{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,2,10]],"date-time":"2026-02-10T21:41:40Z","timestamp":1770759700716,"version":"3.50.0"},"reference-count":34,"publisher":"Oxford University Press (OUP)","issue":"4","license":[{"start":{"date-parts":[[2025,1,28]],"date-time":"2025-01-28T00:00:00Z","timestamp":1738022400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/academic.oup.com\/pages\/standard-publication-reuse-rights"}],"funder":[{"name":"German Federal Ministry of Health"},{"DOI":"10.13039\/501100003107","name":"Bundesministerium f\u00fcr Gesundheit","doi-asserted-by":"publisher","id":[{"id":"10.13039\/501100003107","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Merging and Validating Cancer Registry Data using AI Methods"},{"name":"Zusammenf\u00fchren und Validieren von Krebsregisterdaten durch KI-Verfahren"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2025,4,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:sec>\n                  <jats:title>Objectives<\/jats:title>\n                  <jats:p>Explanations help to understand why anomaly detection algorithms identify data as anomalous. This study evaluates whether robustly standardized explanation scores correctly identify the implausible variables that make cancer data anomalous.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Materials and Methods<\/jats:title>\n                  <jats:p>The dataset analyzed consists of 18\u00a0587 truncated real-world cancer registry records containing 8 categorical variables describing patients diagnosed with bladder and lung tumors. We identified 800 anomalous records using an autoencoder\u2019s per-record reconstruction error, which is a common neural network-based anomaly detection approach. For each variable of a record, we determined a robust explanation score, which indicates how anomalous the variable is. A variable\u2019s robust explanation score is the autoencoder\u2019s per-variable reconstruction error measured by cross-entropy and robustly standardized across records; that is, large reconstruction errors have a small effect on standardization. To evaluate the explanation scores, medical coders identified the implausible variables of the anomalous records. We then compare the explanation scores to the medical coders\u2019 validation in a classification and ranking setting. As baselines, we identified anomalous variables using the raw autoencoder\u2019s per-variable reconstruction error, the non-robustly standardized per-variable reconstruction error, the empirical frequency of implausible variables according to the medical coders\u2019 validation, and random selection or ranking of variables.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>When we sort the variables by their robust explanation scores, on average, the 2.37 highest-ranked variables contain all implausible variables. For the baselines, on average, the 2.84, 2.98, 3.27, and 4.91 highest-ranked variables contain all the variables that made a record implausible.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Discussion<\/jats:title>\n                  <jats:p>We found that explanations based on robust explanation scores were better than or as good as the baseline explanations examined in the classification and ranking settings. Due to the international standardization of cancer data coding, we expect our results to generalize to other cancer types and registries. As we anticipate different magnitudes of per-variable autoencoder reconstruction errors in data from other medical registries and domains, these may also benefit from robustly standardizing the reconstruction errors per variable. Future work could explore methods to identify subsets of anomalous variables, addressing whether individual variables or their combinations contribute to anomalies. This direction aims to improve the interpretability and utility of anomaly detection systems.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Conclusions<\/jats:title>\n                  <jats:p>Robust explanation scores can improve explanations for identifying implausible variables in cancer data.<\/jats:p>\n               <\/jats:sec>","DOI":"10.1093\/jamia\/ocaf011","type":"journal-article","created":{"date-parts":[[2025,1,28]],"date-time":"2025-01-28T15:57:11Z","timestamp":1738079831000},"page":"724-735","source":"Crossref","is-referenced-by-count":1,"title":["Evaluating robustly standardized explainable anomaly detection of implausible variables in cancer data"],"prefix":"10.1093","volume":"32","author":[{"ORCID":"https:\/\/orcid.org\/0000-0003-3376-6670","authenticated-orcid":false,"given":"Philipp","family":"R\u00f6chner","sequence":"first","affiliation":[{"name":"Information Systems and Business Administration, Johannes Gutenberg University , Mainz 55128,","place":["Germany"]},{"name":"Cancer Registry Rhineland-Palatinate, Institute for Digital Health Data , Mainz 55116,","place":["Germany"]}]},{"given":"Franz","family":"Rothlauf","sequence":"additional","affiliation":[{"name":"Information Systems and Business Administration, Johannes Gutenberg University , Mainz 55128,","place":["Germany"]}]}],"member":"286","published-online":{"date-parts":[[2025,1,28]]},"reference":[{"key":"2025041716422186200_ocaf011-B1","doi-asserted-by":"crossref","first-page":"144","DOI":"10.1136\/amiajnl-2011-000681","article-title":"Methods and dimensions of electronic health record data quality assessment: enabling reuse for clinical research","volume":"20","author":"Weiskopf","year":"2013","journal-title":"J. Am. Medical Informatics Assoc"},{"key":"2025041716422186200_ocaf011-B2","author":"Martos","year":"2014"},{"key":"2025041716422186200_ocaf011-B3","author":"Martos","year":"2018"},{"key":"2025041716422186200_ocaf011-B4","author":"Martos","year":"2023"},{"key":"2025041716422186200_ocaf011-B5","author":"Ferlay","year":"2005"},{"key":"2025041716422186200_ocaf011-B6","doi-asserted-by":"crossref","first-page":"1250195","DOI":"10.3389\/fonc.2023.1250195","article-title":"The Joint Research Centre\u2013European Network of Cancer Registries Quality Check Software (JRC-ENCR QCS)","volume":"13","author":"Giusti","year":"2023","journal-title":"Front Oncol"},{"key":"2025041716422186200_ocaf011-B7","doi-asserted-by":"crossref","first-page":"125","DOI":"10.1186\/s12874-023-01946-0","article-title":"Unsupervised anomaly detection of implausible electronic health records: a real-world evaluation in cancer registries","volume":"23","author":"R\u00f6chner","year":"2023","journal-title":"BMC Med Res Methodol"},{"key":"2025041716422186200_ocaf011-B8","doi-asserted-by":"crossref","DOI":"10.1007\/978-94-015-3994-4","volume-title":"Identification of Outliers. Monographs on Applied Probability and Statistics","author":"Hawkins","year":"1980"},{"key":"2025041716422186200_ocaf011-B9","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/1541880.1541882","article-title":"Anomaly detection: a survey","volume":"41","author":"Chandola","year":"2009","journal-title":"ACM Comput Surv"},{"key":"2025041716422186200_ocaf011-B10","doi-asserted-by":"crossref","first-page":"756","DOI":"10.1109\/JPROC.2021.3052449","article-title":"A unifying review of deep and shallow anomaly detection","volume":"109","author":"Ruff","year":"2021","journal-title":"Proc IEEE"},{"key":"2025041716422186200_ocaf011-B11","doi-asserted-by":"crossref","first-page":"101946","DOI":"10.1016\/j.datak.2021.101946","article-title":"Anomaly explanation: a review","volume":"137","author":"Yepmo","year":"2022","journal-title":"Data Knowl Eng"},{"key":"2025041716422186200_ocaf011-B12","first-page":"1","article-title":"A survey on explainable anomaly detection","volume":"18","author":"Li","year":"2024","journal-title":"ACM Trans Knowl Discov Data"},{"key":"2025041716422186200_ocaf011-B13","first-page":"1","article-title":"Deep learning for unsupervised anomaly localization in industrial images: a survey","volume":"71","author":"Tao","year":"2022","journal-title":"IEEE Trans Instrum Meas"},{"key":"2025041716422186200_ocaf011-B14","doi-asserted-by":"crossref","first-page":"977","DOI":"10.1007\/s00778-021-00721-1","article-title":"A survey on outlier explanations","volume":"31","author":"Panjei","year":"2022","journal-title":"VLDB J"},{"key":"2025041716422186200_ocaf011-B15","first-page":"133","volume-title":"MLCN\/DLF\/iMIMIC@MICCAI, Lecture Notes in Computer Science","author":"Silva","year":"2018"},{"key":"2025041716422186200_ocaf011-B16","first-page":"1803","article-title":"How to explain individual classification decisions","volume":"11","author":"Baehrens","year":"2010","journal-title":"J Mach Learn Res"},{"key":"2025041716422186200_ocaf011-B17","first-page":"304","volume-title":"ECML\/PKDD (3), Lecture Notes in Computer Science","author":"Dang","year":"2013"},{"key":"2025041716422186200_ocaf011-B18","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3583558","article-title":"From anecdotal evidence to quantitative evaluation methods: a systematic review on evaluating explainable AI","volume":"55","author":"Nauta","year":"2023","journal-title":"ACM Comput Surv"},{"key":"2025041716422186200_ocaf011-B19","author":"Sakurada","year":"2014"},{"key":"2025041716422186200_ocaf011-B20","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-031-45468-4","volume-title":"Deep Learning \u2013 Foundations and Concepts","author":"Bishop","year":"2024"},{"key":"2025041716422186200_ocaf011-B21","volume-title":"Robust Statistics: Theory and Methods (with R)","author":"Maronna","year":"2019"},{"key":"2025041716422186200_ocaf011-B22","author":"Allaire","year":"2023"},{"key":"2025041716422186200_ocaf011-B23","doi-asserted-by":"crossref","first-page":"149","DOI":"10.32614\/RJ-2015-027","article-title":"Working with multilabel datasets in R: the mldr package","volume":"7","author":"Charte","year":"2015","journal-title":"R J"},{"key":"2025041716422186200_ocaf011-B24","first-page":"667","volume-title":"Data Mining and Knowledge Discovery Handbook","author":"Tsoumakas","year":"2010"},{"key":"2025041716422186200_ocaf011-B25","doi-asserted-by":"crossref","first-page":"103","DOI":"10.2298\/CSIS0501103H","article-title":"FP-outlier: frequent pattern based outlier detection","volume":"2","author":"He","year":"2005","journal-title":"ComSIS"},{"key":"2025041716422186200_ocaf011-B26","doi-asserted-by":"crossref","first-page":"80","DOI":"10.2307\/3001968","article-title":"Individual comparisons by ranking methods","volume":"1","author":"Wilcoxon","year":"1945","journal-title":"Biometrics"},{"key":"2025041716422186200_ocaf011-B27","first-page":"65","article-title":"A simple sequentially rejective multiple test procedure","volume":"6","author":"Holm","year":"1979","journal-title":"Scand J Stat"},{"key":"2025041716422186200_ocaf011-B28","first-page":"1","article-title":"Statistical comparisons of classifiers over multiple data sets","volume":"7","author":"Demsar","year":"2006","journal-title":"J Mach Learn Res"},{"key":"2025041716422186200_ocaf011-B29","first-page":"5:1","article-title":"Should we really use post-hoc tests based on mean-ranks?","volume":"17","author":"Benavoli","year":"2016","journal-title":"J Mach Learn Res"},{"key":"2025041716422186200_ocaf011-B30","first-page":"518","volume-title":"ICDM","author":"Micenkov\u00e1","year":"2013"},{"key":"2025041716422186200_ocaf011-B31","first-page":"831","volume-title":"PAKDD, Lecture Notes in Computer Science","author":"Kriegel","year":"2009"},{"key":"2025041716422186200_ocaf011-B32","first-page":"379","volume-title":"ICDM","author":"Kriegel","year":"2012"},{"key":"2025041716422186200_ocaf011-B33","volume-title":"TNM Classification of Malignant Tumours","author":"Brierley","year":"2017"},{"key":"2025041716422186200_ocaf011-B34","doi-asserted-by":"crossref","first-page":"356","DOI":"10.1016\/j.jtcvs.2017.08.138","article-title":"The eighth edition TNM stage classification for lung cancer: what does it mean on main street?","volume":"155","author":"Detterbeck","year":"2018","journal-title":"J Thorac Cardiovasc Surg"}],"container-title":["Journal of the American Medical Informatics Association"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/jamia\/article-pdf\/32\/4\/724\/61662735\/ocaf011.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/jamia\/article-pdf\/32\/4\/724\/61662735\/ocaf011.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,4,17]],"date-time":"2025-04-17T20:42:42Z","timestamp":1744922562000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/jamia\/article\/32\/4\/724\/7985530"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,1,28]]},"references-count":34,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2025,1,28]]},"published-print":{"date-parts":[[2025,4,1]]}},"URL":"https:\/\/doi.org\/10.1093\/jamia\/ocaf011","relation":{},"ISSN":["1067-5027","1527-974X"],"issn-type":[{"value":"1067-5027","type":"print"},{"value":"1527-974X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2025,4]]},"published":{"date-parts":[[2025,1,28]]}}}