{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,4,22]],"date-time":"2026-04-22T04:04:50Z","timestamp":1776830690902,"version":"3.51.2"},"reference-count":33,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2025,8,12]],"date-time":"2025-08-12T00:00:00Z","timestamp":1754956800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/100007709","name":"Michigan State University","doi-asserted-by":"publisher","id":[{"id":"10.13039\/100007709","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100005825","name":"National Institute of Food and Agriculture","doi-asserted-by":"publisher","award":["2020-67021-32855"],"award-info":[{"award-number":["2020-67021-32855"]}],"id":[{"id":"10.13039\/100005825","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/100005825","name":"National Institute of Food and Agriculture","doi-asserted-by":"publisher","award":["2021-67021-34256"],"award-info":[{"award-number":["2021-67021-34256"]}],"id":[{"id":"10.13039\/100005825","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Artif. Intell."],"abstract":"<jats:p>AI-enabled microscopy is emerging for rapid bacterial classification, yet its utility remains limited in dynamic or resource-limited settings due to imaging variability. This study aims to enhance the generalizability of AI microscopy using domain adaptation techniques. Six bacterial species, including three Gram-positive (<jats:italic>Bacillus coagulans, Bacillus subtilis, Listeria innocua<\/jats:italic>) and three Gram-negative (<jats:italic>Escherichia coli, Salmonella<\/jats:italic> Enteritidis, <jats:italic>Salmonella<\/jats:italic> Typhimurium), were grown into microcolonies on soft tryptic soy agar plates at 37\u00b0C for 3\u20135 h. Images were acquired under varying microscopy modalities and magnifications. Domain-adversarial neural networks (DANNs) addressed single-target domain variations and multi-DANNs (MDANNs) handled multiple domains simultaneously. EfficientNetV2 backbone provided fine-grained feature extraction suitable for small targets, with few-shot learning enhancing scalability in data-limited domains. The source domain contained 105 images per species (<jats:italic>n<\/jats:italic> = 630) collected under optimal conditions (phase contrast, 60 \u00d7 magnification, 3-h incubation). Target domains introduced variations in modality (brightfield, BF), lower magnification (20 \u00d7 ), and extended incubation (20x-5h), each with &amp;lt; 5 labeled training images per species (<jats:italic>n<\/jats:italic> \u2264 30) and test datasets of 60\u201390 images. DANNs improved target domain classification accuracy by up to 54.5% for 20 \u00d7 (34.4% to 88.9%), 43.3% for 20x-5h (40.0% to 83.3%), and 31.7% for BF (43.4% to 73.3%), with minimal accuracy loss in the source domain. MDANNs further improved accuracy in the BF domain from 73.3% to 76.7%. Feature visualizations by Grad-CAM and t-SNE validated the model's ability to learn domain-invariant features across conditions. This study presents a scalable and adaptable framework for bacterial classification, extending the utility of microscopy to decentralized and resource-limited settings where imaging variability often challenges performance.<\/jats:p>","DOI":"10.3389\/frai.2025.1632344","type":"journal-article","created":{"date-parts":[[2025,8,12]],"date-time":"2025-08-12T05:32:06Z","timestamp":1754976726000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":5,"title":["Enhancing AI microscopy for foodborne bacterial classification using adversarial domain adaptation to address optical and biological variability"],"prefix":"10.3389","volume":"8","author":[{"given":"Siddhartha","family":"Bhattacharya","sequence":"first","affiliation":[]},{"given":"Aarham","family":"Wasit","sequence":"additional","affiliation":[]},{"given":"J Mason","family":"Earles","sequence":"additional","affiliation":[]},{"given":"Nitin","family":"Nitin","sequence":"additional","affiliation":[]},{"given":"Jiyoon","family":"Yi","sequence":"additional","affiliation":[]}],"member":"1965","published-online":{"date-parts":[[2025,8,12]]},"reference":[{"key":"B1","doi-asserted-by":"publisher","first-page":"125","DOI":"10.3390\/info11020125","article-title":"Albumentations: fast and flexible image augmentations","volume":"11","author":"Buslaev","year":"2020","journal-title":"Information"},{"key":"B2","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00072","article-title":"\u201cProgressive feature alignment for unsupervised domain adaptation,\u201d","author":"Chen","year":"2019","journal-title":"2019 IEEE\/CVF Conference on Computer Vision and Pattern Recognition (CVPR)"},{"key":"B3","doi-asserted-by":"publisher","first-page":"110413","DOI":"10.1016\/j.foodcont.2024.110413","article-title":"Microscopic identification of foodborne bacterial pathogens based on deep learning method","volume":"161","author":"Chen","year":"2024","journal-title":"Food Control"},{"key":"B4","doi-asserted-by":"publisher","first-page":"e2180","DOI":"10.7717\/peerj-cs.2180","article-title":"Bacterial image analysis using multi-task deep learning approaches for clinical microscopy","volume":"10","author":"Chin","year":"2024","journal-title":"PeerJ Comput. Sci"},{"key":"B5","doi-asserted-by":"publisher","DOI":"10.1016\/j.cll.2015.02.002","article-title":"Salmonella, shigella, and yersinia","author":"Dekker","year":"2015","journal-title":"Clin. Lab. Med"},{"key":"B6","doi-asserted-by":"publisher","first-page":"3106","DOI":"10.1111\/1541-4337.12618","article-title":"Microbial detection and identification methods: bench top assays to omics approaches","volume":"19","author":"Ferone","year":"2020","journal-title":"Compr. Rev. Food Sci. Food Saf"},{"key":"B7","first-page":"1","article-title":"Domain-adversarial training of neural networks","volume":"17","author":"Ganin","year":"2016","journal-title":"J. Mach. Learn. Res"},{"key":"B8","article-title":"\u201cGeneralizing microscopy image labeling via layer-matching adversarial domain adaptation,\u201d","author":"Goyal","year":"2024","journal-title":"ICML'24 Workshop ML for Life and Material Science: From Theory to Industry Applications"},{"key":"B9","article-title":"\u201cStochastic neighbor embedding,\u201d","author":"Hinton","year":"2002","journal-title":"Advances in Neural Information Processing Systems, volume 15"},{"key":"B10","doi-asserted-by":"publisher","first-page":"4","DOI":"10.1089\/fpd.2023.0157","article-title":"Economic burden of foodborne illnesses acquired in the United States","volume":"22","author":"Hoffmann","year":"2024","journal-title":"Foodborne Pathog. Dis"},{"key":"B11","doi-asserted-by":"publisher","first-page":"fuaa062","DOI":"10.1093\/femsre\/fuaa062","article-title":"Advances and opportunities in image analysis of bacterial cells and communities","volume":"45","author":"Jeckel","year":"2020","journal-title":"FEMS Microbiol. Rev"},{"key":"B12","doi-asserted-by":"publisher","first-page":"e01828","DOI":"10.1128\/aem.01828-22","article-title":"Accelerating the detection of bacteria in food using artificial intelligence and optical imaging","volume":"89","author":"Ma","year":"2023","journal-title":"Appl. Environ. Microbiol"},{"key":"B13","doi-asserted-by":"publisher","first-page":"463","DOI":"10.1007\/s12551-022-00949-3","article-title":"Deep learning-based image processing in optical microscopy","volume":"14","author":"Melanthota","year":"2022","journal-title":"Biophys. Rev"},{"key":"B14","doi-asserted-by":"publisher","first-page":"42","DOI":"10.1109\/TMI.2022.3203022","article-title":"Domain adapted multitask learning for segmenting amoeboid cells in microscopy","volume":"42","author":"Mukherjee","year":"2023","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B15","doi-asserted-by":"publisher","first-page":"111014","DOI":"10.1016\/j.jfoodeng.2022.111014","article-title":"End-to-end prediction of uniaxial compression profiles of apples during in vitro digestion using time-series micro-computed tomography and deep learning","volume":"325","author":"Olenskyj","year":"2022","journal-title":"J. Food Eng"},{"key":"B16","doi-asserted-by":"publisher","first-page":"914","DOI":"10.1016\/j.solener.2021.03.058","article-title":"Segmentation of cell-level anomalies in electroluminescence images of photovoltaic modules","volume":"220","author":"Otamendi","year":"2021","journal-title":"Solar Energy"},{"key":"B17","doi-asserted-by":"publisher","first-page":"100430","DOI":"10.1016\/j.jfp.2024.100430","article-title":"Detection of viable but nonculturable E. coli induced by low-level antimicrobials using AI-enabled hyperspectral microscopy","volume":"88","author":"Papa","year":"2025","journal-title":"J. Food Prot"},{"key":"B18","doi-asserted-by":"publisher","first-page":"3934","DOI":"10.1609\/aaai.v32i1.11767","article-title":"Multi-adversarial domain adaptation","volume":"32","author":"Pei","year":"2018","journal-title":"Proc. AAAI Conf. Artif. Intell"},{"key":"B19","doi-asserted-by":"publisher","first-page":"e190","DOI":"10.1017\/S0950268821001722","article-title":"Food recalls associated with foodborne disease outbreaks, United States, 2006\u20132016","volume":"149","author":"Qiu","year":"2021","journal-title":"Epidemiol. Infect"},{"key":"B20","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1109\/ISBI56570.2024.10635473","article-title":"\u201cBiointel: real-time bacteria identification using microscopy imaging,\u201d","volume-title":"2024 IEEE International Symposium on Biomedical Imaging (ISBI)","author":"Ramesh","year":"2024"},{"key":"B21","doi-asserted-by":"publisher","first-page":"957","DOI":"10.4315\/0362-028X-46.11.957","article-title":"Comparison of enrichment and plating media for recovery of virulent strains of Yersinia enterocolitica from inoculated beef stew","volume":"46","author":"Schiemann","year":"1983","journal-title":"J. Food Prot"},{"key":"B22","doi-asserted-by":"publisher","first-page":"671","DOI":"10.1038\/nmeth.2089","article-title":"NIH image to ImageJ: 25 years of image analysis","volume":"9","author":"Schneider","year":"2012","journal-title":"Nat. Methods"},{"key":"B23","first-page":"618","article-title":"\u201cGrad-CAM: Visual explanations from deep networks via gradient-based localization,\u201d","volume-title":"Proceedings of the IEEE International Conference on Computer Vision","author":"Selvaraju","year":"2017"},{"key":"B24","doi-asserted-by":"publisher","first-page":"958","DOI":"10.1109\/ICDAR.2003.1227801","article-title":"Best practices for convolutional neural networks applied to visual document analysis","volume":"3","author":"Simard","year":"2003","journal-title":"In Seventh International Conference on Document Analysis and Recognition (ICDAR)"},{"key":"B25","first-page":"6105","article-title":"\u201cEfficientnet: rethinking model scaling for convolutional neural networks,\u201d","volume-title":"International Conference on Machine Learning","author":"Tan","year":"2019"},{"key":"B26","first-page":"10096","article-title":"\u201cEfficientnetv2: smaller models and faster training,\u201d","volume-title":"International Conference on Machine Learning","author":"Tan","year":"2021"},{"key":"B27","doi-asserted-by":"publisher","first-page":"2897","DOI":"10.1109\/TMI.2020.3046334","article-title":"Multi-task multi-domain learning for digital staining and classification of leukocytes","volume":"40","author":"Tomczak","year":"2021","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B28","doi-asserted-by":"publisher","first-page":"e13158","DOI":"10.1111\/jfs.13158","article-title":"Rapid counting of coliforms and Escherichia coli by deep learning-based classifier","volume":"44","author":"Wakabayashi","year":"2024","journal-title":"J. Food Saf"},{"key":"B29","first-page":"1","article-title":"\u201cSemi-supervised cell instance segmentation for multi-modality microscope images,\u201d","author":"Wang","year":"2023","journal-title":"Proceedings of The Cell Segmentation Challenge in Multi-modality High-Resolution Microscopy Images, volume 212 of Proceedings of Machine Learning Research"},{"key":"B30","doi-asserted-by":"publisher","first-page":"1200994","DOI":"10.3389\/frai.2023.1200994","article-title":"Machine learning algorithms in microbial classification: a comparative analysis","volume":"6","author":"Wu","year":"2023","journal-title":"Front. Artif. Intell"},{"key":"B31","doi-asserted-by":"publisher","first-page":"2880","DOI":"10.1109\/TMI.2020.3042789","article-title":"Bidirectional mapping-based domain adaptation for nucleus detection in cross-modality microscopy images","volume":"40","author":"Xing","year":"2021","journal-title":"IEEE Trans. Med. Imaging"},{"key":"B32","doi-asserted-by":"publisher","first-page":"103416","DOI":"10.1016\/j.bspc.2021.103416","article-title":"Te-yolof: Tiny and efficient yolof for blood cell detection","volume":"73","author":"Xu","year":"2022","journal-title":"Biomed. Signal Process. Control"},{"key":"B33","article-title":"\u201cAdversarial multiple source domain adaptation,\u201d","author":"Zhao","year":"2018","journal-title":"Advances in Neural Information Processing Systems, volume 31"}],"container-title":["Frontiers in Artificial Intelligence"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1632344\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,8,12]],"date-time":"2025-08-12T05:32:08Z","timestamp":1754976728000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/frai.2025.1632344\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,8,12]]},"references-count":33,"alternative-id":["10.3389\/frai.2025.1632344"],"URL":"https:\/\/doi.org\/10.3389\/frai.2025.1632344","relation":{},"ISSN":["2624-8212"],"issn-type":[{"value":"2624-8212","type":"electronic"}],"subject":[],"published":{"date-parts":[[2025,8,12]]},"article-number":"1632344"}}