PROJECT TITLE :

Private Facial Prediagnosis as an Edge Service for Parkinson's DBS Treatment Valuation

ABSTRACT:

Facial phenotyping for the purpose of medical prediagnosis has recently been successfully exploited as a novel way for the preclinical assessment of a range of rare genetic diseases. In these cases, facial biometrics has been shown to have rich links to the underlying genetic or medical causes of the condition in question. In this paper, we propose an Artificial Intelligence of Things (AIoT) edge-oriented privacy-preserving facial prediagnosis framework to investigate the effects of Deep Brain Stimulation (DBS) treatment on Parkinson's Disease (PD) patients. Our goal is to extend this facial prediagnosis technology to a more general disease, Parkinson's Disease (PD). In the framework that has been proposed, a novel edge-based privacy-preserving framework is proposed to implement private deep facial diagnosis as a service over an AIoT-oriented information theoretically secure multi-party communication scheme. This is done in light of the fact that data privacy has been a primary concern in the direction of a wider exploitation of electronic health and medical records (EHR/EMR) over cloud-based medical services. In our experiments using a facial dataset collected from PD patients, we demonstrated, for the very first time, that facial patterns can be used to evaluate the facial difference of PD patients who have undergone DBS treatment. This demonstrates the potential of our facial prediagnosis as a reliable edge service for grading the severity of PD in patients. In addition, we implemented a privacy-preserving information theoretical secure deep facial prediagnosis framework. This framework is able to achieve the same level of accuracy as the non-encrypted one.