Adaptive Recursive Self-feedback for Speech Production

Tech ID: 24T047

Advantages

  • Enables self-guided speech therapy with no need for a clinician
  • Uses only the patient’s own voice for personalized feedback
  • Adapts task difficulty automatically based on performance
  • Works on smartphones for flexible, at-home use
  • Showed improved speech fluency in early clinical testing

Summary

Millions of people suffer from speech disorders caused by strokes or other neurological injuries. Traditional speech therapy requires frequent sessions with a trained clinician, which can be costly, time-consuming, and inaccessible for many patients.

This technology offers a self-guided method for improving speech, especially for individuals with nonfluent aphasia. It uses a patient’s own recorded voice as real-time feedback, helping them practice and improve their speech over time. The system automatically adjusts the difficulty of each speech task based on how the patient is performing, making the experience personalized and effective. It can be run entirely on a smartphone, allowing for flexible, at-home therapy without needing a therapist.

The image is a mobile app interface showing the key difference between recursive self-feedback training and standard control training. In the self-feedback version, users record and immediately listen to their speech, allowing them to self-correct and improve through playback loops.

Desired Partnerships

  • License
  • Sponsored Research
  • Co-Development

Technology Transfer
TTOinfo@usf.edu
(813) 974-0994

Patents