Charisse Litchman^1,2*, Noah Rubin³, Larry Rubin⁴, Tess Litchman⁵, Sarah Wesely¹, John Keaney¹, Timothy Vartanian⁵ and Sharon Stoll^1,2

¹Yale New Haven Hospital, United States
²Yale School of Medicine, United States
³Brown University, United States
⁴BeCare Link, Beth Israel, Israel

*Corresponding Author: Charisse Litchman, Yale New Haven Hospital, United States.

Received: September 15, 2023; Published: November 27, 2023

Abstract

Background:  Lack of access to MS specialty care, exacerbated by the COVID-19 pandemic, has prevented many patients suffering from Multiple Sclerosis (MS) and other neurodegenerative diseases from receiving timely in-person care, resulting in less optimal care.  As a result, there has been an increased adoption of remote patient monitoring, often employing machine learning as a tool.

The BeCare MS Link mobile app collects quantitative measurements of neurologic function as users perform different activities on the mobile app.  The app was designed to monitor changes in neurologic function and patient-reported symptomatology. The intention is for the app to serve as a remote equivalent of the Expanded Disability Status Scale (EDSS) and other standard clinical metrics of MS progression.

Methods:  In total 26 subjects were enrolled in the study over a period of 6 months.  The study was concluded early because of the COVID-19 pandemic. Our research compiled MS disability categorizations of 15 subjects obtained by the BeCare MS Link MLA (Machine Learning Algorithm) and by 2 MS neurologists at Yale New Haven Hospital MS Clinic. The machine learning inputs were derived from data collected by specialized mobile app activities incorporated into the BeCare MS Link app.

Results: Each participant was categorized by their Clinical EDSS score into to one of these categories: Normal (<1.0), Mild (1.0- 3.0), Moderate (3.5 – 6.0) or Severe [6.5 – 9.0]. These clinical classifications were compared to MLA classifications that were obtained using only BeCare MS Link-derived data corresponding to each subject’s visit. To increase the number of available data points, subjects were evaluated in an initial study visit and at a follow-up visit at least three months later. Of the 26 enrolled subjects, 15 subjects complied with the study protocol at the time of their initial visit by using the BeCare App with sufficient frequency and completing sufficient activities to generate MLA-calculated EDSS classifications. Seven of the 15 initial study visit subjects returned for a second visit as part of this study. The study population included 15 patients, between the ages of 39-72 years old, and with 13/15 being female.

The clinical categorization of disability as normal, mild, moderate, and severe were compared to the BeCare MS Link categorization. The clinical and BeCare MS Link categorizations were in perfect agreement 86% of the time. BeCare MS Link categorization differed from the clinical categorization by no more than 1 category (e.g. Mild instead of Normal or Moderate) in 100% of assessments. The EDSS score difference between clinician calculation and BeCareLink machine learning (MLA) was less than one point 91% of the time and less than 1.5 points 95% of the time.

Conclusions: The BeCare MS Link categorization of MS patients as having no disability or mild, moderate, or severe disability and the calculation of EDSS demonstrated high correlation to clinical categorization of disability in MS patients.

 Keywords: Machine Learning; Evaluation; MS Patients; COVID

References

1. Kurtzke JF. “Disability rating scales in multiple sclerosis”. Annals of the New York Academy of Sciences 436 (1983): 347-360.
2. Polman CH., et al. “Diagnostic criteria for multiple sclerosis: 2010 Revisions to the McDonald criteria”. Annals of Neurology2 (2011):292-302.
3. Zeeshan Ahmed., et al. “Artificial intelligence with multi-functional machine learning platform development for better healthcare and precision medicine”. Database (Oxford) (2019).
4. Sidey-Gibbons J and Sidey-Gibbons C. “Machine learning in medicine: a practical introduction”. BMC Medical Research Methodology 19(2019): 64.
5. Macaron G., et al. “Technology-enabled assessments to enhance multiple sclerosis clinical care and research”. Neurology: Clinical Practice 10 (2020): 222-231.
6. Rudick R., et al. “The multiple sclerosis performance test (MSPT) an iPad-based disability assessment tool”. Journal of Visualized Experiments 88 (2014): e51318.
7. Katzan IL and Rudick RA. “Time to integrate clinical and research informatics”. Science Translational Medicine 4 (2012):1-5.
8. Arndt BG., et al. “Tethered to the EHR: primary care physician workload assessment using HER event log data and time-motion observations”. Annals of Family Medicine 15 (2017):419-426.
9. Collins CD., et al. “A comparative analysis of Patient-Reported Expanded Disability Status Scale tools”. Multiple Sclerosis 10 (2016):1349-1358.
10. Francis DA., et al. “An assessment of disability rating scales used in multiple sclerosis”. Archives of Neurology 3 (1991):299-301.
11. Spain R and Wooliscroft L. “Diagnosing Multiple Sclerosis Using the McDonald Criteria”. US Department of Veterans Affairs (2023).

Citation

Citation: Charisse Litchman., et al. “AI Classification of MS Disability with the Utilization of a Mobile App”. Acta Scientific Neurology 6.12 (2023): 28-32.

Copyright

Copyright: © 2023 Charisse Litchman., et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.