Modern possibilities for the diagnosis of voice disorders using computer acoustic analysis of voice and artificial intelligence

OBJECTIVE: The purpose of the review is to analyze modern approaches to the diagnosis of voice disorders using AI and acoustic analysis, as well as to evaluate their effectiveness and prospects for implementation in clinical practice. RESEARCH DESIGN: A review of domestic and foreign literature sources on the use of computer acoustic analysis and AI technologies in the diagnosis of diseases of the vocal apparatus is carried out. MATERIAL AND METHODS: The material for the analysis was publications in scientific journals, the results of clinical trials, data on the use of specialized software systems (PRAAT, MDVP, LingWaves) and machine learning-based systems (SVM, CNN, RNN and their combinations). Data on diagnostic accuracy, interpretability of results, and applicability in real clinical practice were considered. RESULTS: Acoustic analysis of the voice makes it possible to detect pathological changes in the early stages, before the appearance of pronounced symptoms. The PRAAT and MDVP programs have become the standard in assessing the spectral and temporal characteristics of a voice. The introduction of AI has significantly improved diagnostic accuracy (up to 95-97% in individual studies). Machine learning algorithms and deep neural networks automate diagnostics, allow you to identify hidden patterns and use data obtained even from mobile devices. However, there are still unresolved issues of database standardization, interpretability of models, and ethical and legal aspects of their implementation. Despite the high efficiency of technologies, their use requires the development of uniform standards, ensuring the transparency of algorithms and multidisciplinary interaction of specialists.