Spatio-temporal multi-task network cascade for accurate assessment of cardiac CT perfusion

• Published in: Medical image analysis Vol. 74; p. 102207
• Main Authors: Chen, Jiaqi, Zhang, Pengfei, Liu, Huafeng, Xu, Lei, Zhang, Heye
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.12.2021; Elsevier BV
• Subjects: Algorithms; Automation; Cardiovascular disease; Coders; Computed tomography; Coronary artery; Coronary artery disease; Encoders-Decoders; Heart; Heart diseases; Humans; Ischemia; Multitask network cascade; Myocardial perfusion; Parameters; Perfusion; Representations; Spatiotemporal representation; Tomography, X-Ray Computed; Spatiotemporal representation; Myocardial perfusion; Multitask network cascade
• ISSN: 1361-8415; 1361-8423; 1361-8423
• DOI: 10.1016/j.media.2021.102207

•We provide a quantitative assessment of myocardial CT perfusion to achieve predictions of perfusion parameters and ischemic regions.•A spatio-temporal encoder-decoder architecture based on pseudo-3D convolutions is proposed to capture patch-based spatio-temporal representations.•We introduce a cascaded structure into multi-task learning, which improves the feature learning of different tasks.•We present the first work to leverage deep learning techniques in myocardial perfusion assessment to simultaneously obtain perfusion parameters and ischemic regions. [Display omitted] The assessment of myocardial perfusion has become increasingly important in the early diagnosis of coronary artery disease. Currently, the process of perfusion assessment is time-consuming and subjective. Although automated methods by threshold processing have been proposed, they cannot obtain an accurate perfusion assessment. Thus, there is a great clinical demand to obtain a rapid and accurate assessment of myocardial perfusion through a standard procedure using an automated algorithm. In this work, we present a spatio-temporal multi-task network cascade (ST-MNC) to provide an accurate and robust assessment of myocardial perfusion. The proposed network captures patch-based spatio-temporal representations for each pixel through a spatio-temporal encoder-decoder network. Then the multi-task network cascade uses spatio-temporal representations as shared features to predict various perfusion parameters and myocardial ischemic regions. Extensive experiments on CT images of 232 subjects demonstrate ST-MNC could produce a good approximation for perfusion parameters and an accurate classification for ischemic regions. These results show that our proposed method can provide a fast and accurate assessment of myocardial perfusion.