Abstract
Our long-term goal is to employ smartphone-embedded sensors to measure various neurological functions in a patient-autonomous manner. The interim goal is to develop simple smartphone tests (apps) and evaluate the clinical utility of these tests by selecting optimal outcomes that correlate well with clinician-measured disability in different neurological domains. In this paper, we used prospectively-acquired data from 112 multiple sclerosis (MS) patients and 15 healthy volunteers (HV) to assess the performance and optimize outcomes of a Level Test. The goal of the test is to tilt the smartphone so that a free-rolling ball travels to and remains in the center of the screen. An accelerometer detects tilting and records the coordinates of the ball at set time intervals. From this data, we derived five features: path length traveled, time spent in the center of the screen, average distance from the center, average speed while in the center, and number of direction changes underwent by the ball. Time in center proved to be the most sensitive feature to differentiate MS patients from HV and had the strongest correlations with clinician-derived scales. Its superiority was validated in an independent validation cohort of 29 MS patients. A linear combination of different Level features failed to outperform time in center in an independent validation cohort. Limited longitudinal data demonstrated that the Level features were relatively stable intra-individually within 4 months, without definitive evidence of learning. In contrast to previously developed smartphone tests that predominantly measure motoric functions, Level features correlated strongly with reaction time and moderately with cerebellar functions and proprioception, validating its complementary clinical value in the MS app suite. The Level Test measures neurological disability in several domains in two independent cross-sectional cohorts (original and validation). An ongoing longitudinal cohort further investigates whether patient-autonomous collection of granular functional data allows measurement of patient-specific trajectories of disability progression to better guide treatment decisions.
Highlights
There is a need for simple, but reliable, testing of diverse neurological functions to identify neurological disability in situations where a neurologist is absent, to track the disability of patients longitudinally, and to produce reliable outcomes for drug development
We recently demonstrated that digitalizing these tests offers additional benefits: can the test be performed by patients in frequent intervals outside of clinic, but input from smartphone sensors and internal clocks can be assembled into novel, secondary outcomes
Time in center correlated with traditional clinical scales that are biased toward motoric functions and gait, such as EDSS, 9-hole peg test (9HPT), and 25-foot walk (25FW), as well as with the statistical-learning-optimized Combinatorial Weight-adjusted clinical scale, CombiWISE (4)
Summary
There is a need for simple, but reliable, testing of diverse neurological functions to identify neurological disability in situations where a neurologist is absent (e.g., third world countries or rural areas), to track the disability of patients longitudinally, and to produce reliable outcomes for drug development. For the reasons outlined below, our long-term goal is to develop an integrated collection (i.e., Multiple Sclerosis [MS] suite) of simple tests (apps) that reliably measure disability in several/most neurological domains, using smartphone embedded censors in a patient-autonomous manner. This goal is approached via several interim aims: (1) To develop individual simple digital tests and to develop/optimize quantitative outcomes derived from the smartphone sensors/time values; (2) To validate clinical utility of developed outcomes by comparing them to the clinical gold standard represented by clinician-generated disability scores in specific neurological functions in cross-sectional cohort(s); (3) To integrate validated outcomes from multiple apps into a combined clinical score that correlates highly with existing, clinician-derived disability scores and that is optimized for change in time, measured in longitudinal cohort(s). Time and expense constraints preclude collecting these optimized outcomes in routine clinical practices
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
