Using Automatic Skeleton Generation To Extract The Body Movement During The ‘Suggested Clinical Immobilization Test’

SCIT #1 Vs. SCIT #2: Framing The Clinical Discussion With An Automatic Skeleton Generation Algorithm

Based on a previous study on the annotation of Restless Legs Syndrome (RLS) video recordings, we determined that using an algorithm for automatic skeleton generation can standardize the annotation, extract more features, and visualize torso postures that were not identified in the previous study. Using OpenPose, we developed an algorithm to detect movements of the body and extract relevant features.

The Suggested Clinical Immobilization Test (SCIT) is an adaptation of the laboratory-based SIT allowing standardized observation of movements during clinical assessments, when the patient sits on a height- wise appropriate chair. The diagnostic application of standardized observations during SCIT vary depending on the professionals training back-ground. In this work, we review our clinical observation-based recognition concepts for the diagnosis of RLS with a skeleton generation algorithm.