Order preserving and shape prior constrained intra-retinal layer segmentation in optical coherence tomography

Area of Science:

• Ophthalmology
• Medical Imaging
• Computer Vision

Background:

• Optical Coherence Tomography (OCT) is crucial for visualizing retinal layers.
• Accurate segmentation of OCT scans is essential for diagnosing and monitoring retinal diseases.
• Existing segmentation methods face challenges with image artifacts.

Purpose of the Study:

• To develop a robust probabilistic method for segmenting retinal tissue in OCT scans.
• To preserve the physiological order of intra-retinal layers during segmentation.
• To improve the accuracy and reliability of OCT scan analysis.

Main Methods:

• Utilizing a probabilistic approach combining discrete exact inference and a global shape prior.
• Implementing computations that allow for significant parallel processing.
• Evaluating segmentation performance against various OCT scan artifacts.

Main Results:

• Achieved accurate segmentation of intra-retinal layers, maintaining physiological order.
• Demonstrated robustness against speckle noise and shadowing from blood vessels.
• The method proved effective in handling common scan artifacts.

Conclusions:

• The proposed probabilistic method offers accurate and reliable retinal layer segmentation from OCT scans.
• The approach is resilient to common artifacts, enhancing clinical applicability.
• Parallelizable computations contribute to efficient processing of OCT data.