MENTAL WORKLOAD MEASUREMENT USING BRAINWAVE ANALYSIS

THIS PROJECT WILL DEVELOP A REAL-TIME MENTAL WORKLOAD MEASUREMENT SYSTEM BASED UPON A PROVEN BRAINWAVE ANALYSIS APPROACH THAT IS CURRENTLY IN RESEARCH USE. THE EXISTING APPROACH WILL BE AUGMENTED BY USE OF NEURAL NETWORK TECHNIQUES FOR FEATURE EXTRACTION AND PATTERN RECOGNITION. THE FEATURE EXTRACTION NEURAL NETWORK WILL UTILIZE A SELF-ORGANIZING, UNSUPERVISED TRAINING APPROACH. THE PATTERN RECOGNITION NEURAL NETWORK WILL BE TRAINED VIA SUPERVISED TRAINING. DURING PHASE I OF THE PROJECT THE ABILITY OF THE PRIMARY ANALYSIS APPROACH AND THE AUGMENTED APPROACH TO SUCCESSFULLY MEASURE MENTAL WORKLOAD FOR A FIXED TASK IN A LABORATORY SETTING WILL BE ASSESSED. A DESIGN FOR A REAL-TIME VERSION OF THE SYSTEM WILL ALSO BE DEVELOPED IN PHASE I. IN PHASE II A PROTOTYPE REAL-TIME REALWORLD SYSTEM FOR AIR FORCE USE WILL BE BUILT AND DEMONSTRATED. IN PHASE III OF THE PROJECT THE PROTOTYPE REAL-TIME SYSTEM WILL BE PRODUCTIZED IN PREPARATION FOR USE IN OPERATIONAL ENVIRONMENTS BY PHYSICIANS AND HOSPITALS IN CLINICAL PSYCHIATRY (E.G., FOR REAL-TIME MONITORING OF PATIENT RESPONSE DURING PSYCHIATRIC INTERVIEWS).