Complex wavelet based quality assessment for AS-OCT images with application to Angle Closure Glaucoma diagnosis

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Angle closure disease detection relies on Anterior Segment Optical Coherence Tomography (AS-OCT).
• Anterior Chamber (AC) characteristic quantification from AS-OCT images is crucial for diagnosis.
• Current AS-OCT image quality assessment lacks objective or automated methods.

Purpose of the Study:

• To develop an objective and automated method for assessing AS-OCT image quality.
• To create a quality index for AS-OCT images to aid in medical diagnosis.
• To validate the proposed objective assessment against expert subjective grading.

Main Methods:

• Utilized complex wavelet-based local binary pattern features for objective AS-OCT image assessment.
• Employed a Naïve Bayes classifier to derive a final quality parameter.
• Conducted subjective assessments by clinical experts grading images as good, fair, or poor based on AC structure visibility.

Main Results:

• The proposed objective assessment method demonstrated the ability to differentiate between good/fair and poor quality AS-OCT images.
• Validation showed the objective assessment correlates with expert subjective grading.
• The method effectively distinguishes images suitable for glaucoma diagnosis.

Conclusions:

• An automated algorithm for evaluating AS-OCT image quality has been developed.
• The proposed quality index provides objective, quantitative assessment, aligning with glaucoma specialist evaluations.
• This tool facilitates the collection of high-quality data for improved medical diagnosis.