{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,2]],"date-time":"2026-06-02T00:36:32Z","timestamp":1780360592330,"version":"3.54.1"},"reference-count":32,"publisher":"Oxford University Press (OUP)","issue":"Supplement_1","license":[{"start":{"date-parts":[[2020,7,13]],"date-time":"2020-07-13T00:00:00Z","timestamp":1594598400000},"content-version":"vor","delay-in-days":12,"URL":"http:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"funder":[{"DOI":"10.13039\/100000001","name":"NSF","doi-asserted-by":"publisher","award":["DBI-1759736"],"award-info":[{"award-number":["DBI-1759736"]}],"id":[{"id":"10.13039\/100000001","id-type":"DOI","asserted-by":"publisher"}]},{"name":"German Research Council"},{"DOI":"10.13039\/100004807","name":"DFG","doi-asserted-by":"publisher","award":["SCHI 871\/11-1"],"award-info":[{"award-number":["SCHI 871\/11-1"]}],"id":[{"id":"10.13039\/100004807","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"published-print":{"date-parts":[[2020,7,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:sec>\n                  <jats:title>Motivation<\/jats:title>\n                  <jats:p>Mass spectrometry imaging (MSI) characterizes the molecular composition of tissues at spatial resolution, and has a strong potential for distinguishing tissue types, or disease states. This can be achieved by supervised classification, which takes as input MSI spectra, and assigns class labels to subtissue locations. Unfortunately, developing such classifiers is hindered by the limited availability of training sets with subtissue labels as the ground truth. Subtissue labeling is prohibitively expensive, and only rough annotations of the entire tissues are typically available. Classifiers trained on data with approximate labels have sub-optimal performance.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Results<\/jats:title>\n                  <jats:p>To alleviate this challenge, we contribute a semi-supervised approach mi-CNN. mi-CNN implements multiple instance learning with a convolutional neural network (CNN). The multiple instance aspect enables weak supervision from tissue-level annotations when classifying subtissue locations. The convolutional architecture of the CNN captures contextual dependencies between the spectral features. Evaluations on simulated and experimental datasets demonstrated that mi-CNN improved the subtissue classification as compared to traditional classifiers. We propose mi-CNN as an important step toward accurate subtissue classification in MSI, enabling rapid distinction between tissue types and disease states.<\/jats:p>\n               <\/jats:sec>\n               <jats:sec>\n                  <jats:title>Availability and implementation<\/jats:title>\n                  <jats:p>The data and code are available at https:\/\/github.com\/Vitek-Lab\/mi-CNN_MSI.<\/jats:p>\n               <\/jats:sec>","DOI":"10.1093\/bioinformatics\/btaa436","type":"journal-article","created":{"date-parts":[[2020,5,3]],"date-time":"2020-05-03T11:06:59Z","timestamp":1588504019000},"page":"i300-i308","source":"Crossref","is-referenced-by-count":25,"title":["Deep multiple instance learning classifies subtissue locations in mass spectrometry images from tissue-level annotations"],"prefix":"10.1093","volume":"36","author":[{"given":"Dan","family":"Guo","sequence":"first","affiliation":[{"name":"Khoury College of Computer Sciences, Northeastern University , Boston, MA 02115, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Melanie Christine","family":"F\u00f6ll","sequence":"additional","affiliation":[{"name":"Institute for Surgical Pathology, Medical Center \u2013 University of Freiburg , 79106 Freiburg, Germany"},{"name":"Faculty of Medicine, University of Freiburg , 79110 Freiburg, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Veronika","family":"Volkmann","sequence":"additional","affiliation":[{"name":"Institute for Surgical Pathology, Medical Center \u2013 University of Freiburg , 79106 Freiburg, Germany"},{"name":"Faculty of Medicine, University of Freiburg , 79110 Freiburg, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Kathrin","family":"Enderle-Ammour","sequence":"additional","affiliation":[{"name":"Institute for Surgical Pathology, Medical Center \u2013 University of Freiburg , 79106 Freiburg, Germany"},{"name":"Faculty of Medicine, University of Freiburg , 79110 Freiburg, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Peter","family":"Bronsert","sequence":"additional","affiliation":[{"name":"Institute for Surgical Pathology, Medical Center \u2013 University of Freiburg , 79106 Freiburg, Germany"},{"name":"Faculty of Medicine, University of Freiburg , 79110 Freiburg, Germany"},{"name":"Tumorbank Comprehensive Cancer Center Freiburg, Medical Center \u2013 University of Freiburg"},{"name":"German Cancer Consortium (DKTK) and Cancer Research Center (DKFZ) , 79106 Freiburg, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Oliver","family":"Schilling","sequence":"additional","affiliation":[{"name":"Institute for Surgical Pathology, Medical Center \u2013 University of Freiburg , 79106 Freiburg, Germany"},{"name":"Faculty of Medicine, University of Freiburg , 79110 Freiburg, Germany"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Olga","family":"Vitek","sequence":"additional","affiliation":[{"name":"Khoury College of Computer Sciences, Northeastern University , Boston, MA 02115, USA"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"286","published-online":{"date-parts":[[2020,7,13]]},"reference":[{"key":"2024021913335994200_btaa436-B1","doi-asserted-by":"crossref","first-page":"422","DOI":"10.1038\/labinvest.2014.156","article-title":"MALDI imaging mass spectrometry: current frontiers and perspectives in pathology research and practice","volume":"95","author":"Aichler","year":"2015","journal-title":"Lab. Invest"},{"key":"2024021913335994200_btaa436-B2","author":"Allaire","year":"2019"},{"key":"2024021913335994200_btaa436-B3","first-page":"577","article-title":"Support vector machines for multiple-instance learning","author":"Andrews","year":"2003"},{"key":"2024021913335994200_btaa436-B4","doi-asserted-by":"crossref","first-page":"1215","DOI":"10.1093\/bioinformatics\/btx724","article-title":"Deep learning for tumor classification in imaging mass spectrometry","volume":"34","author":"Behrmann","year":"2018","journal-title":"Bioinformatics"},{"key":"2024021913335994200_btaa436-B5","doi-asserted-by":"crossref","first-page":"2418","DOI":"10.1093\/bioinformatics\/btv146","article-title":"Cardinal: an R package for statistical analysis of mass spectrometry-based imaging experiments","volume":"31","author":"Bemis","year":"2015","journal-title":"Bioinformatics"},{"key":"2024021913335994200_btaa436-B6","doi-asserted-by":"crossref","first-page":"1761","DOI":"10.1074\/mcp.O115.053918","article-title":"Probabilistic segmentation of mass spectrometry images helps select important ions and characterize confidence in the resulting segments","volume":"15","author":"Bemis","year":"2016","journal-title":"Mol. Cell. Proteomics"},{"key":"2024021913335994200_btaa436-B7","author":"Bemis","year":"2019"},{"key":"2024021913335994200_btaa436-B8","doi-asserted-by":"crossref","first-page":"9978","DOI":"10.1073\/pnas.1423101112","article-title":"MALDI mass spectrometry imaging analysis of pituitary adenomas for near-real-time tumor delineation","volume":"112","author":"Calligaris","year":"2015","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"2024021913335994200_btaa436-B9","doi-asserted-by":"crossref","first-page":"280","DOI":"10.1016\/j.media.2019.03.009","article-title":"Not-so-supervised: a survey of semi-supervised, multi-instance, and transfer learning in medical image analysis","volume":"54","author":"Cheplygina","year":"2019","journal-title":"Med. Image Anal"},{"key":"2024021913335994200_btaa436-B10","doi-asserted-by":"crossref","first-page":"2969","DOI":"10.1007\/s00216-010-4259-6","article-title":"Multivariate statistical differentiation of renal cell carcinomas based on lipidomic analysis by ambient ionization imaging mass spectrometry","volume":"398","author":"Dill","year":"2010","journal-title":"Anal. Bioanal. Chem"},{"key":"2024021913335994200_btaa436-B11","doi-asserted-by":"crossref","first-page":"2897","DOI":"10.1002\/chem.201001692","article-title":"Multivariate statistical identification of human bladder carcinomas using ambient ionization imaging mass spectrometry","volume":"17","author":"Dill","year":"2011","journal-title":"Chemistry"},{"key":"2024021913335994200_btaa436-B12","doi-asserted-by":"crossref","first-page":"2436","DOI":"10.1073\/pnas.1400274111","article-title":"Molecular assessment of surgical-resection margins of gastric cancer by mass-spectrometric imaging","volume":"111","author":"Eberlin","year":"2014","journal-title":"Proc. Natl. Acad. Sci. USA"},{"key":"2024021913335994200_btaa436-B13","doi-asserted-by":"crossref","first-page":"628719","DOI":"10.1093\/gigascience\/giz143","article-title":"Reproducible mass spectrometry imaging data analysis in Galaxy","volume":"8","author":"F\u00f6ll","year":"2019","journal-title":"GigaScience"},{"key":"2024021913335994200_btaa436-B14","first-page":"958","article-title":"MILIS: multiple instance learning with instance selection","volume":"33","author":"Fu","year":"2010","journal-title":"IEEE Trans. Pattern Anal. Mach. Intell"},{"key":"2024021913335994200_btaa436-B15","doi-asserted-by":"crossref","first-page":"2270","DOI":"10.1093\/bioinformatics\/bts447","article-title":"MALDIQUANT: a versatile R package for the analysis of mass spectrometry data","volume":"28","author":"Gibb","year":"2012","journal-title":"Bioinformatics"},{"key":"2024021913335994200_btaa436-B16","doi-asserted-by":"crossref","first-page":"3500","DOI":"10.1039\/C6SC03738K","article-title":"Deep learning and 3D-DESI imaging reveal the hidden metabolic heterogeneity of cancer","volume":"8","author":"Inglese","year":"2017","journal-title":"Chem. Sci"},{"key":"2024021913335994200_btaa436-B17","doi-asserted-by":"crossref","first-page":"4962","DOI":"10.1016\/j.jprot.2012.06.014","article-title":"Imaging mass spectrometry statistical analysis","volume":"75","author":"Jones","year":"2012","journal-title":"J. Proteomics"},{"key":"2024021913335994200_btaa436-B18","doi-asserted-by":"crossref","first-page":"44","DOI":"10.1016\/j.compmedimag.2014.11.010","article-title":"Computer-aided diagnosis from weak supervision: a benchmarking study","volume":"42","author":"Kandemir","year":"2015","journal-title":"Comput. Med. Imaging Graph"},{"key":"2024021913335994200_btaa436-B19","doi-asserted-by":"crossref","first-page":"893","DOI":"10.3892\/ijo.2014.2788","article-title":"MALDI TOF imaging mass spectrometry in clinical pathology: a valuable tool for cancer diagnostics (review)","volume":"46","author":"Kriegsmann","year":"2015","journal-title":"Int. J. Oncol"},{"key":"2024021913335994200_btaa436-B20","doi-asserted-by":"crossref","first-page":"957","DOI":"10.1016\/j.bbapap.2016.09.020","article-title":"An experimental guideline for the analysis of histologically heterogeneous tumors by MALDI-TOF mass spectrometry imaging","volume":"1865","author":"Lou","year":"2017","journal-title":"Biochim. Biophys. Acta Proteins Proteom"},{"key":"2024021913335994200_btaa436-B21","author":"Molnar","year":"2019"},{"key":"2024021913335994200_btaa436-B22","author":"Pedersen","year":"2019"},{"key":"2024021913335994200_btaa436-B23","doi-asserted-by":"crossref","first-page":"1854","DOI":"10.1021\/pr901008d","article-title":"Classification of HER2 receptor status in breast cancer tissues by MALDI imaging mass spectrometry","volume":"9","author":"Rauser","year":"2010","journal-title":"J. Proteome Res"},{"key":"2024021913335994200_btaa436-B24","doi-asserted-by":"crossref","first-page":"2352","DOI":"10.1162\/neco_a_00990","article-title":"Deep convolutional neural networks for image classification: a comprehensive review","volume":"29","author":"Rawat","year":"2017","journal-title":"Neural Comput"},{"key":"2024021913335994200_btaa436-B25","first-page":"1135","article-title":"\u201cWhy should I trust you?\u201d Explaining the predictions of any classifier","author":"Ribeiro","year":"2016"},{"key":"2024021913335994200_btaa436-B26","doi-asserted-by":"crossref","first-page":"2903","DOI":"10.1158\/0008-5472.CAN-16-3044","article-title":"Metabolic markers and statistical prediction of serous ovarian cancer aggressiveness by ambient ionization mass spectrometry imaging","volume":"77","author":"Sans","year":"2017","journal-title":"Cancer Res"},{"key":"2024021913335994200_btaa436-B27","doi-asserted-by":"crossref","first-page":"64","DOI":"10.1021\/ac504543v","article-title":"Mass spectrometry imaging of biomolecular information","volume":"87","author":"Spengler","year":"2015","journal-title":"Anal. Chem"},{"key":"2024021913335994200_btaa436-B28","doi-asserted-by":"crossref","DOI":"10.1117\/12.2512360","article-title":"Cancer detection in mass spectrometry imaging data by dilated convolutional neural networks","author":"van Kersbergen","year":"2019"},{"key":"2024021913335994200_btaa436-B29","doi-asserted-by":"crossref","first-page":"2690","DOI":"10.1039\/C7AN00565B","article-title":"Mass spectrometry imaging for clinical research-latest developments, applications, and current limitations","volume":"142","author":"Vaysse","year":"2017","journal-title":"Analyst"},{"key":"2024021913335994200_btaa436-B30","doi-asserted-by":"crossref","first-page":"218","DOI":"10.1002\/mas.21360","article-title":"Mass spectrometry imaging under ambient conditions","volume":"32","author":"Wu","year":"2013","journal-title":"Mass Spectrom. Rev"},{"key":"2024021913335994200_btaa436-B31","first-page":"1417","article-title":"Multiple instance boosting for object detection","author":"Zhang","year":"2006"},{"key":"2024021913335994200_btaa436-B32","first-page":"1249","article-title":"Multi-instance learning by treating instances as non-IID samples","author":"Zhou","year":"2009"}],"container-title":["Bioinformatics"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/36\/Supplement_1\/i300\/56702205\/bioinformatics_36_supplement1_i300.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article-pdf\/36\/Supplement_1\/i300\/56702205\/bioinformatics_36_supplement1_i300.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,2,19]],"date-time":"2024-02-19T13:43:24Z","timestamp":1708350204000},"score":1,"resource":{"primary":{"URL":"https:\/\/academic.oup.com\/bioinformatics\/article\/36\/Supplement_1\/i300\/5870486"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2020,7,1]]},"references-count":32,"journal-issue":{"issue":"Supplement_1","published-print":{"date-parts":[[2020,7,1]]}},"URL":"https:\/\/doi.org\/10.1093\/bioinformatics\/btaa436","relation":{},"ISSN":["1367-4803","1367-4811"],"issn-type":[{"value":"1367-4803","type":"print"},{"value":"1367-4811","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2020,7]]},"published":{"date-parts":[[2020,7,1]]}}}