Automatic Evaluation of Dysarthric Speech and Telemedical Use in the Therapy

After a stroke, the speech quality of the patients is often reduced. This is usually caused by a deficit of the motor abilities of the vocal tract. The result is slurred speech. In the various patients, however, very different forms can appear. In the course of therapy, evaluation of the speech quality is required to determine the success of the treatment. At the moment, this assessment is performed only perceptually. This form of assessment is subject to strong intraand inter-individual variation. Therefore, an “objective” assessment is not guaranteed. In this study, we present a rater-independent method for evaluating speech disorders in dysarthria. We use methods of automatic speech recognition. The idea is to determine the speech intelligibility – the main outcome parameter of speech – automatically by an automatic speech recognizer. A correlation of -0.89 was obtained between the criterion “intelligibility” and the recognition rate of the automatic system, in a preliminary study. The second part of this paper deals with an additional problem with this kind of patient. Very often, the stroke leads to partial facial paresis and generally to reduced mobility. Therefore, it is desirable that therapy sessions are performed in a telemedical setup. We report on our work towards such a telemedical diagnosis and rehabilitation system which will allow sessions with a therapist and – at the same time – diagnose the patient and track the recovery process. We describe the equipment (web camera, 3-D camera, stereo microphone, and Internet connection), the patient's environment, and the working environment of the therapist. Depending on the network connection, live images of the patient can be sent to the therapist at a rate of 20 frames per second (fps) at existing LAN connections or 3 fps with a DSL 6000 connection. At the patient host, a three-dimensional face model of the patient performing a facial exercise can be generated and transferred to the therapist in real-time (LAN) or three times real-time (DSL 6000).