A new method incorporating deep learning with shape priors for left ventricular segmentation in myocardial perfusion SPECT images

• Published in: Computers in biology and medicine Vol. 160; p. 106954
• Main Authors: Zhu, Fubao, Li, Longxi, Zhao, Jinyu, Zhao, Chen, Tang, Shaojie, Nan, Jiaofen, Li, Yanting, Zhao, Zhongqiang, Shi, Jianzhou, Chen, Zenghong, Han, Chuang, Jiang, Zhixin, Zhou, Weihua
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.06.2023; Elsevier Limited
• Subjects: Accuracy; Algorithms; Cardiovascular disease; Cicatrix; Contours; Correlation coefficients; Datasets; Deep Learning; Dynamic programming; Epicardium; Heart; Heart Ventricles - diagnostic imaging; Heart Ventricles - pathology; Humans; Image processing; Image segmentation; Internal Medicine; Ischemia; Left ventricle; Myocardial perfusion SPECT; Myocardium; Other; Performance evaluation; Perfusion; Quantitative analysis; Retrospective Studies; Segmentation; Single photon emission computed tomography; Software; Stroke Volume; Tomography, Emission-Computed, Single-Photon - methods; Training; V-Net; Ventricle; Ventricular Function, Left; CNN; Segmentation; CAD; Myocardial perfusion SPECT; LVEF; ESV; LV; DP; STN; ROI; Left ventricle; Deep learning; MPS; DSC; EDV; V-Net; DP-ST-V-Net; HD; coronary artery disease; left ventricular ejection fraction; end-diastolic volume; spatial transform network; convolutional neural network; dynamic programming; end-systolic volume; shape-transformation VNet with dynamic programming; Hausdorff distance; region of interest; Dice similarity coefficient

Accurate segmentation of the left ventricle (LV) is crucial for evaluating myocardial perfusion SPECT (MPS) and assessing LV functions. In this study, a novel method combining deep learning with shape priors was developed and validated to extract the LV myocardium and automatically measure LV functi...