Show simple item record

dc.contributor.authorPicon, Artzai
dc.contributor.authorMedela, Alfonso
dc.date.accessioned2020-12-18T16:21:07Z
dc.date.available2020-12-18T16:21:07Z
dc.date.issued2020
dc.identifier.citationPicon, Artzai, and Alfonso Medela. “Constellation Loss: Improving the Efficiency of Deep Metric Learning Loss Functions for the Optimal Embedding of Histopathological Images.” Journal of Pathology Informatics 11, no. 1 (2020): 38. doi:10.4103/jpi.jpi_41_20.en
dc.identifier.issn2229-5089en
dc.identifier.urihttp://hdl.handle.net/11556/1038
dc.description.abstractDeep learning diagnostic algorithms are proving comparable results with human experts in a wide variety of tasks, and they still require a huge amount of well-annotated data for training, which is often non affordable. Metric learning techniques have allowed a reduction in the required annotated data allowing few-shot learning over deep learning architectures. Aims and Objectives: In this work, we analyze the state-of-the-art loss functions such as triplet loss, contrastive loss, and multi-class N-pair loss for the visual embedding extraction of hematoxylin and eosin (H&E) microscopy images and we propose a novel constellation loss function that takes advantage of the visual distances of the embeddings of the negative samples and thus, performing a regularization that increases the quality of the extracted embeddings. Materials and Methods: To this end, we employed the public H&E imaging dataset from the University Medical Center Mannheim (Germany) that contains tissue samples from low-grade and high-grade primary tumors of digitalized colorectal cancer tissue slides. These samples are divided into eight different textures (1. tumour epithelium, 2. simple stroma, 3. complex stroma, 4. immune cells, 5. debris and mucus, 6. mucosal glands, 7. adipose tissue and 8. background,). The dataset was divided randomly into train and test splits and the training split was used to train a classifier to distinguish among the different textures with just 20 training images. The process was repeated 10 times for each loss function. Performance was compared both for cluster compactness and for classification accuracy on separating the aforementioned textures. Results: Our results show that the proposed loss function outperforms the other methods by obtaining more compact clusters (Davis-Boulding: 1.41 ± 0.08, Silhouette: 0.37 ± 0.02) and better classification capabilities (accuracy: 85.0 ± 0.6) over H and E microscopy images. We demonstrate that the proposed constellation loss can be successfully used in the medical domain in situations of data scarcity.en
dc.description.sponsorshipThis work was partially supported by PICCOLO project. This project has received funding from the European Union’s Horizon2020 Research and Innovation Programme under grant agreement No. 732111.en
dc.language.isoengen
dc.publisherWolters Kluwer Healthen
dc.rightsAttribution-NonCommercial-ShareAlike 4.0 International*
dc.rights.urihttp://creativecommons.org/licenses/by-nc-sa/4.0/*
dc.titleConstellation loss: Improving the efficiency of deep metric learning loss functions for the optimal embedding of histopathological imagesen
dc.typearticleen
dc.identifier.doi10.4103/jpi.jpi_41_20en
dc.relation.projectIDinfo:eu-repo/grantAgreement/EC/H2020/732111/EU/Multimodal highly-sensitive PhotonICs endoscope for improved in-vivo COLOn Cancer diagnosis and clinical decision Support/PICCOLOen
dc.rights.accessRightsopenAccessen
dc.subject.keywordsFew-shot learningen
dc.subject.keywordsHistopathologyen
dc.subject.keywordsMetric learningen
dc.identifier.essn2153-3539en
dc.issue.number1en
dc.journal.titleJournal of Pathology Informaticsen
dc.page.initial38en
dc.volume.number11en


Files in this item

Thumbnail

    Show simple item record

    Attribution-NonCommercial-ShareAlike 4.0 InternationalExcept where otherwise noted, this item's license is described as Attribution-NonCommercial-ShareAlike 4.0 International