A computer-aid speech rehabilitation system with mirrored video generating.

Area of Science:

• Neuroscience
• Computer Science
• Speech Therapy

Background:

• Aphasia is a language disorder significantly impacting individuals' lives.
• Computer-aided rehabilitation offers effective solutions for aphasia patients.

Purpose of the Study:

• To develop an effective speech rehabilitation software for aphasia patients.
• To integrate mirrored therapy into a computer-aided system.

Main Methods:

• The system utilizes patient facial photos for mirrored video generation and speech synthesis.
• Machine learning algorithms assess speech similarity for progress evaluation.
• Task-oriented training modules include vowel, word, and sentence exercises.

Main Results:

• The system achieved average scores of 83.9% for vowel exercises, 74.3% for word exercises, and 77.8% for sentence training.
• Real-time performance evaluation was conducted for all tasks.
• The mirrored therapy approach demonstrated engaging and motivating visual feedback.

Conclusions:

• The developed intelligent mirror software system shows promise for reading-based aphasia rehabilitation.
• The user-friendly application facilitates daily training tasks.
• This system offers a novel approach to speech therapy for aphasia.