Application of an active contour model for tracking vocal folds during phonation in videostroboscopic sequences
PBN-AR
Instytucja
Wydział Elektrotechniki, Elektroniki, Informatyki i Automatyki (Politechnika Łódzka)
Informacje podstawowe
Główny język publikacji
EN
Czasopismo
Polski Przegląd Otorynolaryngologiczny
ISSN
2084-5308
EISSN
Wydawca
Index Copernicus International
DOI
Rok publikacji
2016
Numer zeszytu
Strony od-do
42-43
Numer tomu
5
Identyfikator DOI
Liczba arkuszy
Autorzy
Pozostali autorzy
+ 1
Słowa kluczowe
EN
glottal pathologies
glottovibrograms
videolaryngostroboscopy
voice disorders
Streszczenia
Język
EN
Treść
Purpose: Videolaryngostroboscopic images can be used for both qualitative assessment and quantitative description of the vocal folds health status. In particular, a reliable quantitative description is essential for early diagnosis of voice insufficiency and objectification of the phoniatric examination of the larynx. The aim of the study is to apply image processing techniques for assessing the parameters of the vocal folds phonation process using videolaryngostroboscopic examination method. Method: The first stage of the study is focused on implementing image processing methods, for designating the region of interest (ROI), i.e. detecting the space placed directly between the vocal folds in the video. Based on the detected ROI the active contour model allows to track and register the movements of the vocal folds. We propose a new method based on the magnitude of the image gradient that allows to estimate the movement of the edges of the vocal folds. Results: Active contour based image segmentation allows to convert the video into a spatio-temporal image called the glottovibrogram. Such a representation enables extraction of parameters providing a quantitative description of vocal folds kinematics during phonation i.e.: Open Quotient (OQ), Closing Quotient (CQ), Minimal Relative Glottal Area (MRGA) or the Closure Difference Index (CDI). The voice assessment carried out by the described image analysis techniques revealed that in the group of patients with incomplete glottal closure the evaluated parameters assumed different values than in the control group of normophonic subjects.
Inne
System-identifier
90664