{"type":"ScholarlyArticle","title":"Learning Sparse Interpretable Features For NAS Scoring From Liver Biopsy Images","authors":[{"name":"Chong Yin","slug":"chong-yin","given":"Chong","family":"Yin","affiliation":"Department of Computer Science , Hong Kong Baptist University , Hong Kong"},{"name":"Siqi Liu","slug":"siqi-liu","given":"Siqi","family":"Liu","affiliation":"Department of Computer Science , Hong Kong Baptist University , Hong Kong"},{"name":"Vincent Wai-Sun Wong","slug":"vincent-wai-sun-wong","given":"Vincent Wai-Sun","family":"Wong","affiliation":"Department of Medicine and Therapeutics , The Chinese University of Hong Kong , Hong Kong"},{"name":"Pong C Yuen","slug":"pong-c-yuen","given":"Pong C.","family":"Yuen","affiliation":"Department of Computer Science , Hong Kong Baptist University , Hong Kong"}],"abstract":"Liver biopsy images play a key role in the diagnosis of global non-alcoholic fatty liver disease (NAFLD). The NAFLD activity score (NAS) on liver biopsy images grades the amount of histological findings that reflect the progression of NAFLD. However, liver biopsy image analysis remains a challenging task due to its complex tissue structures and sparse distribution of histological findings. In this paper, we propose a sparse interpretable feature learning method (SparseX) to efficiently estimate NAS scores. First, we introduce an interpretable spatial sampling strategy based on histological features to effectively select informative tissue \r\n regions containing tissue alterations. Then, SparseX formulates the feature learning as a low-rank decomposition problem. Non-negative matrix factorization (NMF)-based attributes learning is embedded into a deep network to compress and select sparse features for a small portion of tissue alterations contributing to diagnosis. Experiments conducted on the internal Liver-NAS and public SteatosisRaw datasets show the effectiveness of the proposed method in terms of classification performance and interpretability.  regions containing tissue alterations. Then, SparseX formulates the feature learning as a low-rank decomposition problem. Non-negative matrix factorization (NMF)-based attributes learning is embedded into a deep network to compress and select sparse features for a small portion of tissue alterations contributing to diagnosis. Experiments conducted on the internal Liver-NAS and public SteatosisRaw datasets show the effectiveness of the proposed method in terms of classification performance and interpretability.","doi":"10.24963/ijcai.2022/220","url":"https://ijcai.org/proceedings/2022/220","pdf":"https://ijcai.org/proceedings/2022/0220.pdf","markdown":"https://ijcai.org/proceedings/2022/220.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/220","volume":{"slug":"2022","title":"Proceedings of the Thirty-First International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2022","url":"https://ijcai.org/proceedings/2022/","year":2022,"location":"Vienna","dates":"23-29 July 2022","editors":"Luc De Raedt","isbn":"978-1-956792-00-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2022"},"section":"Computer Vision","track":"Main Track","number":220,"pages":{"first":1580,"last":1586},"keywords":["Computer Vision: Biomedical Image Analysis","Machine Learning: Weakly Supervised Learning","Computer Vision: Interpretability and Transparency"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38986497","length_sec":79,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38984640","length_sec":316,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Arevalo et al., 2014] John Arevalo, Angel Cruz-Roa, and FABIO A GONZ ÁLEZ O. Histopathology image repre- sentation for automatic analysis: A state-of-the-art review. Revista Med, 22(2):79-91, 2014.","title":"Histopathology image representation for automatic analysis: A state-of-the-art review","authors":"Arevalo","year":"2014","venue":"Revista Med","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Chikontwe et al., 2020] Philip Chikontwe, Meejeong Kim, Soo Jeong Nam, Heounjeong Go, and Sang Hyun Park. Multiple instance learning with center embeddings for histopathology classification. In MICCAI, pages 519-528. Springer, 2020.","title":"Multiple instance learning with center embeddings for histopathology classification","authors":"Philip Chikontwe, Meejeong Kim, Soo Jeong Nam, Heounjeong Go, Sang Hyun Park","year":"2020","venue":"MICCAI","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Cordonnier et al., 2021] Jean-Baptiste Cordonnier, Ar- avindh Mahendran, Alexey Dosovitskiy, Dirk Weis- senborn, Jakob Uszkoreit, and Thomas Unterthiner. Differentiable patch selection for image recognition. In CVPR, pages 2351-2360, 2021.","title":"Differentiable patch selection for image recognition","authors":"Jean-Baptiste Cordonnier, Aravindh Mahendran, Alexey Dosovitskiy, Dirk Weissenborn, Jakob Uszkoreit, Thomas Unterthiner","year":"2021","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Forlano et al., 2020] Roberta Forlano, Benjamin H Mullish, Nikolaos Giannakeas, James B Maurice, Napat Angkathunyakul, Josephine Lloyd, Alexandros T Tzallas, Markos Tsipouras, Michael Yee, Mark R Thursz, et al. High-throughput, machine learning-based quan- tification of steatosis, inflammation, ballooning, and fibrosis in biopsies from patients with nonalcoholic fatty liver disease. Clinical Gastroenterology and Hepatology, 18(9):2081-2090, 2020.","title":"High-throughput, machine learning-based quantification of steatosis, inflammation, ballooning, and fibrosis in biopsies from patients with nonalcoholic fatty liver disease","authors":"Roberta Forlano, Nikolaos Benjamin H Mullish, Giannakeas, Napat James B Maurice, Josephine Angkathunyakul, Alexandros T Lloyd, Markos Tzallas, Michael Tsipouras, Yee, Mark R Thursz","year":"2020","venue":"Clinical Gastroenterology and Hepatology","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Geng et al., 2021] Zhengyang Geng, Meng-Hao Guo, Hongxu Chen, Xia Li, Ke Wei, and Zhouchen Lin. Is attention better than matrix decomposition? ArXiv, abs/2109.04553, 2021.","title":"Is attention better than matrix decomposition?","authors":"Geng","year":"2021","venue":null,"doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[He et al., 2016] Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recog- nition. In CVPR, pages 770-778, 2016.","title":"Deep residual learning for image recognition","authors":"Kaiming He, Xiangyu Zhang, Shaoqing Ren, Jian Sun","year":"2016","venue":"CVPR","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Heinemann et al., 2019] Fabian Heinemann, Gerald Birk, and Birgit Stierstorfer. Deep learning enables pathologist- like scoring of nash models. Scientific reports, 9(1):1-10, 2019.","title":"Deep learning enables pathologistlike scoring of nash models","authors":"Fabian Heinemann, Gerald Birk, Birgit Stierstorfer","year":"2019","venue":"Scientific reports","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Jana et al., 2020] Ananya Jana, Hui Qu, Puru Rattan, Car- los D Minacapelli, Vinod Rustgi, and Dimitris Metaxas. Deep learning based nas score and fibrosis stage predic- tion from ct and pathology data. In BIBE, pages 981-986. IEEE, 2020.","title":"Deep learning based nas score and fibrosis stage prediction from ct and pathology data","authors":"Ananya Jana, Hui Qu, Puru Rattan, Carlos D Minacapelli, Vinod Rustgi, Dimitris Metaxas","year":"2020","venue":"BIBE","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Katharopoulos and Fleuret, 2019] Angelos Katharopoulos and Franc ¸ois Fleuret. Processing megapixel images with deep attention-sampling models. In ICML, pages 3282- 3291. PMLR, 2019.","title":"Angelos Katharopoulos and Franc ¸ois Fleuret. Processing megapixel images with deep attention-sampling models","authors":"Fleuret Katharopoulos","year":"2019","venue":"ICML","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Lee and Seung, 1999] Daniel D Lee and H Sebastian Seung. Learning the parts of objects by non-negative matrix fac- torization. Nature, 401(6755):788-791, 1999.","title":"Learning the parts of objects by non-negative matrix factorization","authors":"D Daniel, H Lee, Seung Sebastian","year":"1999","venue":"Nature","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Lee and Seung, 2000] Daniel D. Lee and H. Sebastian Se- ung. Algorithms for non-negative matrix factorization. In NIPS, 2000.","title":"Algorithms for non-negative matrix factorization","authors":"Daniel D Lee, H Sebastian Seung","year":"2000","venue":"NIPS","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Lin et al., 2021] Huapeng Lin, Xinrong Zhang, Guan- lin Li, Grace Lai-Hung Wong, and Vincent Wai-Sun Wong. Epidemiology and clinical outcomes of metabolic (dysfunction)-associated fatty liver disease. Journal of Clinical and Translational Hepatology, 9(6):972, 2021.","title":"Epidemiology and clinical outcomes of metabolic (dysfunction)-associated fatty liver disease","authors":"Lin","year":"2021","venue":"Journal of Clinical and Translational Hepatology","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Liu et al., 2019] Dongnan Liu, Donghao Zhang, Yang Song, C. Zhang, Fan Zhang, Lauren J O'Donnell, and Wei- dong (Tom) Cai. Nuclei segmentation via a deep panoptic model with semantic feature fusion. In IJCAI, 2019.","title":"Nuclei segmentation via a deep panoptic model with semantic feature fusion","authors":"Liu","year":"2019","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2019/121"},{"position":14,"raw":"[Popa et al., 2021] Stefan Lucian Popa, Abdulrahman Is- maiel, Pop Cristina, Mogosan Cristina, Giuseppe Chiar- ioni, Liliana David, and Dan Lucian Dumitrascu. Non- alcoholic fatty liver disease: Implementing complete auto- mated diagnosis and staging. a systematic review. Diag- nostics, 11, 2021.","title":"Nonalcoholic fatty liver disease: Implementing complete automated diagnosis and staging. a systematic review","authors":"Lucian Stefan, Abdulrahman Popa, Pop Ismaiel, Mogosan Cristina, Giuseppe Cristina, Liliana Chiarioni, Dan Lucian David, Dumitrascu","year":"2021","venue":"Diagnostics","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Puri and Sanyal, 2012] Puneet Puri and Arun J. Sanyal. Nonalcoholic fatty liver disease: Definitions, risk factors, and workup. Clinical Liver Disease, 1, 2012.","title":"Nonalcoholic fatty liver disease: Definitions, risk factors, and workup","authors":"Puneet Puri, Arun J Sanyal","year":"2012","venue":"Clinical Liver Disease","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Roy et al., 2020] Mousumi Roy, Fusheng Wang, Hoang Vo, Dejun Teng, George Teodoro, Alton B Farris, Eduardo Castillo-Leon, Miriam B Vos, and Jun Kong. Deep- learning-based accurate hepatic steatosis quantification for histological assessment of liver biopsies. Laboratory In- vestigation, 100(10):1367-1383, 2020.","title":"Deeplearning-based accurate hepatic steatosis quantification for histological assessment of liver biopsies","authors":"Mousumi Roy, Fusheng Wang, Hoang Vo, Dejun Teng, George Teodoro, Alton B Farris, Eduardo Castillo-Leon, Miriam B Vos, Jun Kong","year":"2020","venue":"Laboratory Investigation","doi":null,"url":null,"internal_url":null},{"position":17,"raw":"[Taylor-Weiner et al., 2021] Amaro Taylor-Weiner, Harsha Pokkalla, Ling Han, Catherine Jia, Ryan Huss, Chuhan Chung, Hunter Elliott, Benjamin Glass, Kishalve Pethia, Oscar Carrasco-Zevallos, et al. A machine learning ap- proach enables quantitative measurement of liver histology and disease monitoring in nash. Hepatology, 2021.","title":"A machine learning approach enables quantitative measurement of liver histology and disease monitoring in nash","authors":"Amaro Taylor-Weiner, Harsha Pokkalla, Ling Han, Catherine Jia, Ryan Huss, Chuhan Chung, Hunter Elliott, Benjamin Glass, Kishalve Pethia, Oscar Carrasco-Zevallos","year":"2021","venue":"Hepatology","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Vahadane et al., 2016] Abhishek Vahadane, Tingying Peng, Amit Sethi, Shadi Albarqouni, Lichao Wang, Maximilian Baust, Katja Steiger, Anna Melissa Schlitter, Irene Espos- ito, and Nassir Navab. Structure-preserving color normal- ization and sparse stain separation for histological images. IEEE transactions on medical imaging, 35(8):1962-1971, 2016.","title":"Structure-preserving color normalization and sparse stain separation for histological images","authors":"Abhishek Vahadane, Tingying Peng, Amit Sethi, Shadi Albarqouni, Lichao Wang, Maximilian Baust, Katja Steiger, Anna Melissa Schlitter, Irene Esposito, Nassir Navab","year":"2016","venue":"IEEE transactions on medical imaging","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Yin et al., 2021] Chong Yin, Siqi Liu, Rui Shao, and Pong C Yuen. Focusing on clinically interpretable fea- tures: Selective attention regularization for liver biopsy image classification. In MICCAI, pages 153-162.","title":"Focusing on clinically interpretable features: Selective attention regularization for liver biopsy image classification","authors":"Chong Yin, Siqi Liu, Rui Shao, Pong C Yuen","year":null,"venue":"MICCAI","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Zhang and Zhou, 2017] Ya-Lin Zhang and Zhi-Hua Zhou. Multi-instance learning with key instance shift. In IJCAI, pages 3441-3447, 2017.","title":"Multi-instance learning with key instance shift","authors":"Zhou ; ] Ya-Lin Zhang, Zhi-Hua Zhang, Zhou","year":"2017","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2017/481"},{"position":21,"raw":"[Zhou et al., 2018] Yanning Zhou, Qi Dou, Hao Chen, Jing Qin, and Pheng-Ann Heng. Sfcn-opi: Detection and fine- grained classification of nuclei using sibling fcn with ob- jectness prior interaction. In AAAI, 2018.","title":"Sfcn-opi: Detection and finegrained classification of nuclei using sibling fcn with objectness prior interaction","authors":"Yanning Zhou, Qi Dou, Hao Chen, Jing Qin, Pheng-Ann Heng","year":"2018","venue":"AAAI","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Zhou et al., 2021] Yu-Jie Zhou, Feng Gao, Wen-Yue Liu, Grace Lai-Hung Wong, Sanjiv Mahadeva, Nik Raihan Nik Mustapha, Xiao-Dong Wang, Wah-Kheong Chan, Vincent Wai-Sun Wong, and Ming-Hua Zheng. Screening for compensated advanced chronic liver disease using re- fined baveno vi elastography cutoffs in asian patients with nonalcoholic fatty liver disease. Alimentary pharmacology & therapeutics, 54(4):470-480, 2021.","title":"Screening for compensated advanced chronic liver disease using refined baveno vi elastography cutoffs in asian patients with nonalcoholic fatty liver disease","authors":"Yu-Jie Zhou, Feng Gao, Wen-Yue Liu, Grace Lai, -Hung Wong, Sanjiv Mahadeva, Nik Raihan, Nik Mustapha, Xiao-Dong Wang, Wah-Kheong Chan, Vincent Wai-Sun, Ming-Hua Wong, Zheng","year":"2021","venue":"Alimentary pharmacology & therapeutics","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}