{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/cia-net-robust-nuclei-instance-segmentation","title":"CIA-Net: Robust Nuclei Instance Segmentation with Contour-aware Information Aggregation","arxiv_id":"1903.05358","date":"2019-03-13","proceeding":null,"authors":["Yanning Zhou","Omer Fahri Onder","Qi Dou","Efstratios Tsougenis","Hao Chen","Pheng-Ann Heng"],"abstract":"Accurate segmenting nuclei instances is a crucial step in computer-aided\nimage analysis to extract rich features for cellular estimation and following\ndiagnosis as well as treatment. While it still remains challenging because the\nwide existence of nuclei clusters, along with the large morphological variances\namong different organs make nuclei instance segmentation susceptible to\nover-/under-segmentation. Additionally, the inevitably subjective annotating\nand mislabeling prevent the network learning from reliable samples and\neventually reduce the generalization capability for robustly segmenting unseen\norgan nuclei. To address these issues, we propose a novel deep neural network,\nnamely Contour-aware Informative Aggregation Network (CIA-Net) with multi-level\ninformation aggregation module between two task-specific decoders. Rather than\nindependent decoders, it leverages the merit of spatial and texture\ndependencies between nuclei and contour by bi-directionally aggregating\ntask-specific features. Furthermore, we proposed a novel smooth truncated loss\nthat modulates losses to reduce the perturbation from outliers. Consequently,\nthe network can focus on learning from reliable and informative samples, which\ninherently improves the generalization capability. Experiments on the 2018\nMICCAI challenge of Multi-Organ-Nuclei-Segmentation validated the effectiveness\nof our proposed method, surpassing all the other 35 competitive teams by a\nsignificant margin.","url_abs":"http://arxiv.org/abs/1903.05358v1","url_pdf":"http://arxiv.org/pdf/1903.05358v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[],"tasks":[{"task_slug":"instance-segmentation","task_name":"Instance Segmentation"},{"task_slug":"multi-tissue-nucleus-segmentation","task_name":"Multi-tissue Nucleus Segmentation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/multi-tissue-nucleus-segmentation-on-kumar","task":"Multi-tissue Nucleus Segmentation","dataset":"Kumar","model":"CIA-Net (e)","rank_in_archive_order":5,"of":18,"metrics":{"Dice":"0.818","Hausdorff Distance (mm)":"57.7"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1903.05358","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}