Related Experiment Video
Updated: Jun 22, 2026
Simultaneous Brightfield, Fluorescence, and Optical Coherence Tomographic Imaging of Contracting Cardiac Trabeculae Ex Vivo
Published on: October 2, 2021
Automated segmentation of tissue structures in optical coherence tomography data.
Fernando Gasca1, Lukas Ramrath, Gereon Huettmann
1University at Luebeck, Graduate School for Computing in Medicine and Life Sciences, Institute for Robotics and Cognitive Systems, Ratzeburger Alle 160, Lubeck 23538, Germany. gasca@rob.uni-luebeck.de
Two novel automated methods accurately segment arbitrary structures in optical coherence tomography (OCT) images, overcoming speckle noise challenges for improved medical imaging. These user-independent techniques offer promising results for OCT applications.
Area of Science:
- Medical Imaging
- Biomedical Engineering
- Image Processing
Background:
- Optical coherence tomography (OCT) is vital for medical imaging, but speckle noise complicates structure identification.
- Automated segmentation of OCT images is crucial for extracting meaningful information.
Purpose of the Study:
- To develop and evaluate two user-independent automated methods for segmenting arbitrary structures in OCT images.
- To address the challenge of speckle noise in OCT image analysis.
Main Methods:
- Two seeded region growing algorithms were developed for automated OCT image segmentation.
- Method 1: Adaptive neighborhood homogeneity criterion on unfiltered OCT images, incorporating tissue intensity and speckle noise models.
- Method 2: Region growing on filtered OCT images using local median for homogeneity assessment.
Main Results:
- Both methods demonstrated capabilities in detecting structures and minimizing leakage on artificial data.
- Quantitative evaluation showed the effectiveness of the proposed segmentation approaches.
- Real-world OCT data tested in various scenarios yielded promising results.
Conclusions:
- The proposed automated segmentation methods effectively overcome speckle noise in OCT images.
- These techniques offer reliable, user-independent solutions for OCT image analysis in medical applications.
- The methods show potential for widespread application in OCT imaging.
