Poster Session 09 Program Schedule
02/16/2024
03:30 pm - 04:45 pm
Room: Shubert Complex (Posters 1-60)
Poster Session 09: Epilepsy | Oncology | MS | Infectious Disease
Final Abstract #51
Robustness of Balance-Related Digital Biomarkers in Differentiating People with Multiple Sclerosis (MS) from People Without MS.
Sepideh Heydari, University of Victoria, Victoria, Canada
Jodie Gawryluk, University of Victoria, Victoria, Canada
John Ralston, Neursantys, Menlo Park, United States
Category: Multiple Sclerosis/ALS/Demyelinating Disorders
Keyword 1: multiple sclerosis
Keyword 2: aging disorders
Keyword 3: neuroimaging: structural connectivity
Objective:
This study investigated the extent to which accelerated aging due to neurodegeneration is detectable using digital biomarkers from a wearable sensor during postural sway in people presenting with and without multiple sclerosis.
Participants and Methods:
Participants included 50 individuals without MS (26 Females, mean age= 48.83 ± 14.95, mean education= 19.39 ± 4.24years) and 29 participants with MS (21 Females, mean age= 57.24 ± 11.32, mean education= 16.80 ± 3.09 years). Participants performed a balance test alternating between segments of eyes open and closed while wearing a medical grade wearable sensor that collected physiological acceleration vibration (phybrata) data during postural sway. Subjective measures of fatigue and pain were collected from both groups. Participants with MS also went through an MRI scan where DTI data to correlate white matter integrity with sensor digital biomarkers. DTI data were collected at West Coast Medical Imaging on a 3T GE MRI scanner, using the following parameters: TR = 8000ms, TE=101ms, flip angle=90°, 52 slices, voxel size = 1.4 x 1.4 x 2.0mm. There were 45 diffusion-weighted images (b=1000 s/mm ) and 3 non-diffusion-weighted images (b=0 s/mm ; b0) per scan.
Results:
ANOVA results on phybrata power revealed a main effect of Population (F(1,552)=30.31, p < 0.001), a main effect of Condition (F(1,552)= 12.11, p <0.001) and a significant interaction between Population and Condition (F(1,552)=9.84, p=0.001). Findings also revealed a significant negative relationship between FA and phybrata signal frequency in the fornix (MNI coordinates: 45, 60, 44) and right optic radiation (MNI coordinates: 38, 29, 42). Phybrata power was also higher for the older MS adults compared to the younger adults, but not for the non-MS population. Non-MS group did not have underlying balance issues, so it was expected not to see any difference in the non-MS phybrata power between young and old groups.
Conclusions:
Our wearable sensor shows potential application for use in patient monitoring and signaling need for early intervention in the aging population. Our results indicate that the digital biomarkers extracted from the sensor's phybrata data can differentiate age clusters of young and old adults with and without MS. These results suggest that phybrata data obtained from the sensor could be used as indicators for age-related balance impairment in healthy aging adults, as well as biomarkers to signal increased neurodegeneration and accelerated aging in people with multiple sclerosis.
| 
