Automatic Side Branch Detection in Optical Coherence Tomography Images Using Adjacent Frame Correlation Information

Chenxi Huang,Sizhe Wan,Jinkun Yang,Yue Li,Ning Wang,Tiancheng Gai,Yaqiong Hu,Yongtao Hao
DOI: https://doi.org/10.1166/jmihi.2018.2464
2018-01-01
Journal of Medical Imaging and Health Informatics
Abstract:Side branch detection is particularly important in clinical applications such as surgical planning, quantification of side branch size, and hemodynamics. Intravascular optical coherence tomography (IVOCT) is a novel imaging modality based on high axial resolution (10-16 mu m) and a better signal-to-noise ratio, which provides greater detail of the vessel lumen. Traditionally, physicians spend a significant amount of time manually marking side branches; thus, in the present study, we demonstrated an automatic side branch detection method using adjacent frame correlation. This method was divided into distance computing of a single frame, gradient computing of the width of the candidate region in the adjacent frame, and quantitative analysis of the side branch. Nineteen IVOCT pullback runs including baseline and follow-up datasets were used to validate the algorithm. Experiments determined that the method is robust and effective; detecting special circumstances including side branch and guide wire shadow overlap. Compared with manual segmentation, the best result reached an F-measure score of 87.3% when the gradient threshold was set to 6.
What problem does this paper attempt to address?