the power control or coded via software programming to prevent tampering.

[0310] In some variations, the internal controller may control the functionality of the sensing device based on the level of service available to the user (e.g., as a subscription service). The controller may also authenticate parts of the sensing device (e.g., any disposable components, such as replaceable modules), as well as license data, serial numbers, and / or other data structures, which may allow subscriptions to be identified. Additionally or alternatively, the controller may store the level of service, permissions, subscriptions, and / or any other relevant data to determine the level of service at any given time. The license, serial number, clock and / or calendar data are stored in non-volatile memory so that critical subscription-related information is not lost even under conditions of power loss. The controller may be triggered by receiving a key. The key may, for example, set the level of service, the duration of the service, and / or the number of times the service may be performed.

[0311] Additionally or alternatively, the key may securely link the sensing device to a local and / or cloud-based health record data management solution. A subscriber may, for example, link an authorized computing device, such as a smartwatch, mobile phone, tablet, or computer, directly to the sensing device (e.g., using a keypad, wireless technology such as NFC, and / or optical technology such as reading an image such as a QR code), providing the device with a key through this process. In some variations, the computing device may be a tether to the sensing device, communicating and connecting the device to the cloud, other external devices, etc., via a cable or wirelessly using wireless and / or wired communication modes as described above. In some variations, the information entered into the computing device is combined with encrypted license data, serial number, and / or other data structures on the vibrometer sensing device and relayed to a cloud-based server, which may verify the information and then transmit an operational subscription key combining the subscription start and end dates and / or other details.

[0312] Additionally, in some variations, subscription data may be combined with device usage data (as described above) to support and / or control device functionality, such as for behavior change support in academic or professional settings, institutional policy monitoring, and / or home use.

[0313] 1.i. Sensing Device Modifications As described above, the sensing device may incorporate various components in a modular manner and may have any of a variety of suitable form factors. All-in-one handheld device

[0314] For example, in some variations, the sensing device may include a handheld housing. As described above with reference to FIGS. 3A and 3B and 4A and 4B, in some variations, the housing may be standalone or integrated into the handheld sensing device 300, 400. The sensing device 300, 400 may be used, for example, as a point-of-care device for assessing one or more physical conditions of a subject. The sensing device 300, 400 may have a wide bandwidth capable of detecting the full spectrum of vibroacoustic signals associated with the subject (e.g., between about 0.01 Hz and about 160 kHz). In some variations, the sensing device 300, 400 may be used, for example, as part of a telemedicine platform. It should also be understood that the stackup shown in FIGS. 3B and 4B is exemplary only, and that the components may be packaged in other suitable configurations within the housing.

[0315] 27A and 27B depict an example variation of a sensing device 2700 that includes a handheld housing as part of an integrated handheld device. Similar to sensing device 300, sensing device 2700 may include a handheld housing 2710 having a base 2702, with the housing 2710 substantially enclosing one or more of the modules and various other components (e.g., a vibroacoustic sensor module 2720, an electronics system 2740, at least one power source 2742, an impedance matching diaphragm 2750) as described above. In some variations, the housing 2710 may be assembled from multiple components that couple together via one or more fasteners, mechanical interlocking features (e.g., mating features), or the like. For example, as shown in FIG. 27B , housing 2710 may include a housing body portion 2711 (e.g., which may include a power port 2714) coupled to top cover 2712 and bottom cover 2713 to enclose the components of sensing device 2700. In some variations, sensing device 2700 may further include a context sensor module similar to that described above. In contrast to sensing device 300 shown in FIGS. 3B and 4B , sensing device 2700 may be cuboidal and omit a display. Sensing device 2700 may be used, for example, as a point-of-care wideband scanner device for assessing one or more conditions of a subject, where sensing device 2700 has a wide bandwidth capable of detecting the full spectrum of vibroacoustic signals associated with the subject (e.g., between about 0.01 Hz and about 160 kHz). Like sensing device 300, in some variations, sensing device 2700 may be used, for example, as part of a telemedicine platform. It should also be understood that the stackup shown in FIG. 27B is exemplary only, and that the components may be packaged in other suitable configurations within the housing.

[0316] 28A and 28B depict an exemplary variation of a sensing device 2800 that includes an ergonomic handheld housing as part of an integrated handheld device. The sensing device 2800 may be similar to sensing devices 300, 400, and 2700 described above, except that the housing 2810 within the sensing device may be shaped and contoured to provide a more ergonomic handheld grip. In some variations, the sensing device 2800 may be designed to be held in the left hand, or the right hand, or may be designed to be comfortably held in either the left or right hand of a user. In some variations, as shown in FIG. 28, the housing 2810 may be assembled from multiple components that couple together via one or more fasteners, mechanical interlocking features (e.g., mating features), or the like. For example, as shown in FIG. 28B, the housing may include a housing body having coupleable body portions 2811a and 2811b. The housing body portion may further be coupled to a top cover 2812 and a bottom cover 2813. Additionally, the housing 2810 may be coupled to the base 2802 and configured to substantially enclose one or more of the modules and various other components (e.g., vibro-acoustic sensor module 2828, electronics system 2840, at least one power source 2842, impedance matching diaphragm 2850) as described above. In some variations, the sensing device 2800 may further include a context sensor module similar to that described above. The housing 2810 may include a power port 2814, for example, for recharging the power source 2842. It should also be understood that the stackup shown in FIG. 28B is exemplary only, and that the components may be packaged in other suitable configurations within the housing.

[0317] Wearable devices In some variations, the sensing device may include a wearable housing. The wearable housing may, for example, be coupled to an adhesive patch configured to attach to a surface (e.g., skin) of the subject at an appropriate body location (e.g., chest, stomach, etc.). As another example, the wearable housing may be coupled to suitable clothing (e.g., clothing such as a shirt or jacket, a chest strap, a belly band, an armband, etc.) for detecting vibroacoustic signals from a subject wearing the clothing. Thus, in such variations, the sensing device may enable continuous monitoring of a subject for one or more physical conditions.

[0318] FIG. 29 shows an exemplary variation of a sensing device 2900 configured as a wearable device. For example, the sensing device 2900 may include a housing coupled to an adhesive patch for attachment to a user. The housing may include a top cover 2912 coupleable (e.g., via mechanical interlocking features, fasteners, etc.) to a bottom cover 2913 to substantially enclose the internal components of the sensing device. Similar to what is described above, the sensing device 2900 may include a vibroacoustic sensor module 2920, at least one power source 2942, and an electronics system 2940. The sensing device 2900 may have a wide bandwidth capable of detecting the full spectrum of vibroacoustic signals associated with the subject (e.g., between about 0.01 Hz and about 160 kHz). Additionally, in some variations, the sensing device 2900 may include an impedance matching diaphragm (not shown) and / or a context sensor module (not shown in FIG. 29) that provides additional sensor data for use in analysis. For example, in some variations, the context sensor module may include auxiliary sensors such as a microphone and / or a 9-axis inertial measurement unit (e.g., including a 3-axis accelerometer, a 3-axis gyroscope, and a 3-axis magnetometer). FIG. 43 shows example data from such an example variation of a context sensor module. In the arrangement shown in FIG. 29 , a continuous opening from the user's surface to the vibroacoustic sensor module 2920 may be provided by aligned openings in the mounting backing 2902, the lower cover 2913, and / or the electronics system 2940. For example, the mounting backing 2902 may include an adhesive patch. It should also be understood that the stackup shown in FIG. 29 is exemplary only, and that the components may be packaged in other suitable configurations within the housing.

[0319] In some variations, a sensing device similar to sensing device 2900 may be coupled to a garment to detect and measure vibroacoustic data and / or other sensors when the garment is worn by a subject. For example, attachment backing 2902 may include an adhesive patch that may be attached to an outer or inner surface of the garment. As another example, the sensing device may be attached to the garment by fastening a housing (e.g., lower cover 2913) to attachment backing 2902 disposed on an opposing surface of the garment, such that a layer of the garment is sandwiched between the backing and the housing. For example, attachment backing 2902 may be disposed adjacent to the inner surface of the garment and the housing may be disposed on the outer surface of the garment, with the attachment backing and housing coupled together via one or more fasteners (e.g., adhesive, mechanical fasteners, magnets, etc.) and / or interlocking components (e.g., threads, snap-fit ​​features, latches, etc.). As yet another example, the housing of sensing device 2900 may be sewn to the garment. Specialized equipment

[0320] The sensing device and / or other components of system 100 may be incorporated into other suitable medical diagnostic equipment, such as a stethoscope. For example, a stethoscope including the sensing device may be used by a clinician to examine a subject and collect data including audible and inaudible vibroacoustic signals from the heart, lungs, intestines, etc. stethoscope

[0321] For example, as shown in FIG. 30 , a stethoscope device 3000 may comprise a sensing device including a handheld housing 3010 (e.g., a chestpiece) having a vibroacoustic sensor module 3020. The vibroacoustic sensor module 3020 may be similar to any of the variations of the vibroacoustic sensor modules described above. For example, as described above, the vibroacoustic sensor module 3020 may comprise at least one deflection structure (e.g., one or more bending arms, membranes, spiders, etc.) and one or more suitable sensors configured to interact with the deflection structure to detect and measure vibroacoustic signals. In some variations, the vibroacoustic sensor module comprises a voice coil. In some variations, the stethoscope tubing leading to the earpiece 3060 may be removably coupled to the vibroacoustic sensor module 3020 (e.g., using a mating connector, magnets, etc.). In some variations, vibroacoustic sensor module 3020 (or suitable components thereof, such as an impedance diaphragm as described above) may be separately manufactured and configured to replace one or more components (e.g., a dome or diaphragm) of an existing stethoscope (e.g., a conventional acoustic stethoscope). In this regard, some components of a sensing device, such as sensing device 400, may be adapted to be retrofitted to a conventional stethoscope.

[0322] The vibroacoustic signal from the vibroacoustic sensor module 3020 may be transmitted to at least one junction box 3040 along the piping of the stethoscope device 3000 or to an electronics system, which may be located in any suitable location for processing the signal. At least the vibroacoustic signal in the audible frequency range traverses through the piping and can be heard by a user via the earpiece 3060. Additionally or alternatively, the junction box 3040 may include one or more connectors that allow at least one peripheral device (e.g., headphones) to be connected to the stethoscope in a wired manner, although in some variations, the acoustic data may be transmitted to the peripheral device in a wireless manner via a communications module (e.g., via WiFi, a cellular network, Bluetooth, etc.), such as those described above. The junction box may also include one or more connection ports by which speakers, headphones, and / or air tubes may be connected to the vibrometer sensing device. Furthermore, in variations in which the electronics system includes a communications module with one or more antennas for wireless transmission, the antennas may be included in the piping to enable optimization for range and / or data rate. In some variations, the antenna may be spaced an appropriate distance from other electronic devices to reduce interference and improve transmission quality.

[0323] In some variations, the sensing device may function as a stethoscope by coupling with a smartphone or other device. panel

[0324] 31A-31C, in some variations, a sensing device 3100 (hereinafter referred to as a "panel 3100" or "panel unit 3100") having a panel-like configuration is provided. The panel 3100 can be mounted on a supporting surface such as a wall, ceiling, or doorway, or can be freestanding. The panel 3100 can be installed in a room, hallway, vehicle, passageway, checkpoint, doorway, vehicle, and other areas to detect sensor signals from a subject to diagnose several physical conditions of the subject. The sensing device 3100 can operate 1 cm, 5 cm, 10 cm, 25 cm, 50 cm, 100 cm, 150 cm, 200 cm, 250 cm, or 300 cm away from the subject or the subject's clothing. The panel 3100 may be camouflaged to prevent it from being recognized by the subject. The thickness of the panel unit 3100 may be less than 1 mm, 2 mm, 3 mm, 4 mm, 5 mm, 7.5 mm, 10 mm, 15 mm, 20 mm, 25 mm, 30 mm, 40 mm, 50 mm, or 75 mm.

[0325] The panel 3100 includes a frame 3110 defining an opening 3120, and a membrane 3130 extending at least partially across the opening 3120 and supported by the frame 3110. A vibroacoustic sensor assembly 3140 is coupled to the frame 3110, such as by a support member 3150, and configured to convert vibrations of the diaphragm 3130 into an analog or digital signal, for example. Panel Vibro-Acoustic Sensor Assembly

[0326] In some variations, the vibroacoustic sensor assembly 3140 may be based on a voice coil type transducer, such as the vibroacoustic transducer 1600 described with respect to Figures 16 and 17. In this regard, the vibroacoustic sensor assembly 3140 may include a conductive coil with a fixed magnet setup connected to the frame 3110. The vibroacoustic transducer may be attached to the frame 3110 by a support member 3150. The vibroacoustic transducer may be mounted centrally or asymmetrically with respect to the frame 3110, the options of which are described below.

[0327] In some other variations, the vibro-acoustic sensor assembly 3140 may be any type of sensor readout element or combination of sensor readout elements, such as, but not limited to, the following: Voice coils with extremely compliant spiders or no spider at all for sensitive membrane bending pickups. Highly compliant spiders can be further shaped as flexures to increase compliance and reduce out-of-plane movement. With no spider at all, the spring constant added before the spider is removed from the moving system. Thus, either the voice coil or the magnet is integrated into the bending membrane, but both have low inertia due to their mass and therefore low resistance to membrane movement. Electrical potential sensors (EPS) can pick up subtle movements of nearby objects due to the electrostatic field disturbances they cause. Thus, an EPS located close to the membrane can sense the movement of the vibrating membrane. In contrast to a voice coil, such a system could do without adding mass or an additional spring constant to the system, thus preserving the original compliance of the membrane. Capacitive pickups, similar to condenser microphones, are a direct alternative to EPS pickups but require a layer capable of generating a charge on top of the membrane. A fixed metal plate is placed behind the membrane in close proximity, completing the two components of a capacitor with the air gap between them acting as a dielectric. The metal plate can be any size, from very small to the entire size of the membrane. Additionally, the layer on top of the membrane can be a conductive material that is polarized through an applied voltage (commonly known as a phantom voltage) or an electret material that provides a quasi-permanent charge or dipole polarization. In either case, the added layer adds mass to the vibrating membrane, resulting in inertia. The magnetic field disturbance sensor is a voice coil sensor, but does not incorporate a voice coil component within the membrane. The sensor's magnetic field is generated through a ferroelectric layer on the membrane. Membrane vibrations modulate the magnetic field, which in turn induces a current in the voice coil winding, generating a signal. A light source (e.g., a laser) and a photodetector are positioned behind the membrane. The light source is positioned to direct an energy beam at the membrane, and the photodetector is positioned to detect the energy beam reflected from the membrane and measure the change in angle of the reflected energy beam (which reflects the movement of the membrane). The angle of reflection may depend on the local bending of the membrane, which vibrates with the incident pressure wave. The photodetector may comprise a photodiode array, from which the angle of reflection is determined by the particular photodiode in the array that captures the majority of the reflected signal intensity. Strain sensors can be positioned directly on the membrane surface at strategic locations to detect the movement of the membrane, e.g., a layer of PVDF. Acoustic echo Dopplers impinge on a high frequency acoustic signal behind a membrane, which is reflected from a detector. The membrane vibration is frequency modulated onto the Doppler carrier frequency, and demodulation results in membrane vibration pickup. Acoustic Dopplers can operate in either pulsed wave or continuous wave mode. Optical time-of-flight (TOF) sensing, which uses a light source that emits a pulse and a sensor that measures the time it takes for the reflected signal to arrive. Laser Doppler interferometry, or self-mixing interferometry, exploits the Doppler effect and interference. As a result, this light-based approach offers higher SNR and amplitude resolution, as opposed to acoustic echo Doppler. Instead of laser light, the setup can include any other commonly known radar sensing technology such as time-of-flight radar, radar Doppler, ultra-wideband radar (UWB), etc.

[0328] In addition to laser and radar, vibration pickups can be based on any frequency of electromagnetic waves and can be combined with the same basic methods such as ToF and the Doppler effect. Positioning of the vibroacoustic sensor assembly relative to the membrane

[0329] The vibroacoustic sensor assembly 3140 or other sensor readout element may be positioned at any suitable location with respect to the edge of the membrane 3130. In some variations, as illustrated, the vibroacoustic transducer is positioned centrally with respect to the edge of the membrane 3130. However, in other variations, the vibroacoustic transducer need not necessarily be at the center of the membrane 3130. There are advantages to placing the vibroacoustic transducer off-center at any suitable location, particularly considering that the membrane 3130 may be excited by higher eigenmodes. For example, if a voice coil transducer is placed at the center and a higher eigenmode has a node at the center, there will be no displacement at the center and no signal will be measured, but in fact the membrane is actually vibrating.

[0330] For example, consider a membrane 3130 that has multiple eigenmodes based on its geometry that form nodes (points of no displacement) on the membrane 3130. For example, if the membrane 3130 has four eigenmodes in a 2x2 configuration, there will be a node at the center of the membrane 3130. This is also true when the membrane 3130 has two more eigenmodes that form nodes (points of no displacement) in the central portion of the membrane 3130. In these cases, and in other eigenmode situations not described, a centrally positioned vibroacoustic transducer 3140 will not be optimally positioned to detect vibrations of the membrane 3130. Therefore, the positioning of the transducer with respect to the membrane 3130 may be selected by considering the eigenmodes of the membrane 3130. In variations of the panel 3100 in which the vibroacoustic sensor assembly 3140 includes EPIC potential sensors and / or capacitive sensors, the electrodes of such sensors may be sized to cover the size of the membrane, which may minimize localized effects of eigenmodes, such as no bending / displacement of the membrane at the nodes. In some other variations, the vibroacoustic sensor assembly 3140 may be configured not to sense beyond the first membrane resonance caused by the first eigenmode, which may also minimize or make redundant the effect of the eigenmode.

[0331] Panel frame The frame 3110 may be of any suitable size or shape, with its dimensions and configuration selected based on the desired application and desired frequency range of detection. The frame 3110 may be fabricated from plastic, wood, metal, composite, glass, ceramic, or any other suitable material capable of withstanding the tension of and / or supporting the attached membrane 3130. Although illustrated as a rectangle, the frame 3110 may be circular, oval, trapezoidal, regular polygonal, or irregular polygonal. In some variations, the frame 3110 is subdivided to define multiple openings for coupling with separate membranes 3130.

[0332] Panel membrane A membrane 3130 may be attached to a first side 3152 of the panel and a back cover 3153 may be provided on a second side 3155 of the panel 3100, thereby defining a cavity 3156 between the membrane 3130 and the cover 3153. The membrane 3130 is configured to vibrate at a frequency related to a desired detection frequency range, such as a vibroacoustic range of a subject. One or more of the parameters of the material, weight, size, and tension of the membrane 3130, as well as the shape or size of the cavity 3156 behind the membrane 3130, may be tailored to achieve the desired frequency range.

[0333] A larger membrane 3130 with less stiffness tends to pick up low frequencies well, while a stiffer membrane picks up higher frequencies but attenuates lower frequencies. The weight of the membrane 3130 itself, or anything connected to it in general, creates inertia during vibration, opposing and attenuating the incident vibration acoustic signal (and potentially increasing the reflection of acoustic waves). A voice coil transducer connected to the membrane 3130 means that the spider component represents an additional spring in the system, which increases the stiffness of the membrane and reduces the compliance of the sensor pickup. An attached voice coil portion may also add inertia to the membrane.

[0334] For example, a highly compliant membrane provides a good signal-to-noise ratio that favors low frequencies (e.g., only 0-100 Hz). Similarly, a larger membrane also favors lower frequencies. A smaller membrane 3130 can detect a higher bandwidth or higher frequencies. A thicker membrane 3130 can detect a higher bandwidth, higher frequencies, generally due to a higher membrane bending stiffness. A thinner membrane 3130 can detect lower frequencies because it tends to be more compliant, all other parameters being equal. A membrane with higher tension can detect a higher bandwidth, and less deflection can result in a lower sensor amplitude and therefore a lower signal-to-noise ratio. A membrane with lower tension can detect a lower bandwidth (better signal-to-noise ratio) as it becomes more compliant and higher deflection caused by the same incident acoustic wave.

[0335] In general, there is a trade-off between different values ​​of bending stiffness and, therefore, the ability to pick up low-amplitude waves. Low bending stiffness results in a compliant membrane being able to pick up very low-amplitude waves (e.g., <20 Hz). However, the resonant frequency and subsequent roll-off of a very compliant membrane is very low, thus hindering its ability to pick up higher frequencies, especially above some threshold frequency, e.g., >100 Hz. In contrast, high bending stiffness results in higher resonant modes of the membrane, trading off the ability to pick up higher frequencies for lower frequencies with small amplitudes.

[0336] In some variations, instead of making a trade-off for the overall frequency range, the panel 3100 can be divided into smaller sub-panels 3156 (FIG. 31C), each with the same or different membrane 3130 stiffness and respective transducer for a particular optimal frequency range. Additionally, the sub-panels 3156 can be different sizes, e.g., a membrane 3130 for picking up <20 Hz can have a larger surface area than a membrane 3130 for detecting >100 Hz. It will be understood that panels 3100 without sub-panels as well as panels 3100 with sub-panels 3156, with some variations, are within the scope of the present technology.

[0337] In this manner, a wider overall frequency range may be detected by using the sub-panel 3156. In some variations, the sub-panels may be separate from one another. In other variations, the configuration of the sub-panels illustrated in FIG. 31C may differ from the configuration illustrated in a manner known by those skilled in the art. In some variations, the panel 3100 may be considered to include multiple transducers operating in the same or different operating modes. The multiple transducers may be arranged in an array.

[0338] In some variations, the membrane 3130 is a compliant material such as a thermoplastic or thermoset elastomer. In other variations, the membrane 3130 may comprise a metal, or an inorganic material such as silica, alumina, or mica, or a woven fabric, fiberglass, Kevlar™, cellulose, carbon fiber, or combinations and composites thereof. In some variations, the membrane 3130 is provided with a protective layer, which may include an acoustically transparent layer, such as foam, positioned on the outward-facing side of the membrane 3130 at a distance of about 1 mm to about 100 mm.

[0339] The membrane 3130 may be attached to the frame 3110 in any manner, such as with an adhesive. The profile of the membrane 3130 when attached to the frame 3110 may be planar, convex, or concave. If the membrane 3130 is under tension, it may be attached to the frame 3110 in a manner that applies uniform tension or differential tension along different orthogonal axes. The membrane 3130 may be a stretched sheet. In some variations, the membrane 3130 may be free-standing or under compression rather than tension. Damping material may be provided to damp movement of the membrane.

[0340] With respect to the cavity 3156, in some variations of the panel 3100, the back cover 3154 provides a varying degree of sealing against the cavity 3156. For example, in some variations, the back cover 3154 may be omitted. In this case, the pressure on either side of the membrane 3130 may be quickly equalized. However, the membrane 3130 can generally only bend / vibrate if there is a difference in pressure on both sides. In particular, at low frequencies, it is not possible to measure such low signals, as there is sufficient time for the air to continuously equalize the pressure on both sides of the membrane 3130 after the pressure wave is incident. It is then also clear that in an open-back setup, static pressure cannot be measured.

[0341] In some other variations in which a back cover 3154 is included in the panel 3110, the back cover 3154 may function to seal the cavity 3156 to different degrees. In an extreme example, the back cover 3154 may include a solid piece that seals the cavity 3156. This can be thought of like a pressure sensor that measures static pressure relative to an internal reference pressure. While it measures down to DC (static pressure), the static pressure opposes movement of the membrane 3130 in response to AC signals, especially the larger the input vibration amplitude. Additionally, a completely sealed cavity will cause the membrane 3130 to bend outward or inward when the external and internal pressures are not equal, for example, when altitude changes. As a result, depending on the volume of the cavity, dynamic (AC) measurements at higher frequencies and larger amplitudes may have a lower signal-to-noise ratio (SNR).

[0342] In some variations, the back cover 3154 includes openings 3162 that allow airflow through and into the cavity 3156. The size, number, and placement of these openings 3162 may be optimized according to the desired frequency detection range and acceptable signal-to-noise ratio, and may also be considered a cavity impedance optimization due to the cavity volume itself. For low-frequency detection (below 20 Hz), low-frequency pressure waves allow sufficient time for an equilibrium state to be created on either side of the membrane 3130. Thus, the configuration of the openings 3162 must consider a trade-off between allowing air in and out of the interior of the cavity 3156 (in response to positive or negative pressure waves) to reduce pressure and delaying the equilibrium process long enough to capture infrasonic pressure waves. Thus, pressure on either side of the membrane 3130 equalizes with a time constant, and vibrations at frequencies corresponding to a time period equal to or less than that equilibrium time constant can be measured. In some variations, the dimensions of the panel are about 7 inches (width), about 9.75 inches (height), and about 0.5 inches (depth). Experimental data obtained with this variation of the sensing device is presented in Example 9.

[0343] The openings 3162 can be of any shape (round, square, rectangular) and size and number. Instead of simple openings, the openings can be structures such as tubes of various diameters and lengths, such as those commonly found in acoustic subwoofers. The opening structures can be anything that allows air flow between the cavity and the outside environment, and are therefore not limited to tubes only. In one exemplary embodiment, the back cover 3154 can have a single small tube to equalize the internal DC pressure of a low-frequency-optimized panel with a large cavity.

[0344] In some other variations, the cavity 3156 inside the panel 3100 may be divided into two lateral sections. The partition between the two cavities is perforated based on design requirements to allow air exchange between the two cavities. In one variation, the cavity closest to the membrane 3130 is a small cavity, while the cavity toward the rear is a larger cavity, serving as an air "reservoir." The entire unit is either completely sealed from the environment or sealed with a small hole or tube to allow pressure equalization with the environment in the event of gradual, nearly DC-type pressure changes, such as those caused by altitude changes. A dual-cavity setup is particularly important when capacitance or potential sensors are used. For example, in a capacitive sensing approach, the conductive plate behind the membrane 3130, required to form a capacitor, may be of a similar size to the membrane to maximize sensitivity. This plate should be close to the membrane to maximize the capacitance between the membrane and the plate, so the cavity formed is small, creating increased air pressure under the vibrating membrane when an unperforated plate is used. Thus, the perforations in the plate connect the small cavity to the larger rear cavity, reducing the pressure.

[0345] In summary, the vibration acoustic detection range of the sensing device 3100 when embodied as a panel 3100 can be considered as a function of various parameters related to the membrane 3130 (e.g., stiffness, material, surface area, etc.), the sensor element that reads the vibrations (e.g., voice coil, capacitance, optical, acoustic (echo Doppler), radar, etc.), and pressure equalization based on, for example, the size of the cavity 3156 and the opening 3162 in the back cover 3154.

[0346] Panel front cover Other variations of the sensing device 3100 having a panel-like configuration are illustrated in FIGS. 31D and 31E . A front cover 3159 is provided on a first side 3152 of the panel 3100. The front cover 3159 may be more rigid than the membrane 3130. The front cover 3159 may provide environmental and mechanical protection for the membrane 3130. The front cover 3159 may have any type of surface finish or configuration. For example, in some variations, the front cover 3159 is highly reflective, such as a mirror. In some variations, the front cover 3159 may include an output display unit 3170. The output display unit 3170 may include any manner of markings and indicators, such as one or more of the likelihood that the subject has a given physical condition (e.g., displayed as a red or green light or other indicator), at least a portion of the data acquired by the sensor (e.g., subject physiological data such as an ECG reading, environmental data), etc. In some variations, the front cover 3159 may be at least partially a mirror and at least partially an output display, such as the output display unit 3170. The front cover 3159 may be configured to extend substantially vertically when supported on a supporting surface, such as a wall or floor. The front cover 3159 is perforated to allow sound pressure to pass through without significant attenuation, and the perforations may be down to a micrometer in size.

[0347] Adding a Sensor Module The panel unit 3100 may incorporate other sensor assemblies based on non-contact detection of signals associated with the subject or the environment, such as, but not limited to, one or more of an echo Doppler sensor module, a motion sensor module, a temperature sensor module, a VOC sensor module, a machine vision sensor module, a contextual sensor module, etc. The sensor modules of the panel unit 3100 may be configured to monitor or detect, for example, Covid-19 infection in the subject by detecting signals related to respiratory function, body temperature, gastrointestinal function, bladder motility, water / fluid retention in the legs (edema), peripheral vascular disease, etc.

[0348] Thus, in some embodiments, a device for detecting vibroacoustic signals related to a subject is provided, the device comprising: a frame defining an opening; a sheet extending across the opening and supported by the frame, the sheet configured to vibrate at a frequency related to the subject's bioacoustic range; and a transducer for converting the vibrations of the sheet into a data stream comprising vibroacoustic data. In some embodiments, the frame is more rigid than the sheet. In some embodiments, the sheet assumes one or more of a planar, convex, or concave shape when supported by the frame. In some embodiments, the transducer comprises one or more of a voice coil, a piezoelectric element, a capacitive element, or a laser microphone. In some embodiments, the data stream comprises one or both of an analog signal and a digital signal. In some embodiments, the frame defines a periphery of the sheet. In some embodiments, the frame is configured to support the sheet in one or more of a tensioned state, a neutral state, and a compression state across the opening. In some embodiments, the sheet is supported by the frame such that a first tension in the sheet along a first axis differs from a second tension along a second axis. In some embodiments, a damping element is further provided for damping movement of the seat. In some embodiments, the seat is connected to or integral with the transducer. In some embodiments, the transducer includes multiple transducers operating in the same or different operating modes. In some embodiments, one or more optical sensors are further provided for capturing optical data related to the subject, and the data stream includes optical data from the one or more optical sensors and vibro-acoustic data from the transducer.

[0349] Base Unit 32A and 32B, a panel-like sensing device 3200 (also referred to as "panel unit 3200") that may correspond to the panel unit 3100 is provided, along with a base unit 3210 incorporating one or more sensor modules. The one or more sensor modules may be any one or more of the sensor modules described herein, such as a vibro-acoustic sensor module, a bioelectric sensor module, or a capacitive sensor module. The panel-like sensing device 3200 and the base unit 3210 may be an integrated single unit (FIG. 32A) or may comprise separate units spaced apart from one another (FIG. 32B). The panel unit 3200 and the base unit 3210 may be positioned orthogonal with respect to one another. Advantageously, this orientation may enable the sensor modules associated with the panel unit 3200 and the base unit 3210 to detect physiological parameters along different planes of the subject's body, providing the ability to obtain complementary data sets for physical state determination.

[0350] The base unit 3210 may be adapted to support a body part of the subject, such as the foot, leg, arm, back, chest, head, etc. In these cases, it will be understood that the system 100 provides both contactless signal detection (from the panel unit 3210) and contact-based signal detection (from the base unit 3210), whether by direct skin contact or indirect contact through clothing and / or footwear.

[0351] The base unit 3210 may include a platform having an upper surface 3220 for the subject to stand on, configured to be supported on a supporting surface such as the ground during use. Markings may be provided on the upper surface 3220 to indicate where the subject should place their feet. As can be seen, the base unit 3210 may be relatively flat or have a stepped structure. One or more sensor assemblies included in the base unit 3210 may be configured to acquire data from the subject while the subject is wearing footwear such as shoes or socks. In some variations, the base unit includes a bioelectric sensor assembly. Other sensor assemblies may include those that detect vascularization, heat, weight, etc. The base unit 3210 may also be configured to emit one or more signals to the subject. For example, the base unit 3210 may be configured to vibrate to detect the subject's physiological response to the vibrations. The base unit 3210 may include a BCG sensor module.

[0352] In some variations, the sensing device 3200 comprises a panel unit 3100 embodied in the configuration of FIG. 32A or 32B , and optionally includes a base unit 3210. The panel unit 3100 includes a vibro-acoustic sensor module, an echo-Doppler sensor (continuous wave), optionally a thermal sensor module (e.g., a thermopile with mKelvin to 1 Kelvin accuracy, and optionally a blackbody reference for calibrating temperature readings based on environmental temperature), a 3D machine vision camera, optionally one or more environmental sensors (e.g., one or more of ambient temperature, GPS, barometric pressure, altitude, ambient noise, light, etc.), and optionally a MEMS microphone. The panel unit and sensor module are configured so that the panel unit can be positioned substantially vertically and a subject can stand in front of the panel unit. Including the echo-Doppler sensor module along with the voice coil sensor module can improve the signal-to-noise ratio of the detected vibro-acoustic signal.

[0353] The base unit 3210 includes markers on its outer surface that indicate where the subject should stand. In some variations, the subject is invited to stand with their chest facing the standing unit 3325. The markers may be images of feet. Optionally, the base unit 3310 may also include a capacitance sensor module (e.g., for measuring galvanic skin response), and optionally a BCG sensor module.

[0354] In some variations, an adjustment mechanism is provided within the panel unit to adjust the position of one or both of the echo Doppler sensor module and the vibro-acoustic sensor module to optimize the height of the sensor modules for optimal or sufficient signal detection. In some variations, the adjustment mechanism allows for vertical position adjustment. In some variations, the adjustment mechanism may also allow for left-right position adjustment. The adjustment mechanism may include a linear motion system (not shown). An additional sensor module (such as a 3D camera) may be included to detect the subject's height and automatically adjust the height of one or both of the echo Doppler sensor module and the vibro-acoustic sensor module. The system 10 may be configured to obtain data from the 3D camera (machine vision sensor module) to detect the subject's eye height and estimate the subject's torso height. If one or both of the echo Doppler sensor module and the vibro-acoustic sensor module are not aligned with the subject's torso height, the system 10 may be configured to adjust the height of one or both of the echo Doppler sensor module and the vibro-acoustic sensor module using the linear motion system. Optionally, the sensing device may include a thermal sensor module for detecting the temperature of the subject, in communication with a 3D camera for detecting the location of the subject's tear duct. A light, such as in the form of a ring, may be provided to illuminate the subject. Illuminating the subject's tear duct can facilitate locating the tear duct and measuring its temperature. In some variations, the thermal sensor module may be configured to move to target the identified tear duct. In other variations, the system 10 may be configured to receive data regarding the subject's chest height or otherwise the desired height of the echo Doppler sensor module and / or the vibro-acoustic sensor module, and cause the adjustment mechanism to move the echo Doppler sensor module and / or the vibro-acoustic sensor module in response. Sufficient signal detection from the subject may be obtained in about 5 to about 15 seconds, and in some variations, in 10 seconds.

[0355] Gateway / Kiosk / Walkthrough Yet another variation in sensing device form factor is illustrated in FIGS. 33A-33C , which provides a sensing device 3300, such as sensing device 3100 or 3200, integrated into a gateway configuration. This is also referred to as a “kiosk” configuration. The gateway configuration may include a base unit 3310, such as base unit 3210, a top unit 3320, and an upright unit 3325. Other support structures may also be provided. The sensing device 3300 is configured for contactless measurement of vibroacoustic data as well as other sensor data of a subject in proximity to the sensing device 3300. The top unit 3320 may be configured to be positioned near the subject's head, such as on top of the head. The top unit 3320 may include sensors, such as sensors associated with EEG. The top unit 3320 may have an adjustable height from the floor or from the base unit 3310 to accommodate subjects of different heights.

[0356] In some variations, the standing unit 3325 is configured to house one or more sensor modules, such as a vibro-acoustic sensor module, for example based on the voice coil transducer 1600 of Figure 16, a continuous-echo Doppler sensor module (alternatively may be pulsed wave), a thermal sensor module (such as a thermopile thermal sensor module) with mKelvin to 1 Kelvin accuracy and optionally a blackbody reference for calibrating temperature readings based on environmental temperature, a MEMS microphone (optional), environmental sensors (optionally one or more of ambient temperature, GPS, barometric pressure, altitude, ambient noise, light, etc.), a 3D machine vision depth camera, and optionally one or more of a capacitive sensor module, such as a charged plate, for measuring galvanic skin response. This may be embodied as a base unit for the subject to stand on.

[0357] In variations including an echo Doppler sensor module, the echo Doppler sensor module may include any combination of emitters and receivers, such as one emitter and one receiver (FIG. 33D), two emitters and one receiver (FIG. 33E), one emitter and two receivers (FIG. 33F), or two emitters and two receivers (FIG. 33G). In variations with dual emitters, an interference pattern focused on the plane of the person's chest may be achieved, resulting in reflection of an amplified signal from that area. Meanwhile, surrounding walls may reflect unamplified signals, increasing the signal-to-noise ratio. Variations using dual receivers may allow for directional preference to be added to the incoming signal, similar to a directional microphone array, so that signals from the person's chest are amplified over signals arriving from the surrounding area. The echo Doppler sensor module may function as either CWD, PWD, or simply time-of-flight. 33D-33G show the signal directionality when the echo Doppler sensor module is positioned in a substantially vertical unit, such as a panel unit component of a gateway.

[0358] In some variations, the signal-to-noise ratio can be adjusted by optimizing the ultrasound carrier frequency. There is a range of achievable ultrasound frequencies with frequency-dependent attenuation characteristics. Overall, ultrasound attenuation varies exponentially with distance, but also exponentially as frequency increases. By selecting a higher ultrasound carrier frequency, the attenuation of reflected signals from walls farther away from the subject is exponentially higher than at lower frequencies. Because reflected signals are unwanted and considered noise, reducing their influence improves the signal-to-noise ratio of the desired signal reflected from the subject.

[0359] For example, when a subject is sitting 50 cm away from the echo Doppler sensor module, there is a total ultrasound signal path of 1 m (twice the distance). If the wall behind the subject is 2 m away from the echo Doppler sensor module, the total signal path is 4 m.

[0360] Referring to Figure 33H, a 50 kHz carrier signal is attenuated at approximately 2 dB / m at room temperature. As a result, the signal reflected from the subject is attenuated by approximately 2 dB, and the signal reflected from the wall is attenuated by approximately 8 dB. Therefore, when the signals are recombined in the ultrasound receiver, the difference between the two is 6 dB, which means that the wall-reflected signal is 50% of the amplitude (or 25% of the power) of the subject's signal, which is a very significant disturbance to the signal.

[0361] In contrast, when a 200 kHz carrier is selected, the attenuation at room temperature is approximately 9 dB / m, resulting in a signal reflected from the subject being attenuated by approximately 9 dB, and a signal reflected from the wall being attenuated by approximately 36 dB. The difference between the two is 27 dB, meaning that the wall-reflected signal is 4.5% of the subject's signal's amplitude (or 0.2% of its power). Because the subject's signal is attenuated by 9 dB, which corresponds to approximately 30% of its original amplitude—or 10% of its original power—it is possible that either the transmitter signal is further amplified, or amplification is added after the signal is received. However, because the amplification operates linearly on the overall signal, the improved signal-to-noise ratio due to the exponential loss remains.

[0362] By way of background, the emitted ultrasonic signal (carrier signal) can be defined as: s e (t)=A c cos(ω c t), where A c is the magnitude of the carrier, ω c ω c =2πf c Carrier angular frequency based on s e is the emitted signal. The emitter signal is frequency modulated by the chest vibrations, and the received signal, including the Doppler shift, results in

[0363]

number

[0364] where c is the speed of sound in the medium used, such as approximately 345 m / s in air at room temperature, and d(t) is the skin displacement of the target body placement. The received signal is then coupled to an ultrasonic carrier wave s e As a result of demodulation by (t), the demodulated signal is obtained by the following equation.

[0365]

number

[0366] Solving for chest displacement yields the following equation:

[0367]

number

[0368] Sensor module positioning In some variations, both the echo Doppler sensor module and the vibroacoustic sensor module are included in the stand unit of the gateway sensing device 3300, sensing device 3200, or sensing device 3100. Both sensor modules are positioned at approximately chest height of an average subject. In some variations, an adjustment mechanism is provided in the stand unit to adjust the position of one or both of the echo Doppler sensor module and the vibroacoustic sensor module. In some variations, the adjustment mechanism allows for vertical position adjustment. In some variations, the adjustment mechanism may also allow for left-right position adjustment. The adjustment mechanism may include a linear motion system (not shown). An additional sensor module (such as a 3D camera) may be included to detect the subject's height and automatically adjust the height of one or both of the echo Doppler sensor module and the vibroacoustic sensor module. System 10 may be configured to obtain data from the 3D camera to detect the subject's eye height and estimate the subject's torso height. If one or both of the echo Doppler sensor module and the vibro-acoustic sensor module are not aligned with the height of the subject's torso, the system 10 may be configured to adjust the height of the echo Doppler sensor module and / or the vibro-acoustic sensor module using a linear motion system. In other variations, the system 10 may be configured to receive data regarding the subject's chest height or otherwise desired height of the echo Doppler sensor module and / or the vibro-acoustic sensor module and to cause the adjustment mechanism to move the echo Doppler sensor module and / or the vibro-acoustic sensor module in response.

[0369] In some variations, the receiver and emitter of the echo Doppler sensor module each have specific signal characteristics depending on the angle. In some variations, the echo Doppler can function within a range of approximately +90 degrees to -90 degrees. In some other variations, the echo Doppler can function within a range of approximately 360 degrees. In other variations, the echo Doppler sensor module is configured to function within a range of approximately ±45 degrees to concentrate the signal energy into a smaller volume while reducing ubiquitous reflections.

[0370] In some variations, the echo Doppler sensor module includes a receiver component, such as an ultrasonic microphone, such as the avisoft-bioacoustics CM16 / CMPA, or a MEMS microphone, such as the invensense ics-41352 / . In some variations, the echo Doppler sensor module includes an emitter component, such as the Prowave 400EP250 or Prowave 400st-R160.

[0371] Modularity In a variation of the gateway-like form factor of the sensing device, the units of the gateway (upright unit, top unit, and base unit) may be configured as modules that allow the components to be flat-packed or otherwise compacted to facilitate mobility and transportation between uses (FIGS. 33A-33C). To this end, one or more portions of the gateway setup may reflect a tessellation pattern that naturally allows for efficient folding and unfolding of the components.

[0372] One such folding pattern, known as the Miura fold, a periodic method of tiling a plane using the simplest mountain and valley folds in origami, is used as the basis for the tessellation pattern of the gateway components. The folded Miura folds are ideal for packaging rigid structures such as solar panels because they can be folded into a flat, compact shape and unfolded in one continuous motion. They also appear in nature in various contexts, such as insect wings and certain leaves. One or more of the panel unit, base unit, and top unit may implement the tessellation pattern.

[0373] In some variations, one or more accessories, such as a beacon or a conversion patch, may be provided that can be attached to the subject and communicate with the processor. A beacon is a device attached to a part of the subject's body that facilitates sensing a specific state, such as movement or posture. Over the past few decades, state-of-the-art techniques and algorithms have been developed for cooperative and non-cooperative attitude determination using electro-optical (EO) sensors. EO sensors consume low power and can be used to estimate all attitude parameters. As a result, such sensors are preferred for this application. Generally, EO sensor systems can be classified as passive systems, systems consisting of a single (monocular) or multiple (stereo) cameras, and active light detection and ranging (LIDAR) systems. Among these systems, monocular vision systems minimize hardware complexity and cost and can be used for remote monitoring. Stereo vision systems use multiple cameras to obtain three-dimensional (3D) information about a target. However, monocular and stereo vision systems suffer from the same handicaps as other vision systems, such as sensitivity to lighting conditions and difficulty separating objects from complex backgrounds. In contrast, LIDAR is robust to illumination differences and can acquire both position and intensity data in 3D, but LIDAR systems consume more energy, require significant computational effort, and have high complexity, resulting in poor real-time performance. Therefore, after weighing the pros and cons of various methods, many research institutes and scholars have opted to rely on monocular vision-based pose determination.

[0374] Typical attitude determination methods typically rely on artificial beacons precisely attached to the target. A proximity sensing sensor (PXS) consists of a camera and light-emitting diode (LED) array on a chaser and a set of passive markers on the target. The LEDs emit pulsed visible light within a 30° cone to illuminate the markers. Simultaneously, the camera captures an image containing the markers. A data processing unit then calculates the relative attitude using complex image processing algorithms. Experimental results demonstrate the high performance of this method: the PXS measurement frequency is 2 Hz, and the accuracy of relative position is on the centimeter scale and the accuracy of relative attitude is on the tenth-degree scale. Similar to PXS, the Advanced Video Guidance Sensor (AVGS) designed by the Marshall Space Flight Center and the Vision-Based System (VBS) designed by the Technical University of Denmark both require artificial beacons, either passive (reflector) or active (LED) markers.

[0375] Bluetooth Low-Energy (BLE) beacon-based indoor positioning is a promising approach, especially for location-based services (PbS) applications. It has low deployment costs and is suitable for a wide range of mobile devices. Existing BLE beacon-based positioning methods can be categorized into range-based and fingerprinting methods. In range-based methods, the beacon location must be known before positioning. In fingerprinting-based methods, a reference fingerprinting map (RFM) is a prerequisite. Many existing methods focus on performing positioning assuming that the beacon location or RFM is known. However, in practical applications, determining the beacon location or RFM in indoor environments is usually a challenging task. In this paper, we propose an efficient graph optimization-based method for estimating beacon location and RFM that combines range-based and fingerprinting methods. This method exists without the need for dedicated sensing equipment. A user wearing a BLE-enabled mobile device walks through an area and collects inertial readings and BLE received signal strength indication (RSSI) readings. Inertial measurements are processed through a pedestrian relative positioning (PDR) method to generate constraints on neighboring postures. Additionally, BLE fingerprints are employed to generate constraints between postures (with similar fingerprints), and RSSI generates distance constraints between postures and beacon locations (according to a predefined path loss model). These constraints are then employed to form a cost function with a least-squares structure. By minimizing this cost function, optimal user postures and beacon locat...

Claims

1. 1. A sensing system for detecting a vibroacoustic signal, comprising: A sensing device, 1. A vibroacoustic sensor module for detecting a vibroacoustic signal, the vibroacoustic sensor module comprising: a voice coil component including a coil holder supporting a winding; a magnet component comprising a magnet supported by a frame, the magnet having a magnet gap configured to receive at least a portion of the voice coil component in a spaced-apart, moveable manner; a connector connecting the voice coil component to the magnet component, the connector allowing relative movement between the voice coil component and the magnet component; a diaphragm configured to induce movement of the voice coil component within the magnet gap in response to incident acoustic waves; a vibration and acoustic sensor module comprising: a housing for holding the vibroacoustic sensor module, the sensing device being a handheld device, the housing having a handle end and a sensor end, the sensor end having a sensor end surface having an opening defined therethrough, the vibroacoustic sensor module being positioned in the housing such that at least a portion of the diaphragm of the vibroacoustic sensor module extends across the opening; A sensing system comprising a sensing device comprising:

2. 10. The sensing system of claim 1, further comprising a computer system processor in operative communication with the sensing device, the processor configured to execute a method for characterizing a physical state of a subject based at least in part on at least a portion of the detected vibroacoustic signal using a trained machine learning model.

3. The sensing system of claim 1 or claim 2, wherein the vibro-acoustic sensor module is configured to detect vibro-acoustic signals having a bandwidth in the range of 0.01 Hz to 160 kHz.

4. The detected vibroacoustic signal comprises: a first vibroacoustic signal component originating from the subject; and a second vibroacoustic signal component not originating from the subject; and Including, 3. The sensing system of claim 2, wherein the processor is configured to extract the first vibro-acoustic signal component from the detected vibro-acoustic signal and characterize the physical state of the subject based on the first vibro-acoustic signal component and the trained machine learning model.

5. 5. The sensing system of claim 4, wherein the processor is configured to extract a biological vibroacoustic signal from the first vibroacoustic signal component, and the detected vibroacoustic signal is detected through remote monitoring of the subject, through contact with the subject through clothing, or through direct skin contact with the subject.

6. The sensing device is (a) an electronics component contained within the housing and connected to the voice coil component configured to convert induced current in the winding from the sound wave to the detected vibroacoustic signal; (b) a power source contained within the housing and connected to one or both of the electronics component and the voice coil component; (c) a wireless communication module configured to communicate one or more of the detected vibroacoustic signal, the extracted first vibroacoustic signal component, and the body state characterization to a processor or another processor; The sensing system according to any one of claims 2 to 5, further comprising one or more of:

7. The sensing system of any one of claims 1 to 6, wherein the sensing device further comprises a context sensor module comprising an inertial measurement unit (IMU).

8. 8. The sensing system of claim 7, wherein the sensing device further comprises a bioelectric sensor module for detecting electrical signals from the subject, the bioelectric sensor module comprising at least one capacitive electrode at the sensor end surface and at least one drive right leg (DRL) electrode for providing a feedback circuit for the capacitive electrode.

9. The sensing system of any one of claims 1 to 8, wherein the voice coil component has an impedance of between 5 ohms and 170 ohms.

10. The sensing system of any one of claims 1 to 9, wherein the connector comprises a deflection structure extending radially from the voice coil component and having an opening formed therein.

11. 11. The sensing system of claim 1, wherein the connector comprises at least two bent arms extending radially from the voice coil component, the at least two bent arms spaced apart from each other to define at least one opening therebetween.

12. The sensing system of any one of claims 1 to 11, wherein the relative movement between the voice coil component and the magnet component is in the range of 0.1 mm / N to 5.0 mm / N.

13. The sensing system of any one of claims 1 to 12, wherein the sensing device further comprises an outer cover at the sensor end for covering an interface between the diaphragm and the housing.

14. 1. A method for detecting COVID-19 infection in a subject, the method being executed by a processor of a computer system, the method comprising: acquiring vibroacoustic data detected by the sensing system of claim 2, wherein the subject and the sensing device are positioned within a distance of 2 meters from each other; extracting a vibroacoustic signal component originating from the subject from the detected vibroacoustic signal; characterizing the subject as having COVID-19 infection based at least in part on the extracted vibroacoustic signal components using a machine learning model; and A method comprising:

15. 15. The method of claim 14, wherein the detected vibroacoustic data comprises a detected vibroacoustic signal having a bandwidth in the range of 0.01 Hz to 160 kHz, the detected vibroacoustic signal including a portion inaudible to the human ear.

Images