<p>Physiological Vibration Acceleration (Phybrata) Sensor Assessment of Multi-System Physiological Impairments and Sensory Reweighting Following Concussion</p>

Phybrata Sensors and Machine Learning for Enhanced Neurophysiological Diagnosis and Treatment

“…We recently introduced physiological vibration acceleration (phybrata) sensing [58], a non-invasive balance and neurophysiological impairment assessment tool that utilizes a head-mounted accelerometer to detect the contributions of individual physiological systems to a phenomenon that is unique to a head-mounted sensor-the biomechanical stabilization of the head and eyes as the reference platform that the body relies on for balance and movement. The phybrata sensor attaches behind the ear using a small disposable adhesive and testing simply requires the patient to stand still for 20 s with their eyes open (Eo), and again for 20 s with their eyes closed (Ec).…”

“…During this static balance testing, the phybrata sensor uses a microelectromechanical system (MEMS)-based inertial motion unit (IMU) to detect microscopic involuntary motions of the body, both normal motions characteristic of healthy individuals and pathological motions caused by physiological impairments that impact balance and postural stability. As we previously described in detail [58], the direct measurement of acceleration motions at the head allows the phybrata sensor to detect signal components with a much wider range of frequencies compared to motion sensors mounted elsewhere on the body, which in turn allows impairments to individual physiological systems (CNS, PNS, sensory, musculoskeletal) to be identified, quantified, and monitored using the unique biomechanical vibrational signature of each system, and reveals the sensory reweighting across multiple physiological systems that is triggered by impairment to any single physiological system. Although many wearable motion sensors contain both an accelerometer and a gyroscope, accelerometers are significantly more power-efficient than gyroscopes [59], and the use of a single accelerometer in the phybrata sensor makes it more practical for long-term activity monitoring [60,61].…”

“…The above impairment classification leverages previous spectral analyses of postural sway time series data in which distinct frequency bands are observed to correspond to specific mechanisms of postural control [64][65][66][67][68] for example >1 Hz (spinal reflexive loops, proprioception, multi-joint and muscle activity); 0.5-1 Hz (CNS participation, both cerebellar and cortical); 0.1-0.5 Hz (vestibular regulation); 0.02-0.1 Hz (visual regulation). The much wider spectral content of phybrata signals compared to motion sensors mounted elsewhere on the body also reveals the sensory reweighting across multiple physiological systems that is triggered by impairment to any single physiological system, as the postural control system dynamically regulates to reduce dependence upon impaired inputs [58]. Disruptions to these neuromotor and neurosensory systems accompany many neurological conditions, including concussions [58,69,70], and are closely tied to concussion symptoms [7,15].…”

“…The much wider spectral content of phybrata signals compared to motion sensors mounted elsewhere on the body also reveals the sensory reweighting across multiple physiological systems that is triggered by impairment to any single physiological system, as the postural control system dynamically regulates to reduce dependence upon impaired inputs [58]. Disruptions to these neuromotor and neurosensory systems accompany many neurological conditions, including concussions [58,69,70], and are closely tied to concussion symptoms [7,15]. Digital biomarkers derived from unique spatial-domain, time-domain, and frequency-domain features and sensory reweighting profiles detected in phybrata signals were shown to demonstrate well-defined normative values that distinguish between healthy and impaired individuals across a wide population base with no previous baseline measurement required [69].…”

“…Digital biomarkers derived from unique spatial-domain, time-domain, and frequency-domain features and sensory reweighting profiles detected in phybrata signals were shown to demonstrate well-defined normative values that distinguish between healthy and impaired individuals across a wide population base with no previous baseline measurement required [69]. For the specific case of concussion injuries, receiver operating characteristic (ROC) analyses of phybrata digital biomarkers have demonstrated classification of healthy vs. concussed individuals with sensitivity, specificity, accuracy, and area under the curve (AUC) all above 90%, as well as separate independent classification of neurological and vestibular impairments, also with sensitivity, specificity, accuracy, and AUC all above 90% [58].…”