Longitudinal Non-invasive Analysis of VOCAle Cord Function Based on Trans-laryngeal Ultrasound Acquisitions and Voice Recordings

Overview

The main objective of VOCALISE study is to propose a new approach allowing a better characterization of postoperative dysphonia. This involves associating with dynamic translaryngeal ultrasound optimized acquisitions of the vibration of each vocal fold in phonation simultaneously with voice recordings. A software program to analyze the displacement of arytenoids, markers of substitution of the vocal cords, will be developed to finely quantify the mobility of laryngeal structures, by combining classical methods of motion analysis and deep learning methods. This approach will be evaluated to follow speech therapy rehabilitation in patients with post-operative dysphonia following recurrent nerve injury.

Dynamic translaryngeal ultrasound (dTLUS), a non-invasive and inexpensive technique, has emerged in recent years as an alternative to nasofibroscopy for assessing vocal cord paralysis. This paralysis is the major risk (3 to 5%) associated with cervical surgery (100,000 procedures per year in France). Initial work by our consortium has demonstrated the performance of dTLUS after thyroid or parathyroid surgery in the early diagnosis of vocal cord paralysis. The aim of VOCALISE is to propose a new approach for better characterisation of post-operative dysphonia. This involves combining optimised dTLUS acquisitions with acquisitions of the vibration of each vocal cord during phonation, simultaneously with voice recordings. Software will be developed to analyse the displacement of the arytenoids, which are surrogate markers for the vocal cords, in order to quantify the mobility of laryngeal structures in fine detail, using a combination of conventional motion analysis methods and deep learning methods. This approach will be evaluated to monitor speech therapy rehabilitation in patients with post-operative dysphonia following a lesion of the recurrent nerve.