Dysarthria diagnosis via respiration and phonation

This report discusses the implementation of a computerized application - the Computerised Frenchay Dysarthria Assessment Procedure (CFDA) - which uses digital signal processing (DSP) techniques to objectively evaluate digitised speech recordings in order to detect any symptoms of dysarthria (a type of motor speech disorder). This investigation focuses specifically on two CFDA diagnostic sub-applications which assess a patient's ability to maintain a prolonged exhalation (the “Respiration at Rest” task) as well as execute a steady state phonation (i.e. the “Sustained Phonation” task). It is demonstrated that combining the functionality of these two sub-applications substantially increases their diagnostic effectiveness in the context of identifying a particular variety of dysarthria known as spastic dysarthria. It is further demonstrated that certain patterns of energy fluctuations are closely correlated with manifestations of spastic dysarthria. The CFDA is intended for use in real-world conditions to assist speech and language therapists when conducting clinical evaluations.