MUMBAI, India, Nov. 2 -- Intellectual Property India has published a patent application (202641125858 A) filed by Srm Trp Engineering College on October 22, 2026, for Development Of Voice Assistive System For Tracheostomy And Speech-Impaired Patients Using M-Rls Algorithm.
Inventors include Nivethitha G; Premeela S; Rajalakshmi C; and Martin Sagayam K.
The application for the patent was published on October 30, 2026, under issue no. 44/2026.
Abstract: A voice assistive system and method are disclosed for enhancing the intelligibility of speech produced by tracheostomy and speech-impaired patients using assistive voicing devices such as an electrolarynx or a tracheo-oesophageal speaking valve. The system comprises a signal acquisition unit, a signal-conditioning front end, an embedded processing core implementing a Modified Recursive Least Squares (M-RLS) adaptive filtering algorithm, and an output rendering unit. The M-RLS algorithm employs a variable forgetting factor and a regularised covariance update to rapidly track and suppress the quasi-periodic mechanical excitation noise characteristic of such devices, while substantially preserving the underlying speech formant structure. The processing core is suitable for implementation on a low- power application-specific integrated circuit or field-programmable gate array, enabling real-time, continuous operation on a compact, wearable, or bedside assistive device. Numerical simulation of a representative speech segment corrupted by mechanical excitation noise demonstrates a 78.6% reduction in root-mean-square signal error and a 13.4 dB improvement in signal-to-noise ratio for the disclosed M-RLS algorithm relative to the unprocessed signal, validating the invention's core mechanism ahead of physical hardware deployment and evidencing its potential to improve communication and quality of life for tracheostomy and speech-impaired patients.
Disclaimer: Curated by HT Syndication.
