VCOM Research Day Program Book 2023

Medical Student Research Biomedical

27 Use of Artificial Intelligence in the Diagnosis of Huntington Disease

Fatima Siddiqui, OMS III; Irushi Dissanayake, DO Corresponding author: fsiddiqui@vt.vcom.edu

Edward Via College of Osteopathic Medicine-Virginia Campus

develop machine learning models. Performance data included accuracy, sensitivity, and specificity, all of which were reported to be >80%. Conclusion: Machine learning models trained and tested with HD patient data have been reported in the literature. This review provides a summary and analysis of the models that have been developed for the purpose of detecting and/or classifying severity of HD.

Context: Huntington Disease (HD) is classified as a neurodegenerative disease that presents with impairment of motor, cognitive, and behavioral function. Genetic testing for the mutant protein, HTT, is the only gold standard diagnostic test available. Recent technological advances are a promising approach to addressing the limitation in diagnostic tests available. We carried out a systematic review to identify and evaluate current artificial intelligence (AI) methods being developed or used for clinical diagnosis of HD. To our knowledge, current literature does not have a systematic review investigating the use of AI in diagnosis of HD.

Objective: To identify and evaluate current AI methods, their efficacy, and their use in diagnosis of HD. Methods: For this systematic review, we searched the literature for articles discussing AI or machine learning and HD. Next, inclusion and exclusion criteria were applied to extract articles relevant to the objectives of our study. Finally, a quality assessment was performed. 16 articles met the search criteria, inclusion/exclusion criteria, and quality assessment. Results: We identified gait dynamics as the primary data type used in most studies (56%) to

82