This paper proposes a novel method of measuring muscle surface deformation using flexible piezoelectric thin film sensors to allow PC operation by people with severe physical disabilities. Piezoelectric signals are first measured using thin film sensors that are flexible, compact and sensitive. Strength and velocity information relating to muscle surface deformation is extracted from the signals measured and used as input for a neural network. In this way, the motions of users can be estimated by learning differences in muscle surface deformation information for particular individuals. A PC can also be operated with only a few motions if an assistive operation mode is introduced. This paper details the development of a mask-type interface with piezoelectric thin film sensors and outlines experiments conducted on the measurement of muscle surface deformation. The results showed that multiple facial motions can be discriminated accurately from the muscle deformation information obtained. PC operation experiments involving three subjects (including one with paralysis from a cervical cord injury) also produced outcomes indicating that a PC can be operated voluntarily based on facial motions by people with severe physical disabilities.
