Novel Computational Video Diagnostic Tool for Parkinson’s Disease asymmetry of facial features in medicine
Links – Summary
Here we develop a novel system intended to serve as a diagnostic tool to identify patients with Parkinson’s disease (PD), to quantify the disease’s progression, and evaluate the efficacy of different treatments.
The system is based on an algorithm of scalar and non-scalar quantitative measurements of facial features, acquired through time evolving 2D and 3D data (video sequences and 3D mesh sequences), and converted into a form on which time-space analysis of symmetry is performed. The system also aims at developing quantitative measurements of the spatial-temporal pattern of facial motion, mainly but not exclusively, fluctuating asymmetries as well as position and temporal characteristics.
The study focus on testing and establishing the accuracy, sensitivity and specificity of the system in its different possible function:
1. A non-invasive objective diagnostic tool, i.e. accurately differentiate between healthy subjects and those affected by neurological movement disorders.
2. An objective mode of evaluating of the disorder’s progression.
3. A remote data acquisition system able to follow-up modifications of patients’ state (telemedicine).
4. A system enabling to test the efficacy of medications on disease progression.
Main Collaborators
(i) Dr. Ilana Schlesinger, Department of Neurology, Rambam medical center
(ii) Dr. Hagit Hel-Or, Depertment of Computer Science,HaifaUniversity
(iii) Dr. Noam Amir, Communication Disorders, Faculty of Medicine,TelAvivUniversity
Students
(i) Noa Privman Horesh (B.Sc) Depertment of Computer Science, Haifa University. M.Sc student co-advised with DR. Hagit Hel-Or
shmulikraz.wordpress.com 