This research introduces an intelligent throat (IT) system that captures muscle vibrations and pulse signals, processing them through large language models to enable natural speech for patients with speech impairments.

• Integrates ultrasensitive textile strain sensors and carotid pulse monitoring
• Processes throat muscle vibrations through LLMs to generate coherent, emotionally expressive speech
• Demonstrates clinical efficacy for stroke patients with dysarthria
• Offers potential applications across various neurological conditions

This breakthrough technology represents a significant advancement in medical assistive devices, potentially restoring communication abilities and improving quality of life for millions of patients with speech disorders worldwide.

Wearable intelligent throat enables natural speech in stroke patients with dysarthria