3D reconstruction of cochlea using optical coherence tomography

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Ophthalmology

Background:

• Optical coherence tomography (OCT) enables in vivo visualization of the cochlea.
• OCT imaging is crucial for diagnosing inner ear diseases, such as Meniere's disease, by quantifying endolymphatic hydrops.

Purpose of the Study:

• To present a novel 3D segmentation method for cochlear OCT images.
• To enable accurate visualization and automated analysis of cochlear structures.

Main Methods:

• A novel 3D segmentation approach utilizing superpixels and diffusion maps.
• Input: grayscale volumetric OCT images.
• Output: binary image with segmented cochlea.

Main Results:

• The proposed method successfully segments cochlear OCT images.
• The segmentation is suitable for both visualization and preprocessing for automated disease grading.

Conclusions:

• The developed superpixel and diffusion map-based method is effective for 3D segmentation of cochlear OCT data.
• This technique supports the diagnosis and automated grading of endolymphatic hydrops.