Classification of Pharynx from MRI Using a Visual Analysis Tool to Study Obstructive Sleep Apnea

Area of Science:

• Medical Imaging
• Computer Vision
• Sleep Medicine

Background:

• Obstructive sleep apnea (OSA) is a prevalent chronic sleep disorder.
• Understanding pharyngeal anatomy is key to OSA pathogenesis.
• Pharyngeal classification is essential for OSA analysis.

Purpose of the Study:

• To develop an automated classifier for pharyngeal segmentation from MRI data.
• To improve the efficiency and accuracy of pharyngeal analysis in OSA research.

Main Methods:

• A visual analysis-based classification pipeline was developed.
• Stages include pre-processing, feature extraction (categorical and numerical), and supervised classification.
• The classifier was trained using visual analytics and silhouette coefficient.

Main Results:

• The automated classifier achieved an 86% Jaccard coefficient for pharyngeal classification on MRI datasets.
• Expert knowledge can optimize feature selection and weighting during classifier training.

Conclusions:

• The proposed classifier is accurate and computationally efficient.
• It offers insights into individual and collective feature influences.
• Applications include large-scale epidemiological studies for OSA analysis.