Image-based Nystagmus Analysis in BPPV Disorders: Polar Coordinate System and Template Matching

• Published in: 2024 IEEE 3rd International Conference on Computing and Machine Intelligence (ICMI) pp. 1 - 5
• Main Authors: Dogru, Hasibe Busra, Kaplan, Ahmet, Zengin, Aydin Tarik, Ozkul, Tarik
• Format: Conference Proceeding
• Language: English
• Published: IEEE 13-04-2024
• Subjects: Accuracy; Analytical models; bppv; Computational modeling; Deep learning; Ear; Feature extraction; nystagmus; Object recognition; polar coordinate system; template matching
• DOI: 10.1109/ICMI60790.2024.10585886

Benign Paroxysmal Positional Vertigo (BPPV) is an inner ear disorder characterized by dizziness and a sense of imbalance, resulting from a disturbance in the balance center located in the inner ear, often due to dislodged otoliths. This study focuses on the challenges associated with detecting horizontal, vertical, and torsional nystagmus in BPPV patients. To detect horizontal and vertical movements, a method involving conversion to a polar coordinate system is used. Torsional movements are more complex compared to other movement components, making their accurate detection difficult. To overcome this difficulty, the effectiveness of the template matching method in the analysis of torsional nystagmus is being evaluated. The proposed feature extraction methods, aim to overcome the challenges encountered in the diagnosis and treatment of BPPV.