Publications

As digital health technology becomes more pervasive, machine learning (ML) provides a robust way to analyze and interpret the myriad of collected features. The purpose of this preliminary work was to use ML classification to assess the benefits and relevance of neurocognitive features both tablet-based assessments and self-reported metrics, as they relate to Parkinson's Disease (PD) and its stages [Hoehn and Yahr (H&Y) Stages 1-5]. Further, this work aims to compare perceived versus sensor-based neurocognitive abilities. In this study, 75 participants ([Formula: see text] PD; [Formula: see text] control) completed 14 tablet-based neurocognitive functional tests (e.g., motor, memory, speech, executive, and multifunction), functional movement assessments (e.g., Berg Balance Scale), and standardized health questionnaires (e.g., PDQ-39). Decision tree classification of sensor-based features allowed for the discrimination of PD from healthy controls with an accuracy of [Formula: see text], and early and advanced stages of PD with an accuracy of [Formula: see text]; compared to the current gold standard tools [e.g., standardized health questionnaires ([Formula: see text] accuracy) and functional movement assessments ([Formula: see text] accuracy)]. Significant features were also identified using decision tree classification. Device magnitude of acceleration was significant in 12 of 14 tests ([Formula: see text]), regardless of test type. For classification between diagnosed and control populations, 17 motor (e.g., device magnitude of acceleration), 9 accuracy (e.g., number of correct/incorrect interactions), and 8 timing features (e.g., time to between interactions) were significant. For classification between early (H&Y Stages 1 and 2) and advanced (H&Y Stages 3, 4, and 5) stages of PD, 7 motor, 12 accuracy, and 14 timing features were significant. Finally, this work depicts that perceived functionality of individuals with PD differed from sensor-based functionalities. In early-stage PD was shown to be [Formula: see text] lower than sensor-based scores with notable perceived deficits in memory and executive function. However, individuals in advanced stages had elevated perceptions (1.57x) for executive and behavioral functions compared to early-stage populations. Machine learning in digital health systems allows for a more comprehensive understanding of neurodegenerative diseases and their stages and may also depict new features that influence the ways digital health technology should be configured.

BACKGROUND: Mobile devices and their capabilities (e.g., device sensors and human-device interactions) are increasingly being considered for use in clinical assessments and disease monitoring due to their ability to provide objective, repeatable, and more accurate measures of neurocognitive performance. These mobile-based assessments also provide a foundation for the design of intervention recommendations.

OBJECTIVE: The purpose of this work was to assess the benefits of various physical intervention programs as they relate to Parkinson's disease (PD), its symptoms, and stages (Hoehn and Yahr (H&Y) Stages 1-5).

METHODS: Ninety-five participants (n = 70 PD; n = 25 control) completed 14 tablet-based neurocognitive functional tests (e.g., motor, memory, speech, executive, and multi-function) and standardized health questionnaires. 208 symptom-specific digital features were normalized to assess the benefits of various physical intervention programs (e.g., aerobic activity, non-contact boxing, functional strength, and yoga) for individuals with PD. While previous studies have shown that physical interventions improve both motor and non-motor PD symptoms, this paper expands on previous works by mapping symptom-specific neurocognitive functionalities to specific physical intervention programs across stages of PD.

RESULTS: For early-stage PD (e.g., H&Y Stages 1 & 2), functional strength activities provided the largest overall significant delta improvement (Δ= 0.1883; p = 0.0265), whereas aerobic activity provided the largest overall significant delta improvement (Δ= 0.2700; p = 0.0364) for advanced stages of PD (e.g., H&Y Stages 3-5).

CONCLUSIONS: As mobile-based digital health technology allows for the collection of larger, labeled, objective datasets, new ways to analyze and interpret patterns in this data emerge which can ultimately lead to new personalized medicine programs.

Previous studies indicate that moderate-to-high ethanol (EtOH) concentrations enhance dopamine (DA) neurotransmission in the mesolimbic DA system from the ventral tegmental area (VTA) and projecting to the nucleus accumbens core (NAc). However, voltammetry studies demonstrate that moderate-to-high EtOH concentrations decrease evoked DA release at NAc terminals. The involvement of γ-aminobutyric acid (GABA) receptors (GABAA Rs), glycine (GLY) receptors (GLYRs) and cholinergic interneurons (CINs) in mediating EtOH inhibition of evoked NAc DA release were examined. Fast scan cyclic voltammetry, electrophysiology, optogenetics and immunohistochemistry techniques were used to evaluate the effects of acute and chronic EtOH exposure on DA release and CIN activity in C57/BL6, CD-1, transgenic mice and δ-subunit knockout (KO) mice (δ-/-). Ethanol decreased DA release in mice with an IC50 of 80 mM ex vivo and 2.0 g/kg in vivo. GABA and GLY decreased evoked DA release at 1-10 mM. Typical GABAA R agonists inhibited DA release at high concentrations. Typical GABAA R antagonists had minimal effects on EtOH inhibition of evoked DA release. However, EtOH inhibition of DA release was blocked by the α4 β3 δ GABAA R antagonist Ro15-4513, the GLYR antagonist strychnine and by the GABA ρ1 (Rho-1) antagonist TPMPA (10 μM) and reduced significantly in GABAA R δ-/- mice. Rho-1 expression was observed in CINs. Ethanol inhibited GABAergic synaptic input to CINs from the VTA and enhanced firing rate, both of which were blocked by TPMPA. Results herein suggest that EtOH inhibition of DA release in the NAc is modulated by GLYRs and atypical GABAA Rs on CINs containing δ- and Rho-subunits.

BACKGROUND: Investigators have proposed that various physical head and neck characteristics, such as neck strength and head and neck size, are associated with protection from mild traumatic brain injury (mTBI/concussion).

OBJECTIVES: To systematically review the literature and investigate potential relationships between physical head and neck characteristics and mTBI risk in athletic and military populations.

METHODS: A comprehensive search of seven databases was conducted: MEDLINE, EMBASE, CINAHL, Scopus, SPORTDiscus, Cochrane Library, and Web of Science. Potential studies were systematically screened and reviewed. Studies on military and athletic cohorts were included if they assessed the relationship between physical head-neck characteristics and mTBI risk or proxy risk measures such as head impact kinematics.

RESULTS: The systematic search yielded a total of 11,723 original records. From these, 22 studies met our inclusion criteria (10 longitudinal, 12 cross-sectional). Relevant to our PECO (Population, Exposure, Comparator, and Outcomes) question, exposures included mTBI incidence and head impact kinematics (acceleration, velocity, displacement) for impacts during sport play and training and in controlled laboratory conditions. Outcome characteristics included head and neck size (circumference, mass, length, ratios between these measures), neck strength and endurance, and rate of force development of neck muscles.

DISCUSSION: We found mixed evidence for head and neck characteristics acting as risk factors for and protective factors against mTBI and increased susceptibility to head impacts. Head-neck strength and size variables were at times associated with protection against mTBI incidence and reduced impact kinematics (14/22 studies found one or more head-neck variable to be associated with protection); however, some studies did not find these relationships (8/22 studies found no significant associations or relationships). Interestingly, two studies found stronger and larger athletes were more at risk of sustaining high impacts during sport. Strength and size metrics may have some predictive power, but impact mitigation seems to be influenced by many other variables, such as behaviour, sex, and impact anticipation. A meta-analysis could not be performed due to heterogeneity in study design and reporting.

CONCLUSION: There is mixed evidence in the literature for the protective capacity of head and neck characteristics. We suggest field-based mTBI research in the future should include more dynamic anthropometric metrics, such as neck stiffness and response to perturbation. In addition, laboratory-based mTBI studies should aim to standardise design and reporting to help further uncover these complicated relationships.

OBJECTIVE: Multiple Sclerosis (MS) is often accompanied by myelopathy, which may be associated with progressive worsening. A specific finding of MS-associated myelopathy is McArdle sign, wherein neck flexion is associated with prominent increased limb weakness relative to that detected with neck extension. In this study, we characterized neuromotor control properties of finger extensors in association with the McArdle sign.

METHODS: A custom-built device was utilized to monitor torque production of the wrist extensors with simultaneous recording of surface electromyography of the extensor digitorum. The electromyography was decomposed and analyzed via both linear and nominal regressions.

RESULTS: Linear regressions demonstrated a strong difference between groups for MS from healthy controls and other myelopathies for motor unit action potential amplitude and average firing rate (p < 0.001). Further, linear regression demonstrated good correlations of neuromotor variables to mechanical torque output (0.24 ≤ R2  ≤ 0.76). Nominal regression distinguished MS from healthy controls with an AUC of 0.87, specificity of 0.97, and sensitivity of 0.64. Nominal regression of MS from other myelopathies demonstrated an AUC of 0.88, specificity of 0.85, and sensitivity of 0.79.

INTERPRETATION: These data demonstrate the neuromotor control factors that largely determine muscle force production change with the observation of McArdle sign; these neuromotor control factors can differentiate MS from both healthy controls and other myelopathy conditions.

BACKGROUND: Previous literature has postulated a relationship between greater hamstring stiffness and a higher risk of sustaining injury. Shear wave elastography (SWE) presents a relatively new means for non-invasive evaluation of soft tissue elasticity pre- and post- injury or intervention.

PURPOSE: 1. To establish baseline hamstring stiffness measures for young competitive athletes and (2) determine effect of targeted neuromuscular training (TNMT) on shear wave stiffness of the hamstring.

STUDY DESIGN: Un-blinded, prospective, non-randomized, cohort study.

METHODS: Six-hundred forty-two lower extremities from 321 high school and collegiate basketball athletes (177 F: 139 M) were examined for hamstring stiffness prior to the start of their competitive basketball season. Teams were cluster assigned to either the control or intervention (TNMT) group. Subjects in the control group underwent regular season activities as directed, with no influence from the research team. For the TNMT group, the research team introduced a hamstring targeted dynamic warm-up program as an intervention focused on activating the hamstring musculature.

RESULTS: Collegiate status was significant to hamstring stiffness for both sexes (p ≤ 0.02), but hamstring stiffness did not correlate to age or sex (r2 ≤ 0.08). Intervention was a significant factor to hamstring stiffness when the hip was positioned in extension (p ≤ 0.01), but not in deeper flexion (p = 0.12). This effect was sex-specific as TNMT influenced hamstring stiffness in females (p = 0.03), but not in males (p ≥ 0.13). Control athletes suffered three HAM injuries; TNMT athletes suffered 0 hamstring injuries.

CONCLUSION: Higher SWE measurements correlated with increased risk of injury, male sex, and collegiate athletics. TNMT intervention can lessen muscle stiffness which may reduce relate to injury incidence. Intervention effectiveness may be sex specific.

LEVEL OF EVIDENCE: II.

BACKGROUND: Neuromuscular deficits exist following anterior cruciate ligament (ACL) injury. To observe these deficits, we examined nonlinear characteristics of vastus medialis electromyography (EMG) signals during submaximal isometric knee extensor contractions. Our purpose was to examine if determinism and entropy in EMG signals reflected neuromuscular control deficits in individuals with ACL-deficient limbs.

METHODS: 24 participants (12 male, 12 female, mean age = 18.8 ± 3.1 years) with unilaterally injured ACLs and 25 age-similar healthy controls (11 male, 14 female, mean age = 18.8 ± 3.1 years) volunteered. Isometric knee extensions were tested at 10%, 25%, 35%, and 50% maximum voluntary contractions. Surface electrodes adhered over the vastus medialis captured EMG signals. EMG data were processed with recurrence quantification analyses. Specifically, determinism (an index of system predictability) and entropy (an index of system disorder) were calculated from recurrence plots.

FINDINGS: Determinism and entropy in EMG signals were lower in the injured than uninjured limb, and lower than that from healthy controls (P < .05).

INTERPRETATION: Vastus medialis EMG signals from the injured limb were less predictable and less complex than those from healthy limbs. The findings reflect impaired neuromuscular control in the injured limb's quadriceps and are consistent with a 'loss of complexity' hypothesis in physiologic signals emanating from pathologic states. Determinism and entropy in EMG signals may represent biomarkers of one's neuromuscular control system.

BACKGROUND: Due to the COVID-19 pandemic, beneficial physical intervention classes for individuals with Parkinson's disease (PD) were cancelled.

OBJECTIVE: To understand effects of the COVID-19 stay-at-home mandate and the inability to participate in recommended and structured physical interventions as a consequence of these mandates, specifically designed mobile assessments were used that collected both self-reporting information and objective task-based metrics of neurocognitive functions to assess symptom changes for individuals with PD.

METHODS: Self-reporting questionnaires focusing on overall quality of life (e.g., when individuals typically feel at their best, changes in activity levels, and symptom progression) were given to all individuals (n = 28). In addition, mobile-based neurocognitive assessments were administered to a subset of the population (n = 8) to quantitatively assess changes due to COVID-19 restrictions.

RESULTS: The highest self-reported factors in which individuals denoted feeling their best were after exercise (67.86%) and being in a comfortable and supportive environment (60.71%). Objective measures found overall duration of physical activity during the stay-at-home mandate decreased significantly (p = 0.022). With the lack of overall activity, 82.14%of individuals self-reported having at least one symptom that worsened moderately or higher. Further testing, using mobile-based assessments, showed average completion times of functional tasks increased, taking about 2.1 times longer, while accuracy metrics showed overall degradation.

CONCLUSION: Although the COVID-19 stay-at-home mandate was intended to help protect individuals at high risk from coming into contact with the virus, it also prevented individuals from receiving recommended supervised exercise interventions resulting in significant negative effects in social well-being and across motor and speech neurocognitive tasks for individuals with PD.

A large variety of data filtration techniques exist in biomechanics literature. Data filtration is both an 'art' and a 'science' to eliminate noise and retain true signal to draw conclusions that will direct future hypotheses, experimentation, and technology development. Thus, data consilience is paramount, but is dependent on filtration methodologies. In this study, we utilized ligament strain, vertical ground reaction force, and kinetic data from cadaveric impact simulations to assess data from four different filters (12 vs. 50 Hz low-pass; forward vs. zero lag). We hypothesized that 50 Hz filtered data would demonstrate larger peak magnitudes, but exhibit consilience of waveforms and statistical significance as compared to 12 Hz filtered data. Results demonstrated high data consilience for matched pair t test correlations of peak ACL strain (≥ 0.97), MCL strain (≥ 0.93) and vertical ground reaction force (≥ 0.98). Kinetics had a larger range of correlation (0.06-0.96) that was dependent on both external load application and direction of motion monitored. Coefficients of multiple correlation demonstrated high data consilience for zero lag filtered data. With respect to in vitro mechanical data, selection of low-pass filter cutoff frequency will influence both the magnitudes of discrete and waveform data. Dependent on the data type (i.e., strain and ground reaction forces), this will not likely significantly alter conclusions of statistical significance previously reported in the literature with high consilience of matched pair t-test correlations and coefficients of multiple correlation demonstrated. However, rotational kinetics are more sensitive to filtration selection and could be suspect to errors, especially at lower magnitudes.

BACKGROUND: McArdle sign is a phenomenon of impaired gait and muscle weakness that occurs with neck flexion, immediately reversible with neck extension. A recent report measured the specificity of this sign for multiple sclerosis by measuring differences in peak torque of the extensor digitorum between neck extension and flexion.

METHODS: This substudy included 73 participants (29 multiple sclerosis, 20 non-multiple sclerosis myelopathies, 5 peripheral nerve disorders, and 19 healthy controls). The effect of neck position was assessed on muscle stiffness and neuromechanical error of the extensor digitorum.

FINDINGS: Patients with multiple sclerosis had greater neuromechanical error (sum of squared error of prediction) compared to controls (P = 0.023) and non-multiple sclerosis myelopathies (P = 0.003). Neuromechanical error also provided improved sensitivity/specificity of McArdle sign. Peak torque, muscle stiffness, and neuromechanical error could distinguish multiple sclerosis from other myelopathies with 80% specificity and 97% sensitivity (AUC = 0.95).

INTERPRETATION: A decrease in muscle stiffness and neuromechanical error in neck flexion compared to extension are additional indicators for a diagnosis of multiple sclerosis. Analysis of muscle stiffness may provide insights into the pathophysiology of this specific clinical sign for multiple sclerosis. Furthermore, muscle stiffness may provide an additional accurate, simple assessment to evaluate multiple sclerosis therapeutic interventions and disease progression.