Validating DystoniaNet Deep Learning for Diagnosing Isolated Dystonia and Similar Movement Disorders

What this Trial Is About

This research focuses on isolated dystonia, a movement disorder that causes involuntary muscle contractions leading to abnormal twisting movements and postures. Diagnosing dystonia is challenging because there is no biomarker or gold standard test, resulting in frequent misdiagnoses and delays averaging over 10 years. The study aims to clinically validate DystoniaNet, a deep learning platform designed to improve the accuracy and speed of dystonia diagnosis through retrospective and prospective studies. The study consists of two parts retrospective studies will validate DystoniaNets diagnostic ability by comparing patients with dystonia to healthy individuals and to patients with other neurological or non-neurological conditions that mimic dystonia symptoms. The prospective randomized study will test DystoniaNets performance for fast, objective diagnosis in actual clinical settings. The intervention involves using DystoniaNet as a diagnostic test to distinguish dystonia from similar disorders. Participants will undergo clinical evaluations and brain imaging data will be analyzed by the DystoniaNet algorithm to measure the correctness and speed of dystonia diagnosis over four years. Researchers will monitor diagnostic accuracy, time to diagnosis, and compare results across patient groups. This study aims to advance dystonia diagnosis to clinical practice, enabling earlier treatment and better patient outcomes, with participation lasting throughout the study period ending in 2028.

Clinical Validation of DystoniaNet Deep Learning Platform for Diagnosis of Isolated Dystonia