{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,28]],"date-time":"2026-09-28T01:55:14Z","timestamp":1790560514609,"version":"4.1.0"},"reference-count":32,"publisher":"Springer Science and Business Media LLC","issue":"S9","license":[{"start":{"date-parts":[[2021,11,1]],"date-time":"2021-11-01T00:00:00Z","timestamp":1635724800000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2021,11,16]],"date-time":"2021-11-16T00:00:00Z","timestamp":1637020800000},"content-version":"vor","delay-in-days":15,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"DOI":"10.13039\/501100012226","name":"Fundamental Research Funds for the Central Universities","doi-asserted-by":"publisher","award":["2682020ZT92"],"award-info":[{"award-number":["2682020ZT92"]}],"id":[{"id":"10.13039\/501100012226","id-type":"DOI","asserted-by":"publisher"}]},{"name":"sichuan key r&d project","award":["2021YFG0136"],"award-info":[{"award-number":["2021YFG0136"]}]}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["BMC Med Inform Decis Mak"],"published-print":{"date-parts":[[2021,11]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:sec><jats:title>Background<\/jats:title><jats:p>Clinical notes are unstructured text documents generated by clinicians during patient encounters, generally are annotated with International Classification of Diseases (ICD) codes, which give formatted information about the diagnosis and treatment. ICD code has shown its potentials in many fields, but manual coding is labor-intensive and error-prone, lead to researches of automatic coding. Two specific challenges of this task are (1) given an annotated clinical notes, the reasons behind specific diagnoses and treatments are \u00a0implicit; (2) explainability is important for practical automatic coding method, the method should not only explain its prediction output but also have explainable internal mechanics. This study aims to develop an explainable CNN approach to address these two challenges.<\/jats:p><\/jats:sec><jats:sec><jats:title>Method<\/jats:title><jats:p>Our key idea is that for the automatic ICD coding task, the presence of informative snippets in the clinical text that correlated with each code plays an important role in the prediction of codes, and an informative snippet can be considered as a local and low-level feature. We infer that there exists a correspondence between a convolution filter and a local and low-level feature. Base on the inference, we come up with the Shallow and Wide Attention convolutional Mechanism (SWAM) to improve the CNN-based models\u2019 ability to learn local and low-level features for each label.<\/jats:p><\/jats:sec><jats:sec><jats:title>Results<\/jats:title><jats:p>We evaluate our approach on MIMIC-III, an open-access dataset of ICU medical records. Our approach substantially outperforms previous results on top-50 medical code prediction on MIMIC-III dataset, the precision of the worst-performing 10% labels in previous works is increased from 0% to 53% on average. We attribute this improvement to SWAM, by which the wide architecture with attention mechanism gives the model ability to more extensively learn the unique features of different codes, and we prove it by an ablation experiment. Besides, we perform manual analysis of the performance imbalance between different codes, and preliminary conclude the characteristics that determine the difficulty of learning specific codes.<\/jats:p><\/jats:sec><jats:sec><jats:title>Conclusions<\/jats:title><jats:p>Our main contributions can be summarized into the following three: (1) We present\u00a0local and low-level features, a.k.a. informative snippets play an important role in the automatic ICD coding task, and the informative snippets extracted from the clinical text provide explanations for each code. (2) We propose that there exists a correspondence between a convolution filter and a local and low-level feature. A combination of wide and shallow convolutional layer and attention layer can help the CNN-based models better learn local and low-level features. (3) We improved the precision of the worst-performing 10% labels from 0 to 53% on average.<\/jats:p><\/jats:sec>","DOI":"10.1186\/s12911-021-01615-6","type":"journal-article","created":{"date-parts":[[2021,11,17]],"date-time":"2021-11-17T10:04:28Z","timestamp":1637143468000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":30,"title":["An explainable CNN approach for medical codes prediction from clinical text"],"prefix":"10.1186","volume":"21","author":[{"given":"Shuyuan","family":"Hu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-9535-7245","authenticated-orcid":false,"given":"Fei","family":"Teng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lufei","family":"Huang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jun","family":"Yan","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Haibo","family":"Zhang","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2021,11,16]]},"reference":[{"issue":"9","key":"1615_CR1","doi-asserted-by":"publisher","first-page":"1205","DOI":"10.1093\/bioinformatics\/btq126","volume":"26","author":"JC Denny","year":"2010","unstructured":"Denny JC, Ritchie MD, Basford MA, Pulley JM, Bastarache L, Brown-Gentry K, et al. PheWAS: demonstrating the feasibility of a phenome-wide scan to discover gene-disease associations. Bioinformatics. 2010;26(9):1205\u201310.","journal-title":"Bioinformatics"},{"key":"1615_CR2","unstructured":"Ranganath R, Perotte AJ, Elhadad N, Blei DM. The survival filter: joint survival analysis with a latent time series. In: Proceedings of Uncertainty in artificial intelligence (UAI); 2015. p. 742\u2013751."},{"issue":"4","key":"1615_CR3","doi-asserted-by":"publisher","first-page":"122","DOI":"10.1186\/s12911-018-0677-8","volume":"18","author":"A Avati","year":"2018","unstructured":"Avati A, Jung K, Harman S, Downing L, Ng A, Shah NH. Improving palliative care with deep learning. BMC Med Inf Decis Mak. 2018;18(4):122.","journal-title":"BMC Med Inf Decis Mak"},{"key":"1615_CR4","unstructured":"Choi E, Bahadori MT, Schuetz A, Stewart WF, Sun J. Doctor AI: predicting clinical events via recurrent neural networks. In: Proceedings of Machine Learning for Healthcare Conference (MLHC); 2016. p. 301\u2013318."},{"key":"1615_CR5","doi-asserted-by":"crossref","unstructured":"Zhang D, He D, Zhao S, Li L. Enhancing automatic ICD-9-CM code assignment for medical texts with pubmed. In: Proceedings of Biomedical Natural Language Processing Workshop (BioNLP); 2017. p. 263\u2013271.","DOI":"10.18653\/v1\/W17-2333"},{"issue":"2","key":"1615_CR6","doi-asserted-by":"publisher","first-page":"155","DOI":"10.1016\/j.artmed.2015.04.007","volume":"65","author":"R Kavuluru","year":"2015","unstructured":"Kavuluru R, Rios A, Lu Y. An empirical evaluation of supervised learning approaches in assigning diagnosis codes to electronic medical records. Artif Intell Med. 2015;65(2):155\u201366.","journal-title":"Artif Intell Med"},{"issue":"1","key":"1615_CR7","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/sdata.2016.35","volume":"3","author":"AE Johnson","year":"2016","unstructured":"Johnson AE, Pollard TJ, Shen L, Li-Wei HL, Feng M, Ghassemi M, et al. MIMIC-III, a freely accessible critical care database. Sci Data. 2016;3(1):1\u20139.","journal-title":"Sci Data"},{"key":"1615_CR8","doi-asserted-by":"crossref","unstructured":"Subotin M, Davis A. A system for predicting ICD-10-PCS codes from electronic health records. In: Proceedings of Biomedical Natural Language Processing Workshop (BioNLP); 2014. p. 59\u201367.","DOI":"10.3115\/v1\/W14-3409"},{"issue":"1","key":"1615_CR9","doi-asserted-by":"publisher","first-page":"92","DOI":"10.1016\/j.jbi.2017.09.004","volume":"74","author":"E Scheurwegs","year":"2017","unstructured":"Scheurwegs E, Cule B, Luyckx K, Luyten L, Daelemans W. Selecting relevant features from the electronic health record for clinical code prediction. J Biomed Inf. 2017;74(1):92\u2013103.","journal-title":"J Biomed Inf"},{"issue":"12","key":"1615_CR10","doi-asserted-by":"publisher","first-page":"3191","DOI":"10.1109\/TKDE.2016.2605687","volume":"28","author":"S Wang","year":"2016","unstructured":"Wang S, Chang X, Li X, Long G, Yao L, Sheng QZ. Diagnosis code assignment using sparsity-based disease correlation embedding. IEEE Trans Knowl Data Eng. 2016;28(12):3191\u2013202.","journal-title":"IEEE Trans Knowl Data Eng"},{"key":"1615_CR11","doi-asserted-by":"crossref","unstructured":"Prakash A, Zhao S, Hasan SA, Datla V, Lee K, Qadir A, et\u00a0al. Condensed memory networks for clinical diagnostic inferencing. In: Proceedings of the Thirty-First AAAI Conference on Artificial Intelligence; 2017. p. 3274\u20133280.","DOI":"10.1609\/aaai.v31i1.10964"},{"key":"1615_CR12","doi-asserted-by":"crossref","unstructured":"Mullenbach J, Wiegreffe S, Duke J, Sun J, Eisenstein J. Explainable prediction of medical codes from clinical text. In: Proceedings of the 2018 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2018;1(1):1101\u201311.","DOI":"10.18653\/v1\/N18-1100"},{"key":"1615_CR13","unstructured":"Shi H, Xie P, Hu Z, Zhang M, Xing EP. Towards automated ICD coding using deep learning; 2017. p. 1\u201311. arXiv preprint arXiv:171104075."},{"key":"1615_CR14","unstructured":"Allamanis M, Peng H, Sutton C. A convolutional attention network for extreme summarization of source code. In: International conference on machine learning (ICML); 2016. p. 2091\u20132100."},{"key":"1615_CR15","doi-asserted-by":"publisher","first-page":"259","DOI":"10.1162\/tacl_a_00097","volume":"4","author":"W Yin","year":"2016","unstructured":"Yin W, Sch\u00fctze H, Xiang B, Zhou B. Abcnn: attention-based convolutional neural network for modeling sentence pairs. Trans Assoc Comput Linguist. 2016;4:259\u201372.","journal-title":"Trans Assoc Comput Linguist"},{"key":"1615_CR16","unstructured":"Santos CD, Tan M, Xiang B, Zhou B. Attentive pooling networks; 2016;p. 1\u201310. arXiv preprint arXiv:160203609."},{"key":"1615_CR17","doi-asserted-by":"crossref","unstructured":"Yin W, Sch\u00fctze H. Attentive convolution: equipping CNNs with RNN-style attention mechanisms; 2017;p. 1\u201316. arXiv preprint arXiv:171000519.","DOI":"10.1162\/tacl_a_00249"},{"key":"1615_CR18","unstructured":"Bahdanau D, Cho K, Bengio Y. Neural machine translation by jointly learning to align and translate. In: Proceedings of 3rd International Conference on Learning Representations; 2014. p. 1\u201315."},{"key":"1615_CR19","doi-asserted-by":"crossref","unstructured":"Yang Z, Yang D, Dyer C, He X, Smola A, Hovy E. Hierarchical attention networks for document classification. In: Proceedings of the 2016 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies; 2016. p. 1480\u20131489.","DOI":"10.18653\/v1\/N16-1174"},{"key":"1615_CR20","doi-asserted-by":"crossref","unstructured":"Rush AM, Chopra S, Weston J. A neural attention model for abstractive sentence summarization. In: Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing; 2015. p. 379\u2013389.","DOI":"10.18653\/v1\/D15-1044"},{"key":"1615_CR21","unstructured":"Rockt\u00e4schel T, Grefenstette E, Hermann KM, Ko\u010disk\u1ef3 T, Blunsom P. Reasoning about Entailment with Neural Attention. In: 4th International Conference on Learning Representations(ICLR); 2015. p. 1\u20139."},{"key":"1615_CR22","unstructured":"Le HT, C, Denis A. Do convolutional networks need to be deep for text classification? In: AAAI Workshop on Affective Content Analysis. 2017;p. 1\u201312."},{"issue":"3","key":"1615_CR23","first-page":"160","volume":"7","author":"R Collobert","year":"2008","unstructured":"Collobert R, Weston J. A unified architecture for natural language processing: Deep neural networks with multitask learning. Proceedings of the 25th international conference on Machine learning. 2008;7(3):160\u20137.","journal-title":"Proceedings of the 25th international conference on Machine learning"},{"key":"1615_CR24","unstructured":"Gong L, Ji R. What Does a TextCNN Learn? 2018. p. 1\u20139. arXiv preprint arXiv:180106287."},{"key":"1615_CR25","doi-asserted-by":"crossref","unstructured":"Kim Y. Convolutional Neural Networks for Sentence Classification. In: Proceedings of the 2014 Conference on Empirical Methods in Natural Language Processing (EMNLP); 2014. p. 1746\u20131751.","DOI":"10.3115\/v1\/D14-1181"},{"key":"1615_CR26","unstructured":"Mikolov T, Sutskever I, Chen K, Corrado GS, Dean J. Distributed representations of words and phrases and their compositionality. In: Proceedings of Advances in Neural Information Processing Systems; 2013. p. 3111\u20133119."},{"key":"1615_CR27","unstructured":"McCallum AK. Multi-label text classification with a mixture model trained by EM. In: Proceedings of Association for the Advancement of Artificial Intelligence 99 workshop on text learning; 1999. p. 1\u20137."},{"key":"1615_CR28","unstructured":"Vaswani A, Shazeer N, Parmar N, Uszkoreit J, Jones L, Gomez AN, et\u00a0al. Attention is all you need. In: Proceedings of Advances in Neural Information Processing Systems; 2017. p. 5998\u20136008."},{"issue":"1","key":"1615_CR29","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1038\/s41746-017-0008-y","volume":"1","author":"A Rajkomar","year":"2018","unstructured":"Rajkomar A, Oren E, Chen K, Dai AM, Hajaj N, Hardt M, et al. Scalable and accurate deep learning with electronic health records. NPJ Digit Med. 2018;1(1):1\u201310.","journal-title":"NPJ Digit Med"},{"issue":"221(01)","key":"1615_CR30","first-page":"20","volume":"46","author":"S Lyu","year":"2018","unstructured":"Lyu S, Liu W. Estimation methods of p value of nonparametric test based on the Bootstrap idea. J Fuzhou Univ (Natural Science Edition). 2018;46(221(01)):20\u20136.","journal-title":"J Fuzhou Univ (Natural Science Edition)"},{"key":"1615_CR31","unstructured":"Aghaebrahimian A, Cieliebak M. Hyperparameter tuning for deep learning in natural language processing. In: Proceedings of 4th Swiss Text Analytics Conference (SwissText); 2019. p. 1\u20137."},{"key":"1615_CR32","doi-asserted-by":"crossref","unstructured":"Prakash A, Storer J, Florencio D, Zhang C. Repr: improved training of convolutional filters. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition; 2019. p. 10666\u201310675.","DOI":"10.1109\/CVPR.2019.01092"}],"container-title":["BMC Medical Informatics and Decision Making"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12911-021-01615-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1186\/s12911-021-01615-6\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1186\/s12911-021-01615-6.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2023,1,15]],"date-time":"2023-01-15T14:46:32Z","timestamp":1673793992000},"score":1,"resource":{"primary":{"URL":"https:\/\/bmcmedinformdecismak.biomedcentral.com\/articles\/10.1186\/s12911-021-01615-6"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,11]]},"references-count":32,"journal-issue":{"issue":"S9","published-print":{"date-parts":[[2021,11]]}},"alternative-id":["1615"],"URL":"https:\/\/doi.org\/10.1186\/s12911-021-01615-6","relation":{},"ISSN":["1472-6947"],"issn-type":[{"value":"1472-6947","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,11]]},"assertion":[{"value":"8 August 2021","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"23 August 2021","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"16 November 2021","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"Not applicable.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Ethics approval and consent to participate"}},{"value":"Not applicable.","order":3,"name":"Ethics","group":{"name":"EthicsHeading","label":"Consent for publication"}},{"value":"The authors declare that they have no competing interests.","order":4,"name":"Ethics","group":{"name":"EthicsHeading","label":"Competing interest"}}],"article-number":"256"}}