{"type":"ScholarlyArticle","title":"“My nose is running.” “Are you also coughing?”: Building A Medical Diagnosis Agent with Interpretable Inquiry Logics","authors":[{"name":"Wenge Liu","slug":"wenge-liu","given":"Wenge","family":"Liu","affiliation":"sen University"},{"name":"Yi Cheng","slug":"yi-cheng","given":"Yi","family":"Cheng","affiliation":"Hong Kong Polytechnic University"},{"name":"Hao Wang","slug":"hao-wang","given":"Hao","family":"Wang","affiliation":"Rutgers University"},{"name":"Jianheng Tang","slug":"jianheng-tang","given":"Jianheng","family":"Tang","affiliation":"Hong Kong University of Science and Technology"},{"name":"Yafei Liu","slug":"yafei-liu","given":"Yafei","family":"Liu","affiliation":"sen University"},{"name":"Ruihui Zhao","slug":"ruihui-zhao","given":"Ruihui","family":"Zhao","affiliation":"Tencent Jarvis Lab"},{"name":"Wenjie Li","slug":"wenjie-li","given":"Wenjie","family":"Li","affiliation":"Hong Kong Polytechnic University"},{"name":"Yefeng Zheng","slug":"yefeng-zheng","given":"Yefeng","family":"Zheng","affiliation":"Tencent Jarvis Lab"},{"name":"Xiaodan Liang","slug":"xiaodan-liang","given":"Xiaodan","family":"Liang","affiliation":"sen University"}],"abstract":"With the rise of telemedicine, the task of developing Dialogue Systems for Medical Diagnosis (DSMD) has received much attention in recent years. Different from early researches that needed to rely on extra human resources and expertise to build the system, recent researches focused on how to build DSMD in a data-driven manner. However, the previous data-driven DSMD methods largely overlooked the system interpretability, which is critical for a medical application, and they also suffered from the data sparsity issue at the same time. In this paper, we explore how to bring interpretability to data-driven DSMD. Specifically, we propose a more interpretable decision process to implement the dialogue manager of DSMD by reasonably mimicking real doctors' inquiry logics, and we devise a model with highly transparent components to conduct the inference. Moreover, we collect a new DSMD dataset, which has a much larger scale, more diverse patterns, and is of higher quality than the existing ones. The experiments show that our method obtains 7.7%, 10.0%, 3.0% absolute improvement in diagnosis accuracy respectively on three datasets, demonstrating the effectiveness of its rational decision process and model design. Our codes and the GMD-12 dataset are available at https://github.com/lwgkzl/BR-Agent.","doi":"10.24963/ijcai.2022/592","url":"https://ijcai.org/proceedings/2022/592","pdf":"https://ijcai.org/proceedings/2022/0592.pdf","markdown":"https://ijcai.org/proceedings/2022/592.md","bibtex":"https://ijcai.org/proceedings/2022/bibtex/592","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":"Natural Language Processing","track":"Main Track","number":592,"pages":{"first":4266,"last":4272},"keywords":["Natural Language Processing: Applications","Multidisciplinary Topics and Applications: Health and Medicine","Natural Language Processing: Dialogue and Interactive Systems"],"videos":[{"label":"Video #1","url":"https://slideslive.com/38986525","length_sec":88,"provider":"slideslive"},{"label":"Video #2","url":"https://slideslive.com/38984668","length_sec":362,"provider":"slideslive"}],"references":[{"position":1,"raw":"[Chen et al., 2020] Jun Chen, Xiaoya Dai, Quan Yuan, Chao Lu, and Haifeng Huang. Towards interpretable clinical diagnosis with bayesian network ensembles stacked on entity-aware cnns. In Proc. of ACL, pages 3143-3153, 2020.","title":"Towards interpretable clinical diagnosis with bayesian network ensembles stacked on entity-aware cnns","authors":"Chen","year":"2020","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Flores et al., 2011] M Julia Flores, Ann E Nicholson, An- drew Brunskill, Kevin B Korb, and Steven Mascaro. Incor- porating expert knowledge when learning bayesian network structure: a medical case study. Artificial intelligence in medicine, 53(3):181-204, 2011.","title":"Incorporating expert knowledge when learning bayesian network structure: a medical case study","authors":"Julia Flores, Ann E Nicholson, Andrew Brunskill, Kevin B Korb, Steven Mascaro","year":"2011","venue":"Artificial intelligence in medicine","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[He et al., 2020] Weijie He, Xiaohao Mao, Chao Ma, José Miguel Hernández-Lobato, and Ting Chen. FIT: A fast and accurate framework for solving medical inquiring and diagnosing tasks. arXiv preprint, arXiv:2012.01065, 2020.","title":"FIT: A fast and accurate framework for solving medical inquiring and diagnosing tasks","authors":"Weijie He, Xiaohao Mao, Chao Ma, José Miguel Hernández-Lobato, Ting Chen","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":4,"raw":"[Kahn Jr et al., 1997] Charles E Kahn Jr, Linda M Roberts, Katherine A Shaffer, and Peter Haddawy. Construction of a bayesian network for mammographic diagnosis of breast cancer. Computers in biology and medicine, 27(1):19-29, 1997.","title":"Construction of a bayesian network for mammographic diagnosis of breast cancer","authors":"Linda M Charles E Kahn Jr, Katherine A Roberts, Peter Shaffer, Haddawy","year":"1997","venue":"Computers in biology and medicine","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Kim et al., 2021] Hyunwoo Kim, Byeongchang Kim, and Gunhee Kim. Perspective-taking and pragmatics for gener- ating empathetic responses focused on emotion causes. In Proc. of EMNLP, pages 2227-2240, 2021.","title":"Perspective-taking and pragmatics for generating empathetic responses focused on emotion causes","authors":"Hyunwoo Kim, Byeongchang Kim, Gunhee Kim","year":"2021","venue":"Proc. of EMNLP","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Kipf and Welling, 2017] Thomas N. Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. In Proc. of ICLR, 2017.","title":"Semi-supervised classification with graph convolutional networks","authors":"Thomas N Kipf, Max Welling","year":"2017","venue":"Proc. of ICLR","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Lei et al., 2018] Wenqiang Lei, Xisen Jin, Min-Yen Kan, Zhaochun Ren, Xiangnan He, and Dawei Yin. Sequic- ity: Simplifying task-oriented dialogue systems with single sequence-to-sequence architectures. In Proc. of ACL, pages 1437-1447, July 2018.","title":"Sequicity: Simplifying task-oriented dialogue systems with single sequence-to-sequence architectures","authors":"Lei","year":"2018","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Liao et al., 2020] Kangenbei Liao, Qianlong Liu, Zhongyu Wei, Baolin Peng, Qin Chen, Weijian Sun, and Xuan- jing Huang. Task-oriented dialogue system for automatic disease diagnosis via hierarchical reinforcement learning. arXiv preprint, arXiv:2004.14254, 2020.","title":"Task-oriented dialogue system for automatic disease diagnosis via hierarchical reinforcement learning","authors":"Kangenbei Liao, Qianlong Liu, Zhongyu Wei, Baolin Peng, Qin Chen, Weijian Sun, Xuanjing Huang","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Lin et al., 2021] Shuai Lin, Pan Zhou, Xiaodan Liang, Jian- heng Tang, Ruihui Zhao, Ziliang Chen, and Liang Lin. Graph-evolving meta-learning for low-resource medical di- alogue generation. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 35, pages 13362-13370, 2021.","title":"Graph-evolving meta-learning for low-resource medical dialogue generation","authors":"Shuai Lin, Pan Zhou, Xiaodan Liang, Jianheng Tang, Ruihui Zhao, Ziliang Chen, Liang Lin","year":"2021","venue":"Proceedings of the AAAI Conference on Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Lincoln and Parker, 1967] Thomas L Lincoln and Rodger D Parker. Medical diagnosis using bayes theorem. Health Services Research, 2(1):34, 1967.","title":"Medical diagnosis using bayes theorem","authors":"L Thomas, Rodger D Lincoln, Parker","year":"1967","venue":"Health Services Research","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Liu et al., 2020] Wenge Liu, Jianheng Tang, Jinghui Qin, Lin Xu, Zhen Li, and Xiaodan Liang. Meddg: A large- scale medical consultation dataset for building medical dialogue system. arXiv preprint arXiv:2010.07497, 2020.","title":"Meddg: A largescale medical consultation dataset for building medical dialogue system","authors":"Wenge Liu, Jianheng Tang, Jinghui Qin, Lin Xu, Zhen Li, Xiaodan Liang","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Liu et al., 2021] Wenge Liu, Jianheng Tang, Xiaodan Liang, and Qingling Cai. Heterogeneous graph reasoning for knowledge-grounded medical dialogue system. Neurocom- puting, 442:260-268, 2021.","title":"Heterogeneous graph reasoning for knowledge-grounded medical dialogue system","authors":"Liu","year":"2021","venue":"Neurocomputing","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[McGeachie et al., 2009] Michael McGeachie, Rachel L Badovinac Ramoni, Josyf C Mychaleckyj, Karen L Furie, Jonathan M Dreyfuss, Yongmei Liu, David Herrington, Xiuqing Guo, João A Lima, Wendy Post, et al. An integrative predictive model of coronary artery calcification in arteriosclerosis. Circulation, 120(24):2448, 2009.","title":"An integrative predictive model of coronary artery calcification in arteriosclerosis","authors":"Michael Mcgeachie, Rachel L Badovinac Ramoni, C Josyf, Karen L Mychaleckyj, Jonathan M Furie, Yongmei Dreyfuss, David Liu, Xiuqing Herrington, Guo, A João, Wendy Lima, Post","year":"2009","venue":"Circulation","doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Milward and Beveridge, 2003] David Milward and Martin Beveridge. Ontology-based dialogue systems. In Proc. of IJCAI, pages 9-18. Citeseer, 2003.","title":"Ontology-based dialogue systems","authors":"Beveridge Milward, David Milward, Martin Beveridge","year":"2003","venue":"Proc. of IJCAI","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Mnih et al., 2016] Volodymyr Mnih, Adrià Puigdomènech Badia, Mehdi Mirza, Alex Graves, Timothy P. Lillicrap, Tim Harley, David Silver, and Koray Kavukcuoglu. Asyn- chronous methods for deep reinforcement learning. In Proc. of ICML, volume 48, pages 1928-1937, 2016.","title":"Asynchronous methods for deep reinforcement learning","authors":"Volodymyr Mnih, Adrià Puigdomènech Badia, Mehdi Mirza, Alex Graves, Timothy P Lillicrap, Tim Harley, David Silver, Koray Kavukcuoglu","year":"2016","venue":"Proc. of ICML","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Pople et al., 1975] Harry E. Pople, Jack D. Myers, and Ran- dolph A. Miller. DIALOG: A model of diagnostic logic for internal medicine. In Proc. of IJCAI, pages 848-855, 1975.","title":"DIALOG: A model of diagnostic logic for internal medicine","authors":"Harry E Pople, Jack D Myers, Randolph A Miller","year":"1975","venue":"Proc. of IJCAI","doi":null,"url":null,"internal_url":"/proceedings/1975/134"},{"position":17,"raw":"[Shortliffe, 1974] Edward H. Shortliffe. MYCIN: A rule- based computer program for advising physicians regarding antimicrobial therapy selection. In Roger C. Brown and Donald E. Glaze, editors, Proc. of the ACM Annual Confer- ence, page 739. ACM, 1974.","title":"MYCIN: A rulebased computer program for advising physicians regarding antimicrobial therapy selection","authors":"H Edward, Shortliffe","year":"1974","venue":"Proc. of the ACM Annual Conference","doi":null,"url":null,"internal_url":null},{"position":18,"raw":"[Wang et al., 1999] Xiao-Hui Wang, Bin Zheng, Walter F Good, Jill L King, and Yuan-Hsiang Chang. Computer- assisted diagnosis of breast cancer using a data-driven bayesian belief network. International Journal of Med- ical Informatics, 54(2):115-126, 1999.","title":"Computerassisted diagnosis of breast cancer using a data-driven bayesian belief network","authors":"Xiao-Hui Wang, Bin Zheng, Walter F Good, Jill L King, Yuan-Hsiang Chang","year":"1999","venue":"International Journal of Medical Informatics","doi":null,"url":null,"internal_url":null},{"position":19,"raw":"[Wei et al., 2018] Zhongyu Wei, Qianlong Liu, Baolin Peng, Huaixiao Tou, Ting Chen, Xuanjing Huang, Kamfai Wong, and Xiangying Dai. Task-oriented dialogue system for automatic diagnosis. In Proc. of ACL, pages 201-207, July 2018.","title":"Task-oriented dialogue system for automatic diagnosis","authors":"Zhongyu Wei, Qianlong Liu, Baolin Peng, Huaixiao Tou, Ting Chen, Xuanjing Huang, Kamfai Wong, Xiangying Dai","year":"2018","venue":"Proc. of ACL","doi":null,"url":null,"internal_url":null},{"position":20,"raw":"[Xia et al., 2020] Yuan Xia, Jingbo Zhou, Zhenhui Shi, Chao Lu, and Haifeng Huang. Generative adversarial regular- ized mutual information policy gradient framework for automatic diagnosis. In Proc. of AAAI, volume 34, pages 1062-1069, 2020.","title":"Generative adversarial regularized mutual information policy gradient framework for automatic diagnosis","authors":"Yuan Xia, Jingbo Zhou, Zhenhui Shi, Chao Lu, Haifeng Huang","year":"2020","venue":"Proc. of AAAI","doi":null,"url":null,"internal_url":null},{"position":21,"raw":"[Xu et al., 2019] Lin Xu, Qixian Zhou, Ke Gong, Xiao- dan Liang, Jianheng Tang, and Liang Lin. End-to-end knowledge-routed relational dialogue system for automatic diagnosis. In Proc. of AAAI, volume 33, pages 7346-7353, 2019.","title":"End-to-end knowledge-routed relational dialogue system for automatic diagnosis","authors":"Lin Xu, Qixian Zhou, Ke Gong, Xiaodan Liang, Jianheng Tang, Liang Lin","year":"2019","venue":"Proc. of AAAI","doi":null,"url":null,"internal_url":null},{"position":22,"raw":"[Zhang and Poole, 1994] Nevin L Zhang and David Poole. A simple approach to bayesian network computations. In Proc. of the Tenth Canadian Conference on Artificial Intel- ligence, 1994.","title":"A simple approach to bayesian network computations","authors":"L Nevin, David Zhang, Poole","year":"1994","venue":"Proc. of the Tenth Canadian Conference on Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":23,"raw":"[Zhao et al., 2020] Xinyan Zhao, Liangwei Chen, and Huan- huan Chen. A graph-based and patient demographics aware dialogue system for disease diagnosis. arXiv preprint, arXiv:2010.10699, 2020.","title":"A graph-based and patient demographics aware dialogue system for disease diagnosis","authors":"Xinyan Zhao, Liangwei Chen, Huanhuan Chen","year":"2020","venue":null,"doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}