Automatic Coregistration Between Coronary Angiography and Intravascular Optical Coherence Tomography
Hui Qin, MSc,
Chunming Li, BSc,
Yingguang Li, PhD,
Jiayue Huang, MSc,
Fan Yang, BSc,
Takashi Kubo, MD, PhD,
Takashi Akasaka, MD, PhD,
Changyan Xiao, PhD,
Juan Luis Gutiérrez-Chico, MD, PhD,
Shengxian Tu, PhD
Affiliations
Hui Qin, MSc
College of Electrical and Information Engineering, Hunan University, Changsha, China; Biomedical Instrument Institute, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Chunming Li, BSc
Biomedical Instrument Institute, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Yingguang Li, PhD
Kunshan Industrial Technology Research Institute Co Ltd, Kunshan, China
Jiayue Huang, MSc
Biomedical Instrument Institute, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Fan Yang, BSc
Biomedical Instrument Institute, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China
Takashi Kubo, MD, PhD
Department of Cardiovascular Medicine, Wakayama Medical University, Wakayama, Japan
Takashi Akasaka, MD, PhD
Department of Cardiovascular Medicine, Wakayama Medical University, Wakayama, Japan
Changyan Xiao, PhD
College of Electrical and Information Engineering, Hunan University, Changsha, China
Juan Luis Gutiérrez-Chico, MD, PhD
Cardiology Department, Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China
Shengxian Tu, PhD
Biomedical Instrument Institute, School of Biomedical Engineering, Shanghai Jiao Tong University, Shanghai, China; Address for correspondence: Dr Shengxian Tu, Room123, Med-X Research Institute, Shanghai Jiao Tong University, No. 1954, Hua Shan Road, Xuhui District, Shanghai 200030, China.
This study sought to evaluate a novel approach for automatic coregistration of optical coherence tomography (OCT) and coronary angiography. Lumen diameters and side branches from both coronary angiography and OCT were used to create 2 feature sets. Subsequently, a 2-step coregistration approach was performed on the feature sets for matching of each OCT cross section on the angiographic centerline. For validation, all side branches with ≥1.0 mm diameter were identified and used as paired fiduciary landmarks. Geographical error was defined as the distance between the automatically coregistered and the true-paired landmarks. Altogether 212 vessels from 181 patients were analyzed. Mismatch of coronary angiography and OCT occurred in 64 of 1,530 reference landmarks. Median geographical error was 0.32 (interquartile range: 0.00-0.56) mm. The mean time for coregistration was 20.69 ± 1.07 seconds. In conclusion, fast and automatic coregistration of OCT and angiography using a single standard angiographic loop is feasible and accurate.