{"entity": "journal", "iuid": "c28a0ae59484457dbdb9ce69aa8850d6", "timestamp": "2026-06-15T07:37:08.336Z", "links": {"self": {"href": "https://publications.scilifelab.se/journal/IEEE%20J%20Biomed%20Health%20Inform.json"}, "display": {"href": "https://publications.scilifelab.se/journal/IEEE%20J%20Biomed%20Health%20Inform"}}, "title": "IEEE J Biomed Health Inform", "issn": "2168-2208", "issn-l": null, "publications_count": 2, "publications": [{"entity": "publication", "iuid": "6f1b382028f74b88ac1be4c6c2cb017d", "links": {"self": {"href": "https://publications.scilifelab.se/publication/6f1b382028f74b88ac1be4c6c2cb017d.json"}, "display": {"href": "https://publications.scilifelab.se/publication/6f1b382028f74b88ac1be4c6c2cb017d"}}, "title": "SimSearch: A Human-in-The-Loop Learning Framework for Fast Detection of Regions of Interest in Microscopy Images.", "authors": [{"family": "Gupta", "given": "Ankit", "initials": "A", "orcid": "0000-0002-9961-1041", "researcher": {"href": "https://publications.scilifelab.se/researcher/ef29c8c1354a4c6d9ca8b0edaa034710.json"}}, {"family": "Sabirsh", "given": "Alan", "initials": "A", "orcid": "0000-0001-5310-0281", "researcher": {"href": "https://publications.scilifelab.se/researcher/0000d5f644194d6f8daf8b7f5dd0540c.json"}}, {"family": "Wahlby", "given": "Carolina", "initials": "C", "orcid": "0000-0002-4139-7003", "researcher": {"href": "https://publications.scilifelab.se/researcher/c50194fbc8524d95b7152663ccf17f29.json"}}, {"family": "Sintorn", "given": "Ida-Maria", "initials": "IM", "orcid": "0000-0002-8307-7411", "researcher": {"href": "https://publications.scilifelab.se/researcher/8ff79494ec3842ffaa7448d2e3277f6b.json"}}], "type": "journal article", "published": "2022-08-00", "journal": {"title": "IEEE J Biomed Health Inform", "issn": "2168-2208", "issn-l": null, "volume": "26", "issue": "8", "pages": "4079-4089"}, "abstract": "Large-scale microscopy-based experiments often result in images with rich but sparse information content. An experienced microscopist can visually identify regions of interest (ROIs), but this becomes a cumbersome task with large datasets. Here we present SimSearch, a framework for quick and easy user-guided training of a deep neural model aimed at fast detection of ROIs in large-scale microscopy experiments.\n\nThe user manually selects a small number of patches representing different classes of ROIs. This is followed by feature extraction using a pre-trained deep-learning model, and interactive patch selection pruning, resulting in a smaller set of clean (user approved) and larger set of noisy (unapproved) training patches of ROIs and background. The pre-trained deep-learning model is thereafter first trained on the large set of noisy patches, followed by refined training using the clean patches.\n\nThe framework is evaluated on fluorescence microscopy images from a large-scale drug screening experiment, brightfield images of immunohistochemistry-stained patient tissue samples, and malaria-infected human blood smears, as well as transmission electron microscopy images of cell sections. Compared to state-of-the-art and manual/visual assessment, the results show similar performance with maximal flexibility and minimal a priori information and user interaction.\n\nSimSearch quickly adapts to different data sets, which demonstrates the potential to speed up many microscopy-based experiments based on a small amount of user interaction.\n\nSimSearch can help biologists quickly extract informative regions and perform analyses on large datasets helping increase the throughput in a microscopy experiment.", "doi": "10.1109/JBHI.2022.3177602", "pmid": "35609108", "labels": {"BioImage Informatics": "Technology development", "Bioinformatics (NBIS)": "Technology development"}, "xrefs": [], "notes": [], "created": "2022-11-02T06:18:25.917Z", "modified": "2023-06-19T12:58:45.375Z"}, {"entity": "publication", "iuid": "539bb94475c4453f951e238b87231dd6", "links": {"self": {"href": "https://publications.scilifelab.se/publication/539bb94475c4453f951e238b87231dd6.json"}, "display": {"href": "https://publications.scilifelab.se/publication/539bb94475c4453f951e238b87231dd6"}}, "title": "Measuring Domain Shift for Deep Learning in Histopathology.", "authors": [{"family": "Stacke", "given": "Karin", "initials": "K"}, {"family": "Eilertsen", "given": "Gabriel", "initials": "G"}, {"family": "Unger", "given": "Jonas", "initials": "J"}, {"family": "Lundstrom", "given": "Claes", "initials": "C"}], "type": "journal article", "published": "2021-02-00", "journal": {"title": "IEEE J Biomed Health Inform", "issn": "2168-2208", "volume": "25", "issue": "2", "pages": "325-336", "issn-l": null}, "abstract": "The high capacity of neural networks allows fitting models to data with high precision, but makes generalization to unseen data a challenge. If a domain shift exists, i.e. differences in image statistics between training and test data, care needs to be taken to ensure reliable deployment in real-world scenarios. In digital pathology, domain shift can be manifested in differences between whole-slide images, introduced by for example differences in acquisition pipeline - between medical centers or over time. In order to harness the great potential presented by deep learning in histopathology, and ensure consistent model behavior, we need a deeper understanding of domain shift and its consequences, such that a model's predictions on new data can be trusted. This work focuses on the internal representation learned by trained convolutional neural networks, and shows how this can be used to formulate a novel measure - the representation shift - for quantifying the magnitude of model-specific domain shift. We perform a study on domain shift in tumor classification of hematoxylin and eosin stained images, by considering different datasets, models, and techniques for preparing data in order to reduce the domain shift. The results show how the proposed measure has a high correlation with drop in performance when testing a model across a large number of different types of domain shifts, and how it improves on existing techniques for measuring data shift and uncertainty. The proposed measure can reveal how sensitive a model is to domain variations, and can be used to detect new data that a model will have problems generalizing to. We see techniques for measuring, understanding and overcoming the domain shift as a crucial step towards reliable use of deep learning in the future clinical pathology applications.", "doi": "10.1109/JBHI.2020.3032060", "pmid": "33085623", "labels": {"AIDA Data Hub": "Service", "Bioinformatics (NBIS)": "Service"}, "xrefs": [], "notes": [], "created": "2022-12-01T15:38:19.734Z", "modified": "2022-12-01T15:38:19.738Z"}], "created": "2022-11-02T06:18:25.924Z", "modified": "2022-11-02T06:18:25.924Z"}