{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/direct-multitype-cardiac-indices-estimation","title":"Direct Multitype Cardiac Indices Estimation via Joint Representation and Regression Learning","arxiv_id":"1705.09307","date":"2017-05-25","proceeding":null,"authors":["Wufeng Xue","Ali Islam","Mousumi Bhaduri","Shuo Li"],"abstract":"Cardiac indices estimation is of great importance during identification and\ndiagnosis of cardiac disease in clinical routine. However, estimation of\nmultitype cardiac indices with consistently reliable and high accuracy is still\na great challenge due to the high variability of cardiac structures and\ncomplexity of temporal dynamics in cardiac MR sequences. While efforts have\nbeen devoted into cardiac volumes estimation through feature engineering\nfollowed by a independent regression model, these methods suffer from the\nvulnerable feature representation and incompatible regression model. In this\npaper, we propose a semi-automated method for multitype cardiac indices\nestimation. After manual labelling of two landmarks for ROI cropping, an\nintegrated deep neural network Indices-Net is designed to jointly learn the\nrepresentation and regression models. It comprises two tightly-coupled\nnetworks: a deep convolution autoencoder (DCAE) for cardiac image\nrepresentation, and a multiple output convolution neural network (CNN) for\nindices regression. Joint learning of the two networks effectively enhances the\nexpressiveness of image representation with respect to cardiac indices, and the\ncompatibility between image representation and indices regression, thus leading\nto accurate and reliable estimations for all the cardiac indices.\n  When applied with five-fold cross validation on MR images of 145 subjects,\nIndices-Net achieves consistently low estimation error for LV wall thicknesses\n(1.44$\\pm$0.71mm) and areas of cavity and myocardium (204$\\pm$133mm$^2$). It\noutperforms, with significant error reductions, segmentation method (55.1% and\n17.4%) and two-phase direct volume-only methods (12.7% and 14.6%) for wall\nthicknesses and areas, respectively. These advantages endow the proposed method\na great potential in clinical cardiac function assessment.","url_abs":"http://arxiv.org/abs/1705.09307v1","url_pdf":"http://arxiv.org/pdf/1705.09307v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"direct-multitype-cardiac-indices-estimation","repo_url":"https://github.com/alejandrodebus/Indices-Net","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"direct-multitype-cardiac-indices-estimation","repo_url":"https://github.com/alejandrodebus/IndicesNet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"feature-engineering","task_name":"Feature Engineering"},{"task_slug":"regression-1","task_name":"regression"}],"methods":[{"method_slug":"convolution","method_name":"Convolution"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}