Magnetic sensing of muscle activity

Abstract

MMG sensors and apparatuses that may be used to detect neuromuscular signals corresponding to a subject, e.g., performing a variety of gestural tasks. MMG signals may be comparable to surface electromyography (sEMG) sensors and may perform better than sEMG and / or may serve as an alternative modality for applications currently dominated by sEMG.

Description

MAGNETIC SENSING OF MUSCLE ACTIVITYCLAIM OF PRIORITY

[0001] This patent application claims priority to U.S. provisional patent application no. 63 / 631,450, titled “MAGNETIC SENSING OF MUSCLE ACTIVITY,” and filed on April 8, 2024, herein incorporated by reference in its entirety.BACKGROUND

[0002] Magnetomyography, or MMG, is a technique for measuring muscle activity through the detection of magnetic fields. Despite its promise in application including humanmachine interfaces and medical applications, including both diagnostics and therapeutics, MMG is not well studied. Sensors required to detect biomagnetic signals are traditionally difficult to implement and use, causing MMG to be largely overlooked as a field of study.

[0003] In general, it would be highly desirable to provide MMG at scale. The methods and apparatuses described herein may address these needs.SUMMARY OF THE DISCLOSURE

[0004] Described herein a MMG sensors and apparatuses. These apparatuses may be used to detect signals corresponding to a subject performing a variety of gestural tasks. As described herein, MMG signals may be comparable to surface electromyography (sEMG) sensors and may perform better than sEMG and / or may serve as an alternative modality for applications currently dominated by sEMG. Additionally, data from MMG and sEMG (in addition to other sensors like accelerometers, gyroscopes, etc.) may be combined to improve system performance. Both MMG and sEMG sensing may have similar characteristics in terms of spectral content (i.e. the data from both modalities looks very similar when overlaid) and may provide direct measures of muscle contraction. MMG can be readily employed in applications requiring detection of degree of muscle contraction and small changes in signal properties, such as fine motor human-computer interfaces, rehabilitation, or athletics.

[0005] MMG signals are unique in ways that can be leveraged to unlock new insights. MMG signals show higher power than sEMG at higher frequencies (>100 Hz) and seem to be more sensitive to sensor placement relative to the source of the signal, indicating potential for greater information content and better localization. This result strongly suggests that MMG may provide better localization and could capture additional complex information compared to EMG. It also alludes to possible use cases of MMG in applications requiring highfrequency content, such as detection of muscle fatigue or assessment of motor unit action potentials.

[0006] In addition, MMG can operate at the same (or better) fidelity as sEMG without direct skin contact, making this modality more robust to real-world conditions. The drop-off in signal strength with distance from the source must be considered when designing MMG- based systems. Magnetic dipoles typically fall off as the cube of the distance from the source; MMG signal strength decreases more slowly than the inverse square law for reasonable recording distances.

[0007] Traditionally, sensor technology is a major limiting factor when it comes to scaling MMG. Commercially available sensors have various limitations, ranging from insufficient sensitivity to capture MMG signals at low contraction strengths to reduced frequency response at higher frequencies, prematurely cutting off the capturable bandwidth, and may require extensive magnetic shielding. MMG may be scaled by using a sensor that is sensitive enough but with high dynamic range to capture the full range of MMG signals in ambient conditions, while also being mass producible. Examples of such MMG systems may include ferromagnetic resonance (FMR) sensors that may be used to measure magnetic properties of materials by detecting the precessional motion of the magnetization in a ferromagnetic sample. Examples of FMR sensors that may be adapted for this use may include those described in: U.S. Patent No. 11,740,192, titled “SYSTEM AND METHOD FOR AN ACOUSTICALLY DRIVEN FERROMAGNETIC RESONANCE SENSOR DEVICE,” filed on December 14, 2020, U.S. Patent Application No. 18 / 222,543, titled “SYSTEM AND METHOD FOR AN ACOUSTICALLY DRIVEN FERROMAGNETIC RESONANCE SENSOR DEVICE,” field on July 17, 2023, U.S. Patent No. 11,903,715, titled “SYSTEM AND METHOD FOR A WEARABLE BIOLOGICAL FIELD SENSING DEVICE USING FERROMAGNETIC RESONANCE,” filed on January 28, 2021, U.S. patent application no. 18 / 413,091, titled, “SYSTEM AND METHOD FOR A WEARABLE BIOLOGICAL FIELD SENSING DEVICE USING FERROMAGNETIC RESONANCE,” filed on January 16, 2024, U.S. Patent No. 11,762,045, titled “SYSTEM AND METHOD FOR A MAGNETIC SENSOR ARRAY CIRCUIT,” filed on September 30, 2021, U.S. Patent Application No. 18,230,650, titled “SYSTEM AND METHOD FOR A MAGNETIC SENSOR ARRAY CIRCUIT,” filed on August 6, 2023; and International Patent Application No. PCT / US2024 / 020805, titled “SYSTEMS AND METHODS FOR MULTI-FERROIC TUNABLE ACOUSTICALLY DRIVEN MAGNETIC RESONANCE SENSORS,” filed on March 20, 2024. Each of these patents and pending application are herein incorporated by reference in their entirety.

[0008] Also described herein are neural control systems based on MMG. A modality like MMG may be remarkably powerful. For example, an MMG sensor may be used for a neural control system for gesture control for human-computer interfaces, prosthetics and rehabilitation devices. In general, these methods and apparatuses may use magnetic sensing as a modality for non-invasive HCI and BCI applications.

[0009] For example, described herein are magnetomyographic (MMG) sensor systems comprising: an array of acoustically driven ferromagnetic resonance (ADFMR) sensors, wherein each ADFMR sensor of the array of ADFMR sensors comprises a piezoelectric base, a magnetostrictive material on the piezoelectric base, and a pair of electrodes on either side of the magnetostrictive material, wherein each ADFMR sensor is configured to have a bandwidth of greater than 200 Hz and a dynamic range of 100 microtesla or greater, wherein the array of ADFMR sensors is coupled to a wearable garment configured to be worn in a ring around a user’s body part; and a processor configured to apply radiofrequency (RF) energy to each of the electrodes to generate an acoustic wave that resonates at a ferromagnetic resonance of the magnetostrictive material and to detect a MMG signal from the user’s body part from the array of ADFMR sensors.

[0010] The ADFMR sensors may be flexibly coupled to the wearable garment. As used herein, the flexible garment may broadly include any structure to be worn by the user, which may be rigid or flexible. The garment may be formed of a fabric and / or a frame. The garment may be configured as a gauntlet, brace, necklace, etc. The wearable garment comprises a rigid frame. In some cases, the wearable garment comprises a form-fitting material.

[0011] Each ADFMR sensor of the MMR system may be sized to fit in a relatively high density on the garment and therefore over the user’s body part. For example, the ADFMR sensors of the array of sensors may each be about 8 cm3or smaller, e.g., about 1 cm3or smaller, about 50 mm3or smaller, etc. The ADFMR sensors may be part of a chip that may include other components, including power supplies, power conditioners, filters, amplifiers, etc. In some cases, each sensor may be self-contained; in some cases, each sensor may be combined with other (e.g., orthogonal) sensors on the same sub-assembly (e.g., chip, etc.).

[0012] The MMG sensors may be configured to be separated from the skin of the user by at least 1 mm (e.g., between about 1 mm and 50 mm, between about 2 mm and 40 mm, between about 1 mm and 30mm, between about 1 mm and 20 mm, etc., greater than 2 mm, greater than about 3 mm, greater than about 4 mm, between 2-50 mm, between 2-40 mm, between 3-50 mm, between 3-40 mm, between 3-30 mm, etc.).

[0013] Any of these systems may include a separation layer (e.g., in some cases an insulating layer), between each ADFMR sensor and a patient-facing surface of the MMGsensor system. This separation layer may be configured to be separated from the skin of the user by at least 1 mm (e.g., at least 2 mm, 3 mm, 4 mm, 5 mm, etc.).

[0014] In general, the wearable garment may be configured to be worn over the user’s arm, hand, leg or neck. For example, the MMG system may be configured to detect a hand gesture(s) or movements, when worn over the subject’s hand, arm, forearm, etc.). In some cases, the MMG system may be configured to sense speech, including sub-vocalized (inaudible) speech. For example, the MMG system may be configured to be worn on or over the subject’s neck.

[0015] Any of these apparatuses (e.g., systems) may be configured to be worn in a fixed or relatively fixed position relative to the user’s body part or region; alternatively the apparatus / system may be configured to adapt to movement of the user’s body part, e.g., by applying movement-compensating logic that may adjust and account for movement of the body part relative to the individual ADFMR sensors in the array of ADFMR sensors, which may be performed by the processor.

[0016] The processor may be integrated with and / or separate from the wearable garment. In some cases, the processor may be a local processor; in some cases, the processor may be a cloud-based (remote) processor. In some cases, the processor may be configured to output an MMG corresponding to motor unit action potentials. The processor may be configured to output an MMG corresponding to muscle fatigue in the user’s body part.

[0017] Also described herein are methods of performing any of these apparatuses. For example, a method of detecting a magnetomyographic (MMG) signal from a user’s body part may include: receiving a plurality of MMG signals from an array of acoustically driven ferromagnetic resonance (ADFMR) sensors positioned over a user’s body part, wherein each ADFMR sensor of the array of ADFMR sensors comprises a piezoelectric base, a magnetostrictive material on the piezoelectric base, and a pair of electrodes on either side of the magnetostrictive material, and wherein each ADFMR sensor has a bandwidth of greater than 200 Hz and a dynamic range of 100 microtesla or greater; and identifying motor unit action potentials from the plurality of MMG signals.

[0018] Identifying the motor unit action potentials may include identifying a movement of a hand distal to the body part. In some cases, identifying the motor unit action potentials comprises identifying speech based on the motor unit action potentials. For example, receiving may comprise receiving the plurality of signals from the array of ADFMR sensors positioned between 1 and 30 mm from the surface of the body part. In some cases, receiving comprises receiving the plurality of signals from the array of ADFMR sensors positioned over the user’s arm. For example, receiving may comprise receiving the plurality of signalsfrom the array of ADFMR sensors positioned over the user’s neck. In any of these examples identifying the motor unit action potentials may comprise identifying muscle fatigue based on the motor unit action potentials. Any of these methods may include positioning the array of ADFMR sensors over the body part.

[0019] In general, these methods may include applying radiofrequency (RF) energy to the electrodes of the array of ADFMR sensors to generate acoustic waves in each of the ADFMR sensors that resonates at a ferromagnetic resonance of the magnetostrictive material of each ADFMR sensor and to detect the MMG signal from the user’s body part.

[0020] All of the methods and apparatuses described herein, in any combination, are herein contemplated and can be used to achieve the benefits as described herein.BRIEF DESCRIPTION OF THE DRAWINGS

[0021] A better understanding of the features and advantages of the methods and apparatuses described herein will be obtained by reference to the following detailed description that sets forth illustrative embodiments, and the accompanying drawings of which:

[0022] FIGS. 1A-1B illustrate one example of an experimental setup. FIG. IB shows 0PM (optically pumped magnetometer) and EMG sensor positions for the gradient force task (left); only the radial axis was recorded for each 0PM sensor. Sensor positions for clench and hold task (middle) with labeled sensors roughly corresponding to flexor carpi radialis (FCR), flexor carpi ulnaris (FCU), extensor digitorum (ED), and brachioradialis (BR). Both radial and tangential axes were recorded for each 0PM sensor. Green dots for both denote the center of EMG pairs. Finally, we also delivered stimulation to activate the median nerve using bipolar EMG electrodes (right). Red dots denote locations of each EMG electrode.

[0023] FIG. 1C shows an alignment of trials using measured force. In FIG. 1C, subjects had changing reaction times to the tone to start each trial that widely varied over time and across subjects (left). Aligning each trial to the maximum contraction strength shows properly aligned trials allowing consistent data analysis (right).

[0024] In general, the apparatuses and methods described herein may apply to, and may refer to, the use of one or more (e.g., an array of) ADFMR sensors as a key part of these systems. ADFMR systems may generally include a piezoelectric substrate holding or supporting a magnetostrictive material and two or more electrodes between which an RF energy at or near the resonant frequency of the magnetostrictive material is applied. Monitoring, comparing, and / or subtracting the signal(s) from the ADFMR sensors may be used to provide one or more signals indicative of motor unit action potentials from theplurality of MMG signals sensed by the system. Although other examples described herein may use other magnetostrictive materials, and / or other non-ADFRM magnetic sensors (such as optically pumped magnetometers, OPMs), as described in greater detail below, any of the methods or apparatuses described herein may include and / or may be incorporated with one or more of the alternative modes described herein.

[0025] FIGS. 2A-2D illustrate gradient force of contraction and changes in EMG and MMG signal strength. FIG. 2 A shows total EMG power between 20 and 300 Hz normalized to maximum power averaged across all trials as well as averaged measured force (left). As there was variability in timing of force between trials, we normalized the time from 10% to 100% of MVC to be uniform duration across trials. Pearson r correlation coefficients are shown for each channel compare with the measured force. The spectrogram of the EMG signals normalized for each frequency for channel EMG1 is shown on the right (FIG. 2C). FIG. 2B shows equivalent analysis applied to MMG signals, 0PM1 is shown on the right (FIG. 2D).

[0026] FIG 2E shows a minimum contraction strength as a percentage of maximum contraction detected by EMG and MMG sensors. The minimum contraction strength a sensor can detect was calculated by the degree of force required to reach the baseline amplitude plus two times the standard deviation of the baseline. Note that MMG was as sensitive to low contraction forces as EMG but seemed to also be more reliant on the specific positioning of the sensor.

[0027] FIGS. 3A-3C show frequency analysis of EMG and MMG during the clench and hold task. In FIG. 3 A, the example grip strength, single channel EMG spectrogram, and single channel MMG spectrogram averaged across an entire session. FIG. 3B shows normalized signal to noise ratio (SNR) of spectra during muscle contraction compared to during rest for each EMG channel and corresponding MMG channel. FIG. 3C shows differences in normalized SNR between MMG and EMG (MMG SNR - EMG SNR). The horizontal dashed line is at 0, or when MMG and EMG have equal SNR. The colored dots show where MMG SNR is greater than (blue) or less than (red) EMG SNR (Wilcoxon signed rank test). Note the cases in which MMG SNR is less than EMG SNR at higher frequencies typically fall within 60 Hz harmonics as the 0PM sensors used had greater line noise or at frequencies higher than 200 Hz likely due to the frequency response of the 0PM sensors (Supplementary FIG. 3).

[0028] FIG. 3D shows a frequency response of an 0PM sensor (Sensor A). In FIG. 3D, the frequency response was measured by playing a magnetic tone at various frequencies(dots) at the same amplitude. The curve across the bandwidth was obtained using cubic interpolation.

[0029] FIGS. 4A-4B shows a frequency analysis with MMG and EMG corrected with sensor frequency responses. In FIG. 4A, the analysis analogous to FIGS. 3A-3C. In FIG. 4A, normalized SNR of EMG and MMG and b, the difference between MMG SNR and EMG SNR. The MMG SNR was corrected by using the frequency response of the 0PM sensors. The EMG SNR was corrected to match an electrode with 3 mm diameter using an estimated frequency response.

[0030] FIGS. 4C-4D illustrates changes in EMG and MMG due to muscle fatigue. FIG. 4C shows an example of a normalized spectrogram of EMG positioned over the brachioradialis averaged for within trials (top left) and across trials (top right) for a single session. Median (middle) and mean (bottom) frequencies within and across trials. Shaded region shows standard error of the mean. FIG. 4D shows the same analyses applied to radial MMG signals positioned at the same location.

[0031] FIG. 5 A illustrates fatigue per subject. In FIG. 5 A, channels that showed signs of fatigue per subject both within and across trials, see FIG. 1 for sensor placements.Significance was determined by comparing mean and median frequencies of the first 1 second or 10 trials to the last 1 second or 10 trials to test for changes both within and across trials. Fatigue was only noted if the differences were statistically significant (Wilcoxon ranksum test) for both comparisons. Note both EMG and MMG (OPMs) show similar locations of fatigue per subject.

[0032] FIGS. 5B-5D illustrate MMG signal strength with respect to distance. In FIG. 5B, the example shows filtered time domain signals from 3 0PM sensors at different distances during a single trial of clench and hold. FIG. 5C shows an averaged spectra for each sensor. FIG. 5D shows an example of a rate of signal decrease with respect to distance (N) at hypothetical source depths. Shaded region shows standard error of the mean.

[0033] FIG. 6 illustrates an example of stimulus evoked potentials. FIG. 6 shows stimulus evoked potentials for three different sensors. The large signal around time 0 is the stimulation artifact and the signals from 5-15 ms from stimulation onset are the evoked responses. Note the similarity between EMG and MMG sensor and the lack of sharp features in the 0PM signals.

[0034] FIGS. 7A-7B illustrate voluntary MMG with TMR sensors. In FIG. 7 A, an SNR of MMG captured with MMG sensors and the ICA component with maximum SNR. FIG. 7B shows an MMG from voluntary movement captured with a commercial TMR sensor (Sensor C). Averaged spectra during contraction (green), control sensor for movement artifact duringcontraction (orange), and during relax periods (blue) (left). Spectra during contract subtracted by spectra of the control signal (right). Blue dots show statistical significance above zero (Wilcoxon rank sum test). FIG. 7B shows a spectra of forearm MMG at 20-100% MVC when measured at the skin compared to the noise floors of the commercially available sensors examined. All spectra were estimated from measured values. Note the MMG power was calculated using signals captured by the OPMs with a correction on the higher frequency components to account for the sensor frequency response (Supplementary FIG. 3).

[0035] FIG. 7C shows an example of positioning of MMG sensors (e.g., Sensor C). In FIG. 7C, the placement of the MMG sensors (black bars) on the forearm with sensitive axes (arrow) labeled for each sensor. The control sensor picks up minimal MMG as the sensitive axis is along the length of the muscle fibers.DETAILED DESCRIPTION

[0036] The measurement of magnetic fields generated by skeletal muscle, magnetomyography (MMG), has seen a renewal of interest from the academic community. Although studies have demonstrated complex models of MMG and experiments classifying between different movements using MMG, there has yet to be time frequency analysis of MMG as well as concurrent recordings of MMG and its electrical counterpart, electromyography (EMG). A comparison of MMG in the context of EMG is described herein, simultaneously recording both modalities during various muscle contraction tasks. MMG, similar to EMG, shows highly linearly correlated power to the degree of muscle contraction, has a unimodal distribution in spectral power, and can detect changes in muscle fatigue via changes in the spectral distribution. MMG typically has more high frequency content compared to EMG, even when accounting for the filtering induced by the size of the EMG electrodes. The decrease in MMG power due to distance from the arm and show MMG decreases slower than the inverse square law and can be measured up to 50 mm from the surface of the skin. Finally, MMG may be captured with non-OPM sensors.

[0037] Magnetic fields produced by the human body, or biomagnetism, was first demonstrated with recordings of the heart, or magnetocardiography (MCG) by Baule and McFee using a superconducting quantum inference device (SQUID). The first magnetic recordings of the brain and muscle, magnetoencephalography (MEG) and magnetomyography (MMG) respectively, were subsequently demonstrated. Both MCG and MEG were continuously developed over the decades and are used today, albeit seldomly. However, MMG has been understudied because existing magnetometers with sufficientsensitivities are difficult to implement, making its electrical counterpart, surface electromyography (sEMG), the dominant technology for most applications.

[0038] The use of miniaturized, optically pumped magnetometers (OPMs), high performance tunneling magnetoresistance (TMR) sensors, and novel sensor technologies like acoustically driven ferromagnetic resonance (ADFMR) sensors may be included in any of the methods and apparatuses (e.g., devices, systems, etc.). These sensors may have sensitivities high enough to capture MMG while being mobile and modular, providing numerous benefits for assessing MMG compared to SQUIDs.

[0039] Robust recordings of MMG have been shown, and used to apply stimulation, including stimulus triggered averages of muscle activity, decoding of individual motor units, and even discrimination between movements that rival results obtained with surface EMG. Models and simulations have also shown that MMG has higher specificity and better localization compared to surface EMG, furthering excitement for the field. Prior work typically reports filtered time traces or stimulus triggered averages of MMG and have even compared the accuracy of gesture discrimination between MMG and EMG, however, there has yet to be a comprehensive time-frequency characterization of voluntary MMG or an empirical comparison between MMG and EMG. Though the origin of both modalities is the currents traveling through muscle fibers, there are notable differences between how they propagate. Surface EMG signals are warped by the conductive properties of the signal path - tissue, skin, and the impedance at the sensor interface - whereas the body is permeable to magnetic fields leaving MMG signals unaffected. This distortion introduced to EMG is highly nonlinear and depends heavily on the subject and day-to-day environmental conditions, introducing difficulties in generalization and scaling with EMG. As such, we need a better understanding of not only the characteristics of MMG but also its differences from EMG.

[0040] To that end, described herein are simultaneously captured MMG and bipolar surface EMG. This study is the first to show characterization of MMG compared to EMG, including time frequency analysis with respect to muscle contraction and muscle fatigue, how MMG falls off as a function of distance from the signal source, and voluntary MMG captured with various sensors. These results provide context and groundwork for future MMG studies and demonstrate MMG recordings with different sensors to highlight the applicability of MMG.Examples

[0041] Methods - Data collection and Sensors and acquisition

[0042] Three different commercial magnetometers were used for measurements: sensors A, B, and C (each MMG sensors). Sensor A was an 0PM, Sensor B was a magnetoimpedance (MI) sensor, and sensor C was a TMR sensor. The sensors were strapped to subjects using a combination of custom 3D printed holders and Velcro straps.

[0043] All EMG recordings were performed with single-use, bipolar, MRI safe, pregelled surface EMG electrodes with non-ferrous contacts (EL508, Biopac Systems Inc.). The gelled area was roughly 10 mm in diameter, and the center-to-center inter-electrode spacing of bipolar pairs was kept to 20 mm. The EMG leads were also non-ferrous to limit introduction of artifactual magnetic fields (LEAD108C, Biopac Systems Inc.). Comparisons with standard EMG electrodes and leads showed no differences in recorded data. To measure the grip strength of subjects we used a pneumatic dynamometer bulb attached to an analog pressure sensor (TR1-0030A-101, Merit Sensor). To measure force gradient produced by finger flexion we used a FlexiForce A201 (Tekscan) and custom readout electronics.

[0044] All sensors provided analog signals which were digitized using NIDAQ ADC modules (National Instruments, 16-bit accuracy for OPMs, 24-bit accuracy for all other sensors) and collected with custom Python data acquisition code.

[0045] Measurements were performed with a magnetically shielded room; a MuROOM (Magnetic Shield Corporation) with inside dimensions of 1.3x1.3x2 meters was used to prevent sensor saturation and contamination with ambient noise sources (FIG. 1 A). The shielded room had around 25,000-fold attenuation of residual fields at DC and up to 8000 fold attenuation at AC. The room was degaussed prior to each recording to guarantee maximum rejection of ambient fields. All recordings were performed inside the shielded room.Participants and tasks

[0046] We had a total of 5 subjects participate in the study (4 males and 1 female in their 20’s and 30’s, all right-handed).

[0047] The study employed three different tasks. The first was a gradient force task in which the subject rested the ventral forearm flush on an acrylic table and slowly increased downward pressure onto the pressure sensor using their middle finger up to 100% maximum voluntary contraction (MVC) across 5 seconds before slowly ramping down back to relax. This task was cued with an audio tone. 2 subjects participated in this task.

[0048] The second involved the clench and hold task in which the subject was instructed to clench their fist at 100% maximum voluntary contraction (MVC) and hold for 10 seconds.Each trial was separated by 5 seconds of rest in which the subjects fully relaxed their arm. The start of the contraction and relaxation periods were cued by an audio tone, and subjects continued the task until they could not hold the contraction for the full 10 seconds. 5 subjects participated in this task when using a wristband of 16 dual-axis OPMs and four bipolar EMG channels (FIG. IB). 2 subjects participated in this task when using 0PM sensors at different distances, e.g., MMG sensors.

[0049] Lastly, we used the same EMG electrodes and leads as recordings to deliver stimulation to the median nerve on one subject. A pair of electrodes were placed along the length of the arm on the midline of the ventral side of the wrist with 20 mm inter-electrode distance to activate the median nerve (Pease et al., 2007). A Digitimer DS7A was used to deliver stimulation and was triggered by an Arduino microcontroller. The stimulus was monophasic and single pulse with 500 ps pulse width and alternating polarity each stimulus. The amplitude was adjusted in 1 mA increments until we observed roughly 80% of maximum recruitment as measured by EMG response amplitudes (typically ~30 mA). We then delivered the stimulation every 2 seconds with a jitter of 500 ms for 5 minutes.Sensor placements

[0050] For the clench and hold task 8 0PM sensors and 4 bipolar surface EMG sensor pairs were placed around the dominant forearm. To capture the largest muscle signals, we placed the sensors at 1 / 3 of the full forearm length from the antecubital fossa (crook of the elbow). The first 0PM sensor was placed roughly over the flexor carpi radialis, and each subsequent sensor was placed 45 degrees clockwise (FIG. IB). The first EMG sensor was also placed over the flexor carpi radialis, and each subsequent sensor was placed 90 degrees clockwise. The center of the bipolar EMG electrodes was positioned to be directly below the corresponding 0PM sensors.

[0051] For the gradient force task 6 0PM sensors were placed in a 3 *2 grid on the other side of the acrylic surface in a custom 3D printed mount (~2.5 cm distance from the surface of the skin). For measurements with MMR sensors as well as when using multiple 0PM sensors at different distances the sensors were placed above the extensor digitorum and held by a custom 3D printed holder.

[0052] FIG. 1 A shows a magnetically shielded room used for all experiments. FIG. IB schematically illustrates 0PM and EMG sensor positions for the gradient force task (left); only the radial axis was recorded for each 0PM sensor. Sensor positions for clench and hold task (middle) with labeled sensors roughly corresponding to flexor carpi radialis (FCR), flexor carpi ulnaris (FCU), extensor digitorum (ED), and brachioradialis (BR). Both radial and tangential axes were recorded for each 0PM sensor. Green dots for both denote thecenter of EMG pairs. Finally, we also delivered stimulation to activate the median nerve using bipolar EMG electrodes (right). Red dots denote locations of each EMG electrode. Data analysis and Preprocessing

[0053] All data were captured at a 2 kHz sampling rate. Signals were first filtered at 60 Hz harmonics with a 2ndorder HR notch filter with a q-factor of 30 to remove line noise. For analysis requiring further processing, we applied a band pass filter from 20 to 300 Hz with a 5thorder Butterworth filter. To account for trial-to-trial variability, we aligned each trial to the time of maximum grip strength as subjects typically started with the strongest grip that loosened with time as muscles were fatigued (FIG. 1C).Median and mean frequencies

[0054] The median frequency is defined as the frequency at which the sum of the spectrum is equal on both sides of the frequency:

[0055] where fminand fmaxare the minimum and maximum frequencies of the bandwidth and 5( ) is the spectrum of the signal. The mean frequency is defined as the average of frequencies weighted by the spectrum:

[0056] Both median and mean frequencies were calculated with a bandwidth of 20-300 Hz.Calculation of distance dependence

[0057] The decrease in signal strength with respect to distance between the sensor and the signal source can be written as:

[0058] Where d0and A are signal amplitudes at the source and sensor respectively, r0is the distance from the source to the surface of the skin, r, is the distance from the surface of the skin to the sensor, and N is the rate of decrease in the signal.

[0059] However, the signal amplitude at the source and the distance between the source and the surface of the skin are both unknown. As such, there needs to be a second sensor at a different distance, which gives:Dividing Equation 3 by Equation 4 and solving for N gives:

[0060] Equation 5 provides a method with which to calculate N as a function of r0. With three sensors at three distances Equation 5 can be modified to directly calculate N:

[0061] Note the subtraction of in Equation 6 rather than addition as in Equation 5 as the distance between the sensors must be calculated.Statistical analysis

[0062] All statistical analyses assumed non-parametric distributions of data and typically consisted of the one- or two-sided Wilcoxon signed rank (e.g., scipy. stats, wilcoxon) or rank sum (e.g., scipy.stats. ranksums) tests. Linear correlations were calculated using the Pearson correlation coefficient (e.g., scipy. stats. pearsonr). Statistical significance was determined at an alpha level of 0.05. Each analysis specifies the tests used to obtain the results.Results

[0063] MMG magnitude is correlated with force of muscle contraction. The amplitude of EMG has been shown to be directly correlated with the strength of muscle contraction, allowing it to be employed in applications in which varying contraction is a significant factor. To determine whether MMG also has the same correlation, we asked the subjects to press down with a finger with increasing strength over time and measured the flexor muscles on the forearm.

[0064] Pearson correlation coefficients were calculated between the measured force and EMG signals and found highly linearly correlated changes in surface EMG signal power with increase in force produced by finger flexion (FIGS. 2A-2C), similar to previous literature. We found the correlation to be consistent across the bandwidth with the Pearson correlation having a standard deviation of 0.02 (FIG. 2C). Though there are reports of frequencies being unchanged or even having negative correlations with contraction strength, these results are often subject-specific, indicating they are due to variability in volume conductor properties and muscle composition.

[0065] Repeating the analyses on MMG showed similar results (FIG. 2B). The correlation in MMG was similarly consistent across the bandwidth with the Pearson correlation also having a standard deviation of 0.02 (FIG. 2D).

[0066] An additional result of note is the variability of MMG. FIG. 2B left, shows normalized power, but it is clear there are specific channels which show high responses to muscle contraction compared to others (i.e. high variability at low contraction strength suggests lower signal -to-noise ratio). EMG channels that were similarly distanced all showed very similar signals across all channels. This potentially suggests higher spatial specificity of MMG, though additional investigation is necessary to draw firm conclusions.

[0067] Finally, we also assessed what percentage of maximum voluntary contraction (MVC) could be detected with either modality by capturing the average and standard deviation of the signal at rest. Setting a threshold to the average plus two times the standard deviation, we found MMG to be able to detect contraction strengths as small as EMG (FIG. 2E). We also found that the variance between the MMG channels was due to the placement of the sensors, with channels in high proximity having more similar detection thresholds.

[0068] In FIG. 2 A, the total EMG power between 20 and 300 Hz normalized to maximum power averaged across all trials as well as averaged measured force (left). As there was variability in timing of force between trials, we normalized the time from 10% to 100% of MVC to be uniform duration across trials. Pearson r correlation coefficients are shown for each channel compare with the measured force. The spectrogram of the EMG signals normalized for each frequency for channel EMG1 is shown in FIGS. 2C and 2D, respectively.MMG has large bandwidth with relatively higher power at higher frequencies.

[0069] After confirming that MMG power is correlated with the strength of contraction, we sought to compare the spectral distribution between MMG and EMG. MMG frequency content was reported during its initial discovery, but the recording device was roughly 4 cm from the surface of the skin and the measurements were of the upper arm and the palm. Toexpand upon these results, we recorded MMG and surface EMG from around the forearm, as gesture recognition from a wearable placed on the wrist or the forearm is a major emerging application of neuromuscular recordings. The subjects were instructed to clench their fists at 100% MVC and sustain the muscle contraction for 10 seconds with 5 seconds of relaxation periods in between each trial.

[0070] FIG. 3 A shows the averaged force across a single session of the clench and hold task and the averaged spectrogram of EMG and MMG measured at the extensor digitorum for a single subject. The timing of the two signals aligns well with the exerted force, and both modalities show significant power up to 100s of Hz. We were able to verify MMG signal power at 300 Hz even though the frequency response of the 0PM sensors we used start decreasing at 100 Hz and roughly -3.5 dB at 300 Hz (FIG. 3D), signifying MMG may have strong high frequency content.

[0071] To compare between the two modalities, we calculated the signal-to-noise ratio (SNR) for each trial by capturing the spectra during contraction and dividing by the average spectra of when the subject is at rest. The SNR for each trial was subsequently normalized to emphasize relative power between frequencies by dividing by the average power within a 20- 300 Hz bandwidth, then averaged across all sessions and subjects (FIG. 3B). Note that both axes of the 0PM sensors (radial and tangential) were used.

[0072] We found that MMG typically had higher relative power in the higher frequencies compared to EMG on all directly comparable sensor pairs. By calculating the difference between the normalized SNR of MMG and EMG, MMG has significantly greater relative power at frequencies greater than 100 Hz and up to 300 Hz even without accounting for the frequency response of the 0PM sensors (FIG. 3C). This suggests MMG may provide more reliable recordings of motor unit action potentials as well as more precise localization of active muscle fibers compared to EMG, though we were unable to test these hypotheses due to the lack of an applicable sensor.

[0073] To have the comparison be as equivalent as possible, we also corrected the MMG spectra with the 0PM frequency response. The EMG signals are also low pass filtered, mainly due to the size of the electrodes. As a result, we estimated the change in frequency response by using previously reported spatial frequency response and assuming a conduction velocity of 4 m / s to match that of a 3 mm wide electrode as the 0PM vapor cell is a 3 mm wide cube. The resulting normalized SNR shows a slight reduction in the difference at high frequencies and pushes the difference up to roughly 150 Hz for FCR and FCU but maintains that MMG has significantly greater relative power at higher frequencies (Supplementary FIG. 4).

[0074] FIGS. 3A-3C shows an example grip strength (FIG. 3A), single channel EMG spectrogram (FIG. 3B), and single channel MMG spectrogram (FIG. 3C) averaged across an entire session. FIG. 3B shows a normalized signal to noise ratio (SNR) of spectra during muscle contraction compared to during rest for each EMG channel and corresponding MMG channel. FIG. 3C shows difference in normalized SNR between MMG and EMG (MMG SNR - EMG SNR). The horizontal dashed line is at 0, or when MMG and EMG have equal SNR. The colored dots show where MMG SNR is greater than (blue) or less than (red) EMG SNR (Wilcoxon signed rank test). Note that the cases in which MMG SNR is less than EMG SNR at higher frequencies typically fall within 60 Hz harmonics as the 0PM sensors used had greater line noise or at frequencies higher than 200 Hz likely due to the frequency response of the 0PM sensors (Supplementary FIG. 3). See FIGS. 4A-4B for MMG SNR corrected to account for the frequency response.MMG can detect muscle fatigue

[0075] We assessed muscle fatigue by evaluating the temporal changes in MMG signal in the clench and hold task. The task was designed such that it continued until the subjects could not complete a full trial reflecting muscle fatigue. Though numerous methods of tracking muscle fatigue with EMG signals exist, we focused on the decrease in average frequency (i.e. relative decrease in high frequency power) due to its spectral relevance.

[0076] To that end, we calculated the spectrogram of the signals for each channel in each trial. To determine changes in spectral content within trials (i.e. throughout the duration of holding the contraction) and across trials (i.e. throughout the duration of the session) we normalized the spectrograms either across time within trials or across trials (FIG. 4C and 4D, top).

[0077] From the spectrograms it was clear that most frequencies decreased over time, but different frequencies behaved differently depending on the degree of fatigue (i.e. time within trials and number of trials within a session). As a result, we calculated the median and mean frequencies for both EMG and MMG signals both within and across trials (FIGS. 4C and 4D, middle and bottom). There is a clear decrease in both measures both within and across trials showing that high frequency signals were decreasing faster, similar to previously reported results.

[0078] Both median and mean frequencies reflected the same results, demonstrating lack of outlier influence. Interestingly, the measured decreased linearly within trials but across trials had a large decrease within the first 10 trials before stabilizing or even going up in the later trials as in the case of EMG in the example. We also found that MMG typically had larger change in the measures, as demonstrated in FIGS. 4C-4D.

[0079] In FIG. 4C, example normalized spectrogram of EMG positioned over the brachioradialis averaged for within trials (top left) and across trials (top right) for a single session. Median (middle) and mean (bottom) frequencies within and across trials. Shaded region shows standard error of the mean. FIG. 4D shows the same analyses applied to radial MMG signals positioned at the same location.

[0080] Not all channels of EMG and MMG displayed such changes. We assessed whether fatigue was detected by statistical comparison of the first second or first 10 trials compared with the last second or the last 10 trials. Fatigue was only noted if the difference was significant both within and across trials (FIG. 5 A). As each subject had slightly differing strategies of grasping the pressure bulb (position and angle of bulb in the hand, positions of the fingers around the bulb, etc.) as well as varying muscle structure, we observed subject dependent differences in which channels displayed fatigue. Note, however, that EMG and MMG channels located in proximity typically showed similar changes.MMG signal strength as a function of distance typically decreases slower than the inverse square law.

[0081] The magnetic field of a dipole falls off as a cube of the distance from the source, the field from a finite wire as the square of the distance, and the field from an infinite wire as the inverse of the distance. Although muscle fibers are often modeled as finite wires, the drop-off is typically considered to be the cube of the distance when recording very close to the source to become the inverse of the distance when further from the source.

[0082] To empirically quantify the rate at which the signal decreases, we measured MMG signals over the extensor digitorum during a clench and release task for 20 trials across 2 subjects. MMG was captured with three 0PM sensors measuring radially with 10 mm between the sensors, translating to vapor cells (i.e. the sensitive element of the sensors) at distances of 6.2, 28.6, and 51 mm from the skin (FIGS. 5B and 5C). The 0PM sensors apply an internal bias field to cancel out external DC fields which can interact between sensors in close proximity resulting in sensor saturation, and greater than 10 mm distance between sensors often led to no signal captured in the furthest sensor. As a result, we could not take measurements at different distances.

[0083] FIG. 5B shows an example filtered time domain signals from 3 0PM sensors at different distances during a single trial of clench and hold. FIG. 5C shows an averaged spectra for each sensor. FIG. 5D shows a rate of signal decrease with respect to distance (N) at hypothetical source depths. Shaded region shows standard error of the mean.

[0084] Using measurements from the near and middle sensors, we can estimate the rate of decrease in the signal at the skin surface with respect to the depth of the signal source (FIG.5D). The change in signal is largely linear if the signal source is superficial (less than 2 mm) and the rate remains below 2 even if the signal source was up to 10 mm deep. Using all three sensors we calculated the rate of decrease between the middle and far sensors to be 0.90 ± 0.05 (average ± SEM), showing the decrease quickly becomes linear.MMG can be captured with commercial non-OPM magnetometers

[0085] Although SQUIDs and OPMs have typically been used to detect MMG due to the high sensitivity requirements, we tested whether we could detect MMG with a commercial magnetoimpedance (MI) sensor. Though previous studies have demonstrated MMG with non-OPM sensors, they were typically performed with custom hardware in research settings. To that end we measured the abductor pollicis brevis while delivering stimulation to the median nerve at the wrist.

[0086] Three sessions on one subject were captured, each with a different type of sensor. The triggered averages of each are shown in FIG. 6, showing stimulus evoked potentials for three different sensors. The large signal around time 0 is the stimulation artifact and the signals from 5-15 ms from stimulation onset are the evoked responses. Note the similarity between EMG and MI sensors and the lack of sharp features in the OPM signals. FIG. 6 shows that the responses captured with the MI sensor (e.g., Sensor B) very similar to the response measured with EMG (biphasic response from 5 - 15 ms). Signals recorded with OPMs resulted in a similar response but were not able to properly capture the sharp features. This may be due to the bandwidth limitations of the OPM (e.g., see FIG. 3D).Sensor noise floor required to capture voluntary MMG

[0087] We explored whether we could capture voluntary MMG with the commercial sensors (Sensor A, Sensor B and Sensor C) as well. To that end we used two sets of 4 MI sensors (Sensor B) aligned and stacked with 3 mm distance between each sensor. We then mounted the sensors on the extensor digitorum to attempt to capture similar signals across all sensors.

[0088] FIG. 7A shows the SNR the highest SNR MI sensor (Sensor B) during muscle contraction in a clench and hold task compared to periods of relaxation averaged across 50 trials. Applying ICA across all sensors resulted in a ~0.1 increase in SNR across the bandwidth. Though not a significant enhancement, it nevertheless suggests we could leverage high density arrays with lower sensitivity to increase SNR when measuring voluntary MMG.

[0089] We repeated the procedure with a second commercially available sensor (Sensor C, FIG. 7B). The Sensor C TMR sensor was particularly sensitive to movement, and it was especially challenging to securely fasten to the arm. Thus, we used another sensor with its sensitive axis down the length of the arm to purposefully pick up minimal MMG and used itsmeasurements to estimate movement artifacts (FIG. 7C). Averaging across 50 trials shows significant signals during contraction compared to the movement control up through 300 Hz.

[0090] These results suggest that the MI sensor noise floor (e.g., with Sensor B) is at the threshold of measuring MVC MMG (FIG. 7C). Though single TMR sensors (Sensor C) were able to reliably capture large MMG signals, they struggled to discern signals during small forces of contraction. Though OPMs were able to fully measure MMG, we could not determine the change in signal strength at frequencies higher than 300 Hz. Nevertheless, our comparisons show that a sensor with a noise floor and an order of magnitude higher than OPMs should be able to reliably capture MMG in this bandwidth.DiscussionMMG reflects muscle activity similar to EMG

[0091] Although magnetic signals recorded in the vicinity of skeletal muscle have been previously reported, there has yet to be validation that these signals directly reflect muscle activity and are directly analogous to EMG. Muscle fibers generate large currents, but the magnetic signal may be obfuscated due to the orientations of the fibers and the asynchronous timing of the motor units; they may also be affected by signals generated by nerve bundles or even blood flow. In addition, movement of the sensors, whether from muscle contraction, heart beats, or spontaneous tremors, could introduce artifacts if there is a residual magnetic field gradient present.

[0092] By measuring MMG simultaneously with EMG at the same area we found that MMG shows very similar characteristics in spectral content, having a single peak at roughly the same frequency as well as similar bandwidth. In addition, we observed MMG to be highly linearly correlated with induced force, demonstrating it to be a direct measure of muscle contraction. We also found MMG can detect muscle fatigue at least as consistently as EMG by comparing spectral content, demonstrating independent narrow band signals. MMG can thus be employed in applications requiring detection of degree of muscle contraction and small changes in signal properties, such as rehabilitation, athletics, or fine motor humancomputer interfaces.MMG has distinct differences from EMG

[0093] Though MMG displayed many similar properties to EMG, we noticed a large difference between the high signal power content. Spectral content in surface EMG is heavily dependent on both the electrode size and the inter-electrode (IED) distance, with smaller sizes and shorter distances allowing for detection of higher frequencies. As a result, although SENIAM (Surface ElectroMyoGraphy for the Non-Invasive Assessment of Muscles) recommended ~10 mm size electrodes and ~20 mm IED for bipolar surface EMG recordings,subsequent studies have suggested smaller electrodes with shorter lEDs to capture a larger bandwidth of the EMG signal.

[0094] We were limited to using the EMG electrodes and distancing used in the study as we required non-ferrous contacts to prevent the OPMs from saturation or contaminate other sensor signals with noise. However, changing the IED from 20 mm to 10 mm has shown the median frequency, or the frequency at which the total spectral power is split in half, to change only by roughly 5 Hz within a 500 Hz bandwidth. Another study measuring mean frequency, or the average frequency weighted by the spectral power, corroborates these results showing a change of ~3 Hz at IED of 18 mm compared to 36 mm. As such, the IED alone does not fully explain the difference in the higher frequency power observed between MMG and EMG.

[0095] The size of the electrodes potentially has a much larger effect on EMG bandwidth as the increased area results in integration resulting in attenuation of higher frequencies. However, previous studies have shown that for 10 mm wide devices the transfer function of spatial filtering results in -3 dB at 50 cycles / meter which translates to 200 Hz when assuming 4 meters / second conduction velocity of muscle fiber signals. Applying corrections based on these transfer functions as well as the 0PM sensor’s frequency response showed that MMG still maintained significantly larger high frequency power (FIGS. 4A-4B).

[0096] The difference in high frequency content is likely due to volume conduction affecting electrical signal transmission. Though MMG can be affected by passive currents traveling through tissue, the effect is minimal compared to the distortion caused by the changes in impedance at tissue boundaries. This result strongly suggests that MMG may provide better localization and could capture additional complex information compared to surface EMG. It also alludes to possible use cases of MMG in applications requiring high frequency content, such as detection of muscle fatigue or assessment of motor unit action potentials.

[0097] One other large difference between MMG and EMG is that MMG can operate with the same fidelity regardless of skin contact. However, as magnetic fields decrease in power with respect to distance, how much this advantage could be leveraged before the signals become too small to detect has not been fully quantified. Conflicting modeling results on whether the signal drops faster or slower with distance has caused further uncertainty. In this study we show conclusively that the signal from the surface of the skin drops slower than the inverse square law assuming the “center” of the signal source is less than 10 mm from the surface of the skin. We were still able to measure the MMG signal at greater than 50 mmfrom the surface of the skin, showing the non-contact aspect of MMG is indeed an advantage that can be utilized.

[0098] OPMs have been a major advancement in magnetic sensor development for biomagnetism. Compared to SQUIDs (superconducting quantum interference device), the original magnetometers used to record biomagnetism, they have comparative sensitivity while allowing for flexible positioning and high density, while being significantly lower in cost. However, compared to EMG sensors OPMs may have lower bandwidth, small dynamic range, and are significantly more expensive.

[0099] The MMG sensors for use as described herein may have a high bandwidth, large enough dynamic range to operate in unshielded conditions near electronic equipment and other magnetically noisy objects, low noise floor to detect the relatively small MMG signals, small size to allow for high density, and low price for ubiquity.

[0100] Magnetic sensing of the brain (magnetoencephalography (MEG)) has been demonstrated to provide higher localization accuracy compared to electrical sensing (electroencephalography (EEG)) in both a phantom as well as an implanted dipole in humans, but the results cannot be extended to muscle sensing. The head has complex conductivity due to the thickness of the skull and boundaries between various tissues; in addition, brain sources are typically assumed to behave like dipoles. In contrast, skeletal muscle is much closer to the surface of the skin and the fields produced by muscle fibers behave more similarly to finite wires. As such, there is a need to empirically compare localization performance between MMG and EMG before drawing conclusions on their efficacies.

[0101] OPMs used in this study may allow flexible and close placement of the sensors to the muscles but had clear limitations in bandwidth. A magnetic sensor with higher bandwidth would allow for detection of voluntarily activated motor unit action potentials, potentially proving further insights into applications of MMG.

[0102] All publications and patent applications mentioned in this specification are herein incorporated by reference in their entirety to the same extent as if each individual publication or patent application was specifically and individually indicated to be incorporated by reference. Furthermore, it should be appreciated that all combinations of the foregoing concepts and additional concepts discussed in greater detail below (provided such concepts are not mutually inconsistent) are contemplated as being part of the inventive subject matter disclosed herein and may be used to achieve the benefits described herein.

[0103] Any of the methods (including user interfaces) described herein may be implemented as software, hardware or firmware, and may be described as a non-transitory computer-readable storage medium storing a set of instructions capable of being executed bya processor (e.g., computer, tablet, smartphone, etc.), that when executed by the processor causes the processor to control perform any of the steps, including but not limited to: displaying, communicating with the user, analyzing, modifying parameters (including timing, frequency, intensity, etc.), determining, alerting, or the like. For example, any of the methods described herein may be performed, at least in part, by an apparatus including one or more processors having a memory storing a non-transitory computer-readable storage medium storing a set of instructions for the processes(s) of the method.

[0104] While various embodiments have been described and / or illustrated herein in the context of fully functional computing systems, one or more of these example embodiments may be distributed as a program product in a variety of forms, regardless of the particular type of computer-readable media used to actually carry out the distribution. The embodiments disclosed herein may also be implemented using software modules that perform certain tasks. These software modules may include script, batch, or other executable files that may be stored on a computer-readable storage medium or in a computing system. In some embodiments, these software modules may configure a computing system to perform one or more of the example embodiments disclosed herein.

[0105] As described herein, the computing devices and systems described and / or illustrated herein broadly represent any type or form of computing device or system capable of executing computer-readable instructions, such as those contained within the modules described herein. In their most basic configuration, these computing device(s) may each comprise at least one memory device and at least one physical processor.

[0106] The term “memory” or “memory device,” as used herein, generally represents any type or form of volatile or non-volatile storage device or medium capable of storing data and / or computer-readable instructions. In one example, a memory device may store, load, and / or maintain one or more of the modules described herein. Examples of memory devices comprise, without limitation, Random Access Memory (RAM), Read Only Memory (ROM), flash memory, Hard Disk Drives (HDDs), Solid-State Drives (SSDs), optical disk drives, caches, variations or combinations of one or more of the same, or any other suitable storage memory.

[0107] In addition, the term “processor” or “physical processor,” as used herein, generally refers to any type or form of hardware-implemented processing unit capable of interpreting and / or executing computer-readable instructions. In one example, a physical processor may access and / or modify one or more modules stored in the above-described memory device. Examples of physical processors comprise, without limitation, microprocessors, microcontrollers, Central Processing Units (CPUs), Field-ProgrammableGate Arrays (FPGAs) that implement softcore processors, Application-Specific Integrated Circuits (ASICs), portions of one or more of the same, variations or combinations of one or more of the same, or any other suitable physical processor.

[0108] Although illustrated as separate elements, the method steps described and / or illustrated herein may represent portions of a single application. In addition, in some embodiments one or more of these steps may represent or correspond to one or more software applications or programs that, when executed by a computing device, may cause the computing device to perform one or more tasks, such as the method step.

[0109] In addition, one or more of the devices described herein may transform data, physical devices, and / or representations of physical devices from one form to another. Additionally or alternatively, one or more of the modules recited herein may transform a processor, volatile memory, non-volatile memory, and / or any other portion of a physical computing device from one form of computing device to another form of computing device by executing on the computing device, storing data on the computing device, and / or otherwise interacting with the computing device.

[0110] The term “computer-readable medium,” as used herein, generally refers to any form of device, carrier, or medium capable of storing or carrying computer-readable instructions. Examples of computer-readable media comprise, without limitation, transmission-type media, such as carrier waves, and non-transitory-type media, such as magnetic-storage media (e.g., hard disk drives, tape drives, and floppy disks), optical-storage media (e.g., Compact Disks (CDs), Digital Video Disks (DVDs), and BLU-RAY disks), electronic-storage media (e.g., solid-state drives and flash media), and other distribution systems.[OHl] A person of ordinary skill in the art will recognize that any process or method disclosed herein can be modified in many ways. The process parameters and sequence of the steps described and / or illustrated herein are given by way of example only and can be varied as desired. For example, while the steps illustrated and / or described herein may be shown or discussed in a particular order, these steps do not necessarily need to be performed in the order illustrated or discussed.

[0112] The various exemplary methods described and / or illustrated herein may also omit one or more of the steps described or illustrated herein or comprise additional steps in addition to those disclosed. Further, a step of any method as disclosed herein can be combined with any one or more steps of any other method as disclosed herein.

[0113] The processor as described herein can be configured to perform one or more steps of any method disclosed herein. Alternatively or in combination, the processor can be configured to combine one or more steps of one or more methods as disclosed herein.

[0114] When a feature or element is herein referred to as being "on" another feature or element, it can be directly on the other feature or element or intervening features and / or elements may also be present. In contrast, when a feature or element is referred to as being "directly on" another feature or element, there are no intervening features or elements present. It will also be understood that, when a feature or element is referred to as being "connected", "attached" or "coupled" to another feature or element, it can be directly connected, attached or coupled to the other feature or element or intervening features or elements may be present. In contrast, when a feature or element is referred to as being "directly connected", "directly attached" or "directly coupled" to another feature or element, there are no intervening features or elements present. Although described or shown with respect to one embodiment, the features and elements so described or shown can apply to other embodiments. It will also be appreciated by those of skill in the art that references to a structure or feature that is disposed "adjacent" another feature may have portions that overlap or underlie the adjacent feature.

[0115] Terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. For example, as used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise. It will be further understood that the terms "comprises" and / or "comprising," when used in this specification, specify the presence of stated features, steps, operations, elements, and / or components, but do not preclude the presence or addition of one or more other features, steps, operations, elements, components, and / or groups thereof. As used herein, the term "and / or" includes any and all combinations of one or more of the associated listed items and may be abbreviated as " / ".

[0116] Spatially relative terms, such as "under", "below", "lower", "over", "upper" and the like, may be used herein for ease of description to describe one element or feature's relationship to another element(s) or feature(s) as illustrated in the figures. It will be understood that the spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. For example, if a device in the figures is inverted, elements described as "under”, or "beneath" other elements or features would then be oriented "over" the other elements or features. Thus, the exemplary term "under" can encompass both an orientation of over and under. The device may be otherwise oriented (rotated 90 degrees or at other orientations) and the spatially relative descriptors used herein interpreted accordingly. Similarly, the terms "upwardly","downwardly", "vertical", "horizontal" and the like are used herein for the purpose of explanation only unless specifically indicated otherwise.

[0117] Although the terms “first” and “second” may be used herein to describe various features / elements (including steps), these features / elements should not be limited by these terms, unless the context indicates otherwise. These terms may be used to distinguish one feature / element from another feature / element. Thus, a first feature / element discussed below could be termed a second feature / element, and similarly, a second feature / element discussed below could be termed a first feature / element without departing from the teachings of the present invention.

[0118] In general, any of the apparatuses and methods described herein should be understood to be inclusive, but all or a sub-set of the components and / or steps may alternatively be exclusive and may be expressed as “consisting of’ or alternatively “consisting essentially of’ the various components, steps, sub-components or sub-steps.

[0119] As used herein in the specification and claims, including as used in the examples and unless otherwise expressly specified, all numbers may be read as if prefaced by the word "about" or “approximately,” even if the term does not expressly appear. The phrase “about” or “approximately” may be used when describing magnitude and / or position to indicate that the value and / or position described is within a reasonable expected range of values and / or positions. For example, a numeric value may have a value that is + / - 0.1% of the stated value (or range of values), + / - 1% of the stated value (or range of values), + / - 2% of the stated value (or range of values), + / - 5% of the stated value (or range of values), + / - 10% of the stated value (or range of values), etc. Any numerical values given herein should also be understood to include about or approximately that value, unless the context indicates otherwise. For example, if the value " 10" is disclosed, then "about 10" is also disclosed. Any numerical range recited herein is intended to include all sub-ranges subsumed therein. It is also understood that when a value is disclosed that "less than or equal to" the value, "greater than or equal to the value" and possible ranges between values are also disclosed, as appropriately understood by the skilled artisan. For example, if the value "X" is disclosed the "less than or equal to X" as well as "greater than or equal to X" (e.g., where X is a numerical value) is also disclosed. It is also understood that the throughout the application, data is provided in a number of different formats, and that this data, represents endpoints and starting points, and ranges for any combination of the data points. For example, if a particular data point “10” and a particular data point “15” are disclosed, it is understood that greater than, greater than or equal to, less than, less than or equal to, and equal to 10 and 15 are considered disclosed as well as between 10 and 15. It is also understood that each unit between two particular unitsare also disclosed. For example, if 10 and 15 are disclosed, then 11, 12, 13, and 14 are also disclosed.

[0120] Although various illustrative embodiments are described above, any of a number of changes may be made to various embodiments without departing from the scope of the invention as described by the claims. Optional features of various device and system embodiments may be included in some embodiments and not in others. Therefore, the foregoing description is provided primarily for exemplary purposes and should not be interpreted to limit the scope of the invention as it is set forth in the claims.

[0121] The examples and illustrations included herein show, by way of illustration and not of limitation, specific embodiments in which the subject matter may be practiced. As mentioned, other embodiments may be utilized and derived there from, such that structural and logical substitutions and changes may be made without departing from the scope of this disclosure. Such embodiments of the inventive subject matter may be referred to herein individually or collectively by the term “invention” merely for convenience and without intending to voluntarily limit the scope of this application to any single invention or inventive concept, if more than one is, in fact, disclosed. Thus, although specific embodiments have been illustrated and described herein, any arrangement calculated to achieve the same purpose may be substituted for the specific embodiments shown. This disclosure is intended to cover any and all adaptations or variations of various embodiments. Combinations of the above embodiments, and other embodiments not specifically described herein, will be apparent to those of skill in the art upon reviewing the above description.

Claims

CLAIMSWhat is claimed is:

1. A magnetomyographic (MMG) sensor system, the system comprising: an array of acoustically driven ferromagnetic resonance (ADFMR) sensors, wherein each ADFMR sensor of the array of ADFMR sensors comprises a piezoelectric base, a magnetostrictive material on the piezoelectric base, and a pair of electrodes on either side of the magnetostrictive material, wherein each ADFMR sensor is configured to have a bandwidth of greater than 200 Hz and a dynamic range of 100 microtesla or greater, wherein the array of ADFMR sensors is coupled to a wearable garment configured to be worn in a ring around a user’s body part; and a processor configured to apply radiofrequency (RF) energy to each of the electrodes to generate an acoustic wave that resonates at a ferromagnetic resonance of the magnetostrictive material and to detect a MMG signal from the user’s body part from the array of ADFMR sensors.

2. The MMG sensor of claim 1, wherein the ADFMR sensors are flexibly coupled to the wearable garment.

3. The MMG sensor system of claim 1, wherein each ADFMR sensor 8 cm3or smaller.

4. The MMG sensor system of claim 1, wherein each ADFMR sensor is 1 cm3or smaller.

5. The MMG sensor system of claim 1, wherein each ADFMR sensor is 50 mm3or smaller.

6. The MMG sensors system of claim 1, wherein the MMG sensor is configured to be separated from the skin of the user by at least 1 mm.

7. The MMG sensors system of claim 1, further comprising an insulating layer between each ADFMR sensor and a patient-facing surface of the MMG sensor system, wherein the MMG sensor is configured to be separated from the skin of the user by at least 1 mm.

8. The MMG sensor of claim 1, wherein the wearable garment comprises a rigid frame.

9. The MMG sensor of claim 1, wherein the wearable garment comprises a form -fitting material.

10. The MMG sensor of claim 1, wherein the wearable garment is configured to be worn over the user’s arm, hand, leg or neck.

11. The MMG of claim 1, wherein the processor is configured to output an MMG corresponding to motor unit action potentials.

12. The MMG of claim 1, wherein the processor is configured to output an MMG corresponding to muscle fatigue in the user’s body part.

13. A method of detecting a magnetomyographic (MMG) signal from a user’s body part, the method comprising: receiving a plurality of MMG signals from an array of acoustically driven ferromagnetic resonance (ADFMR) sensors positioned over a user’s body part, wherein each ADFMR sensor of the array of ADFMR sensors comprises a piezoelectric base, a magnetostrictive material on the piezoelectric base, and a pair of electrodes on either side of the magnetostrictive material, and wherein each ADFMR sensor has a bandwidth of greater than 200 Hz and a dynamic range of 100 microtesla or greater; and identifying motor unit action potentials from the plurality of MMG signals.

14. The method of claim 13, wherein identifying the motor unit action potentials comprises identifying a movement of a hand distal to the body part.

15. The method of claim 13, wherein identifying the motor unit action potentials comprises identifying speech based on the motor unit action potentials.

16. The method of claim 13, wherein receiving comprises receiving the plurality of signals from the array of ADFMR sensors positioned between 1 and 30 mm from the surface of the body part.

17. The method of claim 13, wherein receiving comprises receiving the plurality of signals from the array of ADFMR sensors positioned over the user’s arm.

18. The method of claim 13, wherein receiving comprises receiving the plurality of signals from the array of ADFMR sensors positioned over the user’s neck.

19. The method of claim 13, wherein identifying the motor unit action potentials comprises identifying muscle fatigue based on the motor unit action potentials.

20. The method of claim 13, further comprising positioning the array of ADFMR sensors over the body part.

21. The method of claim 13, further comprising applying radiofrequency (RF) energy to the electrodes of the array off ADFMR sensors to generate acoustic waves in each of the ADFMR sensors that resonates at a ferromagnetic resonance of the magnetostrictive material of each ADFMR sensor and to detect the MMG signal from the user’s body part.

Images