Introduction The human voice is one of the most complex phenomena in nature. Prom the medical point of view, the human voice is a collection of information about the patient's health (way of speaking, intonation, voice timbre, etc.). For an otorhinolaryngologist, it is primarily a somce of information on the voice-forming organ of the larynx. Objective of the work The aim of the study was to try to answer a mun ber of questions regarding the presence of pathology in the larynx and the classification of this pathology based only on the analysis of a voice sample. Method In the conditions of an anechoic chamber, voice samples were recorded in accordance with the established research protocol (properly selected sets of phones and sentences). Then, the obtained samples were subjected to multi-paran1eter analysis - MDVP, and the results obtained in this way were subjected to statistical analysis. Study group The study group consisted of 117 patients aged 22 to 79 who selfreported or were referred by a primary care physician to an otolaryngology clinic due to voice disorders. Results and conclusions The obtained results allowed to determine the possibility of detecting laryngeal pathology in a patient only on the basis of a multiparameter analysis of his voice recording. Sets of features were also found that allow to assign the results obtained in this way t ; o the appropriate categories of pathological conditions of the voice organ. It was not possible to obtain a narrow group of features allowing for such an analysis in the case of voice samples recorded on generally available sound recording devices.
Rada Dyscypliny Nauki medyczne
Aug 27, 2024
Jun 28, 2024
12
0
http://dl.cm-uj.krakow.pl:8080/publication/5139
Edition name | Date |
---|---|
ZB-139916 | Aug 27, 2024 |
Gasiński, Mateusz
Kurzyński, Marian
Sobolewska, Aleksandra