{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,30]],"date-time":"2026-07-30T14:27:17Z","timestamp":1785421637793,"version":"3.56.0"},"reference-count":30,"publisher":"MDPI AG","issue":"7","license":[{"start":{"date-parts":[[2026,6,29]],"date-time":"2026-06-29T00:00:00Z","timestamp":1782691200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"name":"National Science and Technology Council (NSTC) of Taiwan","award":["112-2410-H-415-012-MY2"],"award-info":[{"award-number":["112-2410-H-415-012-MY2"]}]},{"name":"National Science and Technology Council (NSTC) of Taiwan","award":["114-2410-H-415-033-MY2"],"award-info":[{"award-number":["114-2410-H-415-033-MY2"]}]},{"award":["112-2410-H-415-012-MY2"],"award-info":[{"award-number":["112-2410-H-415-012-MY2"]}],"id":[{"id":"https:\/\/ror.org\/02kv4zf79","id-type":"ROR","asserted-by":"publisher"}]},{"award":["114-2410-H-415-033-MY2"],"award-info":[{"award-number":["114-2410-H-415-033-MY2"]}],"id":[{"id":"https:\/\/ror.org\/02kv4zf79","id-type":"ROR","asserted-by":"publisher"}]},{"name":"Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation","award":["DTCRD 112(2)-E-13"],"award-info":[{"award-number":["DTCRD 112(2)-E-13"]}]},{"award":["DTCRD 112(2)-E-13"],"award-info":[{"award-number":["DTCRD 112(2)-E-13"]}],"id":[{"id":"https:\/\/ror.org\/04rbvc675","id-type":"ROR","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Bioengineering"],"abstract":"<jats:p>Medical image segmentation requires dense pixel-level annotations, making large-scale dataset construction expensive and motivating research into data-efficient training. The main objective of this paper is to determine whether fusing two complementary pretrained image encoders into a single similarity space can make training-free dataset pruning robust across heterogeneous polyp segmentation domains and to quantify that robustness against a comprehensive panel of baselines. To achieve this, we propose Multi-Encoder Diverse Pruning (MEDP), a training-free dataset-pruning method. MEDP fuses features from an ImageNet-pretrained ResNet-18 and a self-supervised DINOv2 ViT-S\/14 into a single 896-D similarity space. It partitions the training pool via Louvain modularity maximization and selects per-community samples via maximal-marginal-relevance (MMR) ranking, which effectively balances eigenvector centrality with feature-space diversity. We benchmarked MEDP against 12 baselines at a 20% retention ratio across three polyp segmentation settings (Kvasir-SEG, CVC-ClinicDB, and a Combined cross-domain pool) using a standard 5-level UNet. Based on approximately 330 controlled training runs, the results demonstrate that MEDP achieves the highest mean test Dice of 0.7324 on the most challenging Combined cross-domain pool, significantly outperforming uniform random sampling (Cohen\u2019s d = +1.79, paired Wilcoxon p = 0.002). Conversely, all hand-crafted structure-aware variants failed to outperform uniform random sampling. These findings confirm that combining multi-encoder features with MMR diversity provides a simple and effective strategy for improving robustness across heterogeneous medical imaging settings and that the choice of pretrained image encoder is the dominant factor in segmentation-aware pruning.<\/jats:p>","DOI":"10.3390\/bioengineering13070759","type":"journal-article","created":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T05:58:04Z","timestamp":1782799084000},"page":"759","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Cross-Domain Robust Pruning for Polyp Segmentation: Multi-Encoder Feature Fusion Beats Single-Encoder Baselines"],"prefix":"10.3390","volume":"13","author":[{"given":"Chia-Pei","family":"Tang","sequence":"first","affiliation":[{"name":"Division of Gastroenterology, Department of Internal Medicine, Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Chiayi City 622401, Taiwan"},{"name":"School of Medicine, Tzu Chi University, Hualien City 970374, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-1249-1655","authenticated-orcid":false,"given":"Hong-Yi","family":"Chang","sequence":"additional","affiliation":[{"name":"Department of Management Information System, National Chiayi University, Chiayi City 600355, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Tzu-Shan","family":"Chang","sequence":"additional","affiliation":[{"name":"Institute of Information Security, National Tsing Hua University, Hsinchu City 600355, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yu-Chieh","family":"Chang","sequence":"additional","affiliation":[{"name":"Department of Information Management, National Yunlin University of Science and Technology, Douliu City 64002, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Chia-Hsin","family":"Cheng","sequence":"additional","affiliation":[{"name":"Department of Medical Research, Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Chiayi City 622401, Taiwan"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1968","published-online":{"date-parts":[[2026,6,29]]},"reference":[{"key":"ref_1","unstructured":"Jha, D., Smedsrud, P.H., Riegler, M.A., Halvorsen, P., de Lange, T., Johansen, D., and Johansen, H.D. (2020, January 5\u20138). Kvasir-SEG: A Segmented Polyp Dataset. Proceedings of the 26th International Conference on MultiMedia Modeling (MMM 2020), Daejeon, Republic of Korea."},{"key":"ref_2","doi-asserted-by":"crossref","first-page":"99","DOI":"10.1016\/j.compmedimag.2015.02.007","article-title":"WM-DOVA maps for accurate polyp highlighting in colonoscopy: Validation vs. saliency maps from physicians","volume":"43","author":"Bernal","year":"2015","journal-title":"Comput. Med. Imaging Graph."},{"key":"ref_3","unstructured":"Sener, O., and Savarese, S. (May, January 30). Active Learning for Convolutional Neural Networks: A Core-Set Approach. Proceedings of the 6th International Conference on Learning Representations (ICLR 2018), Vancouver, BC, Canada."},{"key":"ref_4","doi-asserted-by":"crossref","unstructured":"Guo, C., Zhao, B., and Bai, Y. (2022, January 22\u201324). DeepCore: A Comprehensive Library for Coreset Selection in Deep Learning. Proceedings of the 33rd International Conference on Database and Expert Systems Applications (DEXA 2022), Vienna, Austria.","DOI":"10.1007\/978-3-031-12423-5_14"},{"key":"ref_5","unstructured":"Rahman, M.M., and Marculescu, R. (2025, January 9\u201311). Training-Free Dataset Pruning for Polyp Segmentation via Community Detection in Similarity Networks. Proceedings of the Medical Imaging with Deep Learning (MIDL), Salt Lake City, UT, USA."},{"key":"ref_6","doi-asserted-by":"crossref","unstructured":"He, K., Zhang, X., Ren, S., and Sun, J. (2016, January 27\u201330). Deep Residual Learning for Image Recognition. Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, NV, USA.","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"P10008","DOI":"10.1088\/1742-5468\/2008\/10\/P10008","article-title":"Fast unfolding of communities in large networks","volume":"2008","author":"Blondel","year":"2008","journal-title":"J. Stat. Mech. Theory Exp."},{"key":"ref_8","doi-asserted-by":"crossref","unstructured":"Ronneberger, O., Fischer, P., and Brox, T. (2015, January 5\u20139). U-Net: Convolutional Networks for Biomedical Image Segmentation. Proceedings of the Medical Image Computing and Computer-Assisted Intervention (MICCAI), Munich, Germany.","DOI":"10.1007\/978-3-319-24574-4_28"},{"key":"ref_9","doi-asserted-by":"crossref","unstructured":"Fan, D.-P., Ji, G.-P., Zhou, T., Chen, G., Fu, H., Shen, J., and Shao, L. (2020, January 4\u20138). PraNet: Parallel Reverse Attention Network for Polyp Segmentation. Proceedings of the 23rd International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2020), Lima, Peru.","DOI":"10.1007\/978-3-030-59725-2_26"},{"key":"ref_10","doi-asserted-by":"crossref","first-page":"9150015","DOI":"10.26599\/AIR.2023.9150015","article-title":"Polyp-PVT: Polyp Segmentation with Pyramid Vision Transformers","volume":"2","author":"Dong","year":"2023","journal-title":"CAAI Artif. Intell. Res."},{"key":"ref_11","doi-asserted-by":"crossref","first-page":"654","DOI":"10.1038\/s41467-024-44824-z","article-title":"Segment Anything in Medical Images","volume":"15","author":"Ma","year":"2024","journal-title":"Nat. Commun."},{"key":"ref_12","unstructured":"Oquab, M., Darcet, T., Moutakanni, T., Vo, H., Szafraniec, M., Khalidov, V., Fernandez, P., Haziza, D., Massa, F., and El-Nouby, A. (2024). DINOv2: Learning Robust Visual Features without Supervision. arXiv."},{"key":"ref_13","unstructured":"Paul, M., Ganguli, S., and Dziugaite, G.K. (2021, January 6\u201314). Deep Learning on a Data Diet: Finding Important Examples Early in Training. Proceedings of the Advances in Neural Information Processing Systems (NeurIPS 34), Virtual."},{"key":"ref_14","unstructured":"Toneva, M., Sordoni, A., Tachet des Combes, R., Trischler, A., Bengio, Y., and Gordon, G.J. (2019, January 6\u20139). An Empirical Study of Example Forgetting during Deep Neural Network Learning. Proceedings of the 7th International Conference on Learning Representations (ICLR 2019), New Orleans, LA, USA."},{"key":"ref_15","unstructured":"Xia, X., Liu, J., Yu, J., Shen, X., Han, B., and Liu, T. (2023, January 1\u20135). Moderate Coreset: A Universal Method of Data Selection for Real-World Data-Efficient Deep Learning. Proceedings of the 11th International Conference on Learning Representations (ICLR 2023), Kigali, Rwanda."},{"key":"ref_16","unstructured":"Dai, Y., Xiao, L., Tsang, I.W., and He, Y. (2025, January 24\u201328). Training-Free Dataset Pruning for Instance Segmentation. Proceedings of the 13th International Conference on Learning Representations (ICLR 2025), Singapore."},{"key":"ref_17","doi-asserted-by":"crossref","unstructured":"Carbonell, J.G., and Goldstein, J. (1998, January 24\u201328). The Use of MMR, Diversity-Based Reranking for Reordering Documents and Producing Summaries. Proceedings of the 21st Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR \u201998), Melbourne, Australia.","DOI":"10.1145\/290941.291025"},{"key":"ref_18","unstructured":"Darcet, T., Oquab, M., Mairal, J., and Bojanowski, P. (2024, January 7\u201311). Vision Transformers Need Registers. Proceedings of the 12th International Conference on Learning Representations (ICLR 2024), Vienna, Austria."},{"key":"ref_19","doi-asserted-by":"crossref","unstructured":"Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A.C., and Lo, W.-Y. (2023, January 2\u20136). Segment Anything. Proceedings of the 2023 IEEE\/CVF International Conference on Computer Vision (ICCV), Paris, France.","DOI":"10.1109\/ICCV51070.2023.00371"},{"key":"ref_20","unstructured":"Jha, D., Tomar, N.K., Sharma, V., and Bagci, U. (2023, January 10\u201312). TransNetR: Transformer-based Residual Network for Polyp Segmentation with Multi-Center Out-of-Distribution Testing. Proceedings of the Medical Imaging with Deep Learning (MIDL 2023), Nashville, TN, USA."},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1007\/s44267-024-00071-w","article-title":"A Survey on Deep Learning for Polyp Segmentation: Techniques, Challenges and Future Trends","volume":"3","author":"Mei","year":"2025","journal-title":"Vis. Intell."},{"key":"ref_22","unstructured":"Oh, A.H., Agarwal, A., Belgrave, D., and Cho, K. (2022). Beyond neural scaling laws: Beating power law scaling via data pruning. Advances in Neural Information Processing Systems 35, Curran Associates, Inc."},{"key":"ref_23","unstructured":"Zheng, H., Liu, R., Lai, F., and Prakash, A. (2023, January 1\u20135). Coverage-centric coreset selection for high pruning rates. Proceedings of the 11th International Conference on Learning Representations (ICLR 2023), Kigali, Rwanda."},{"key":"ref_24","unstructured":"Maharana, A., Yadav, P., and Bansal, M. (2024, January 7\u201311). D2 Pruning: Message passing for balancing diversity and difficulty in data pruning. Proceedings of the 12th International Conference on Learning Representations (ICLR 2024), Vienna, Austria."},{"key":"ref_25","unstructured":"Abbas, A., Tirumala, K., Simig, D., Ganguli, S., and Morcos, A.S. (2023). SemDeDup: Data-efficient learning at web-scale through semantic deduplication. arXiv."},{"key":"ref_26","unstructured":"Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., and Antiga, L. (2019, January 8\u201314). PyTorch: An Imperative Style, High-Performance Deep Learning Library. Proceedings of the Advances in Neural Information Processing Systems 32, Vancouver, BC, Canada."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"357","DOI":"10.1038\/s41586-020-2649-2","article-title":"Array programming with NumPy","volume":"585","author":"Harris","year":"2020","journal-title":"Nature"},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"261","DOI":"10.1038\/s41592-019-0686-2","article-title":"SciPy 1.0: Fundamental algorithms for scientific computing in Python","volume":"17","author":"Virtanen","year":"2020","journal-title":"Nat. Methods"},{"key":"ref_29","doi-asserted-by":"crossref","first-page":"e453","DOI":"10.7717\/peerj.453","article-title":"scikit-image: Image Processing in Python","volume":"2","author":"Boulogne","year":"2014","journal-title":"PeerJ"},{"key":"ref_30","doi-asserted-by":"crossref","unstructured":"Hagberg, A.A., Schult, D.A., and Swart, P.J. (2008, January 19\u201324). Exploring Network Structure, Dynamics, and Function Using NetworkX. Proceedings of the 7th Python in Science Conference (SciPy2008), Pasadena, CA, USA.","DOI":"10.25080\/TCWV9851"}],"container-title":["Bioengineering"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2306-5354\/13\/7\/759\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,30]],"date-time":"2026-06-30T07:34:08Z","timestamp":1782804848000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2306-5354\/13\/7\/759"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,6,29]]},"references-count":30,"journal-issue":{"issue":"7","published-online":{"date-parts":[[2026,7]]}},"alternative-id":["bioengineering13070759"],"URL":"https:\/\/doi.org\/10.3390\/bioengineering13070759","relation":{},"ISSN":["2306-5354"],"issn-type":[{"value":"2306-5354","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,6,29]]}}}