Novel Noninvasive Brain Disease Detection System Using a Facial Image Sensor.

Area of Science:

• Neurology
• Medical Imaging
• Biomedical Engineering

Background:

• Brain diseases are a leading cause of death, with traditional diagnostics being inconvenient.
• There is a growing need for noninvasive, efficient, and patient-friendly brain disease detection methods.

Purpose of the Study:

• To propose and evaluate a novel noninvasive brain disease detection system.
• To utilize facial color analysis for early detection of brain conditions.

Main Methods:

• A system capturing facial images using a specialized sensor.
• Automatic localization of four facial key blocks for feature extraction.
• Classification using a Probabilistic Collaborative based Classifier on extracted color features.

Main Results:

• The system achieved high performance metrics: accuracy (95%), sensitivity (94.33%), and specificity (95.67%).
• The best results were obtained using a specific facial key block combination.
• Average processing time per sample was under one minute.

Conclusions:

• The proposed facial color analysis system is a viable noninvasive method for brain disease detection.
• The Probabilistic Collaborative based Classifier demonstrated suitability for this application.
• This approach offers a promising, patient-friendly alternative to traditional diagnostic methods.