{"database":"biostudies-literature","file_versions":[],"scores":null,"additional":{"omics_type":["Unknown"],"volume":["21(1)"],"submitter":["Arnay R"],"pubmed_abstract":["The quantification and identification of components in archaeological micromorphology remain subjective and challenging, particularly for early-career researchers. To address this, we developed a deep learning tool for the automatic segmentation of three materials commonly found in Palaeolithic contexts and thin sections: bone, charcoal, and lithic fine-grained debitage (flint and obsidian). Using high-resolution photomicrographs of 57 thin sections in plane-polarised and cross-polarised light, we trained and evaluated state-of-the-art convolutional neural networks (CNNs) for material segmentation. The best-performing configuration, a U-Net with an InceptionV4 encoder, achieved mean intersection over union (IoU) scores of 0.96 for flint/obsidian, 0.80 for bone, and 0.82 for charcoal. The models also classified the relative abundance of each material with balanced accuracies of 0.99 for flint/obsidian, 0.92 for bone, and 0.85 for charcoal. These results demonstrate the potential of deep learning to enhance objectivity, accuracy, and reproducibility in archaeological micromorphology, providing a valuable resource for future geoarchaeological research."],"journal":["PloS one"],"pagination":["e0340353"],"full_dataset_link":["https://www.ebi.ac.uk/biostudies/studies/S-EPMC12818656"],"repository":["biostudies-literature"],"pubmed_title":["Improving micromorphological analysis with CNN-based segmentation of flint/obsidian, bone and charcoal."],"pmcid":["PMC12818656"],"pubmed_authors":["Garcia-Villa P","Rueda-Saiz S","Hernandez-Aceituno J","Mallol C","Arnay R"],"additional_accession":[]},"is_claimable":false,"name":"Improving micromorphological analysis with CNN-based segmentation of flint/obsidian, bone and charcoal.","description":"The quantification and identification of components in archaeological micromorphology remain subjective and challenging, particularly for early-career researchers. To address this, we developed a deep learning tool for the automatic segmentation of three materials commonly found in Palaeolithic contexts and thin sections: bone, charcoal, and lithic fine-grained debitage (flint and obsidian). Using high-resolution photomicrographs of 57 thin sections in plane-polarised and cross-polarised light, we trained and evaluated state-of-the-art convolutional neural networks (CNNs) for material segmentation. The best-performing configuration, a U-Net with an InceptionV4 encoder, achieved mean intersection over union (IoU) scores of 0.96 for flint/obsidian, 0.80 for bone, and 0.82 for charcoal. The models also classified the relative abundance of each material with balanced accuracies of 0.99 for flint/obsidian, 0.92 for bone, and 0.85 for charcoal. These results demonstrate the potential of deep learning to enhance objectivity, accuracy, and reproducibility in archaeological micromorphology, providing a valuable resource for future geoarchaeological research.","dates":{"release":"2026-01-01T00:00:00Z","publication":"2026","modification":"2026-06-06T19:45:41.561Z","creation":"2026-06-04T03:12:32.296Z"},"accession":"S-EPMC12818656","cross_references":{"pubmed":["41557713"],"doi":["10.1371/journal.pone.0340353"]}}