Faculty Bibliography

Recent advances in wearable sensor technology and machine learning (ML) have allowed for the seamless and objective study of human motion in clinical applications, including Parkinson's disease, and stroke. Using ML to identify salient patterns in sensor data has the potential for widespread application in neurological disorders, so understanding how to develop this approach for one's area of inquiry is vital. We previously proposed an approach that combined wearable inertial measurement units (IMUs) and ML to classify motions made by stroke patients. However, our approach had computational and practical limitations. We address these limitations here in the form of a primer, presenting how to optimize a sensor-ML approach for clinical implementation. First, we demonstrate how to identify the ML algorithm that maximizes classification performance and pragmatic implementation. Second, we demonstrate how to identify the motion capture approach that maximizes classification performance but reduces cost. We used previously collected motion data from chronic stroke patients wearing off-the-shelf IMUs during a rehabilitation-like activity. To identify the optimal ML algorithm, we compared the classification performance, computational complexity, and tuning requirements of four off-the-shelf algorithms. To identify the optimal motion capture approach, we compared the classification performance of various sensor configurations (number and location on the body) and sensor type (IMUs vs. accelerometers). Of the algorithms tested, linear discriminant analysis had the highest classification performance, low computational complexity, and modest tuning requirements. Of the sensor configurations tested, seven sensors on the paretic arm and trunk led to the highest classification performance, and IMUs outperformed accelerometers. Overall, we present a refined sensor-ML approach that maximizes both classification performance and pragmatic implementation. In addition, with this primer, we showcase important considerations for appraising off-the-shelf algorithms and sensors for quantitative motion assessment.

Objective: To explore the applicability of an ambulatory inertial sensor (G-walk) to characterize gait function during the Timed Up and Go (TUG) Test under three different conditions.
Background(s): In Parkinson's disease (PD), the current lack of both reliable and feasible biomarkers of gait function and mobility limits the objective characterization of motor ability, clinical progression, and responsiveness to treatments. Current assessments of motor function rely on a clinicians' subjective judgement and/or the patient's self-reported questionnaires, which are not sensitive in capturing subtle changes over time and restrict comparability across raters. Ambulatory inertial sensors allow for non-invasive, wireless transmission of accurate quantitative data and therefore, may represent a useful tool in ambulatory settings. Design/Methods: Nineteen (19) PD patients (H&Y <4) and 10 agematched controls (CTRL) were consecutively enrolled to undergo inertial TUG (iTUG) testing under three experimental conditions: normal walking (iTUGnorm), dual task walking (iTUGcog), and at maximum speed (iTUGfast). The time needed to complete each test was sub-divided into six distinct phases quantified by the sensor: sitto- stand (1), forward gait (2), mid-turn (3), return gait (4), end-turn (5) and stand-to-sit (6). Other assessments included UDPRS Part III, MoCA, depression, fatigue, Benton and Rey-Osterrieth visual tests.
Result(s): A total of nineteen PD patients and ten CTRLs completed all assessments. PD patients were divided into mild (H&Y=2, n=12) and moderate (H&Y=3, n=7) disease severity. One-way-ANOVA and correlation analysis were performed. Different patterns of kinematic performance were observed (figure 1.A and 1.B). In PD, iTUG correlations were found with cognitive function, visual performance and motor severity, while in CTRLs there was only a correlation with motor performance only. iTUGfast performance seemed more sensitive experimental condition when PD was stratify by severity (figure 1.B).
Conclusion(s): iTUG assessed by an ambulatory inertial sensor is a quick, sensitive and feasible tool for objective measurements of functional mobility in PD. Utilizing validate tests for mobility and gait under different stress conditions can provide distinct information of gait function and mobility. Future longitudinal studies are warranted to better characterize the sensitivity to disease progression and the potential for monitoring and optimizing therapeutic interventions in this patient population. (Figure Presented)

During movement, modulation of beta power occurs over the sensorimotor areas, with a decrease just before its start (event-related desynchronization, ERD) and a rebound after its end (event-related synchronization, ERS). We have recently found that the depth of ERD-to-ERS modulation increases during practice in a reaching task and the following day decreases to baseline levels. Importantly, the magnitude of the beta modulation increase during practice is highly correlated with the retention of motor skill tested the following day. Together with other evidence, this suggests that the increase of practice-related modulation depth may be the expression of sensorimotor cortex's plasticity. Here, we determine whether the practice-related increase of beta modulation depth is equally present in a group of younger and a group of older subjects during the performance of a 30-minute block of reaching movements. We focused our analyses on two regions of interest (ROIs): the left sensorimotor and the frontal region. Performance indices were significantly different in the two groups, with the movements of older subjects being slower and less accurate. Importantly, both groups presented a similar increase of the practice-related beta modulation depth in both ROIs in the course of the task. Peak latency analysis revealed a progressive delay of the ERS peak that correlated with the total movement time. Altogether, these findings support the notion that the depth of beta modulation in a reaching movement task does not depend on age and confirm previous findings that only ERS peak latency but not ERS magnitude is related to performance indices.

BACKGROUND:Little is known with respect to behavioral markers of subjective cognitive decline (SCD), a condition initially described in association with Global Deterioration Scale (GDS) stage 2. OBJECTIVE:Two-year interval behavioral markers were investigated herein. METHODS:Subjects from a published 7-year outcome study of GDS stage 2 subjects were selected. This study had demonstrated a hazard ratio of 4.5 for progression of GDS stage 2, in comparison with GDS stage 1 (no subjective or objective cognitive decline) subjects, after controlling for demographic and temporal variables. Because GDS 2 subjects have previously demonstrated impairment in comparison with healthy persons free of complaints, we herein suggest the terminology "SCD(I)" for these persons. 98 SCD(I) persons, 63 women and 35 men, mean baseline age, 67.12±8.75 years, with a mean educational background of 15.55±2.60 years, and mean baseline MMSE scores of 28.9±1.24 were followed for 2.13±0.30 years. RESULTS:Observed annual decline on the GDS was 6.701% per annum, very close to a 1986 published estimate. At follow up, the MMSE, and 7 of 8 psychometric tests did not decline significantly. Of 21 Hamilton Depression Scale items, 2 improved and the remainder were unchanged. Anxieties declined from multiple perspectives. The Brief Cognitive Rating Scale (BCRS) declined significantly (p < 0.001), with component declines in Remote memory (p < 0.01), and Functioning/self-care (p = 0.01). CONCLUSION/CONCLUSIONS:SCD(I) persons decline at an annual rate of approximately 6.7% /year from several recent studies. The BCRS assessments and the Digit Symbol Substitution Test can be sensitive measures for future studies of progression mitigation.