For the purpose of improving the efficiency of aphasia rehabilitation training, artificial intelligence-scheduling function is added in the aphasia rehabilitation software, and the software's performance is improved. With the characteristics of aphasia patient's voice as well as with the need of artificial intelligence-scheduling functions under consideration, the present authors have designed a set of endpoint detection algorithm. It determines the reference endpoints, then extracts every word and ensures the reasonable segmentation points between consonants and vowels, using the reference endpoints. The results of experiments show that the algorithm is able to attain the objects of detection at a higher accuracy rate. Therefore, it is applicable to the detection of endpoint on aphasia-patient's voice.
Algorithms ; Aphasia ; etiology ; rehabilitation ; Artificial Intelligence ; Endpoint Determination ; Humans ; Language Therapy ; instrumentation ; Phonetics ; Software ; Speech Intelligibility ; Speech Production Measurement ; instrumentation ; Stroke ; complications ; Stroke Rehabilitation ; Verbal Behavior