{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,3,3]],"date-time":"2026-03-03T09:14:15Z","timestamp":1772529255336,"version":"3.50.1"},"reference-count":32,"publisher":"PeerJ","license":[{"start":{"date-parts":[[2026,3,3]],"date-time":"2026-03-03T00:00:00Z","timestamp":1772496000000},"content-version":"unspecified","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":[],"abstract":"<jats:p>Accurate medical image segmentation plays a crucial role in diagnosing diseases and planning effective treatments, particularly in detecting brain tumors using Magnetic Resonance Imaging (MRI) scans. These segmentations provide detailed insights required to identify problems and deliver timely and effective care. The contemporary models, like nnU-Net, provide exceptional accuracy. However, their processing requirements make it challenging to reach real-time performance. This study proposes an efficient feature map pruning method that concurrently increases nnU-Net\u2019s segmentation accuracy and reduces inference time. Several pruning criteria are explored to enhance the model\u2019s efficiency, such as the variance, mean activation, and a novel hybrid approach that combines activation statistics with weight norms. Through the assessment of these various approaches, we determine the best strategy for reducing unnecessary complexity without compromising model functionality. This process is further refined by introducing a voting mechanism that acts like a panel of experts, selecting the best pruning decisions based on multiple criteria. This collaborative approach ensures more reliable and consistent outcomes, optimizing the model\u2019s accuracy and latency while maintaining its streamlined and efficient nature. On the BraTS 2023 dataset, the pruned nnU-Net models achieve inference times of 0.0406 s, compared to 2.9258\u00a0s for the original model, which is a considerable improvement. Our pruned models maintain competitive Dice scores of up to 0.7585. These results demonstrate the potential of our approach for time-sensitive medical image processing.<\/jats:p>","DOI":"10.7717\/peerj-cs.3567","type":"journal-article","created":{"date-parts":[[2026,3,3]],"date-time":"2026-03-03T08:08:23Z","timestamp":1772525303000},"page":"e3567","source":"Crossref","is-referenced-by-count":0,"title":["Efficient brain tumor segmentation using Hybrid-Pruned nnU-Net"],"prefix":"10.7717","volume":"12","author":[{"given":"Muhammad","family":"Jameel","sequence":"first","affiliation":[{"name":"FAST School of Computing, National University of Computer and Emerging Sciences, Islamabad, Islamabad, Pakistan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Usman","family":"Haider","sequence":"additional","affiliation":[{"name":"FAST School of Computing, National University of Computer and Emerging Sciences, Islamabad, Islamabad, Pakistan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4793-6239","authenticated-orcid":true,"given":"Usman","family":"Habib","sequence":"additional","affiliation":[{"name":"FAST School of Computing, National University of Computer and Emerging Sciences, Islamabad, Islamabad, Pakistan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Fatima","family":"Khalid","sequence":"additional","affiliation":[{"name":"Faculty of Computer Science and Engineering, Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Topi, KPK, Pakistan"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Muhammad","family":"Nadeem","sequence":"additional","affiliation":[{"name":"College of Engineering and Technology, American University of the Middle East, Egaila, Kuwait"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Saeed Mian","family":"Qaisar","sequence":"additional","affiliation":[{"name":"College of Engineering and Technology, American University of the Middle East, Egaila, Kuwait"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"4443","published-online":{"date-parts":[[2026,3,3]]},"reference":[{"key":"10.7717\/peerj-cs.3567\/ref-1","first-page":"140","article-title":"Structured model pruning for efficient inference in computational pathology","author":"Adnan","year":"2024"},{"key":"10.7717\/peerj-cs.3567\/ref-2","first-page":"221","article-title":"Model ensemble for brain tumor segmentation in magnetic resonance imaging","author":"Capell\u00e1n-Mart\u00edn","year":"2023"},{"key":"10.7717\/peerj-cs.3567\/ref-3","article-title":"Feature map similarity-based neural network pruning","author":"Chen","year":"2023"},{"issue":"12","key":"10.7717\/peerj-cs.3567\/ref-4","doi-asserted-by":"publisher","first-page":"10558","DOI":"10.1109\/tpami.2024.3447085","article-title":"A survey on deep neural network pruning: taxonomy, comparison, analysis, and recommendations","volume":"46","author":"Cheng","year":"2023","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"issue":"4","key":"10.7717\/peerj-cs.3567\/ref-5","doi-asserted-by":"publisher","first-page":"102583","DOI":"10.1101\/2021.11.26.470124","article-title":"STAMP: simultaneous training and model pruning for low data regimes in medical image segmentation","volume":"81","author":"Dinsdale","year":"2022","journal-title":"Medical Image Analysis"},{"key":"10.7717\/peerj-cs.3567\/ref-6","article-title":"Deep compression: compressing deep neural networks with pruning, trained quantization and Huffman coding","author":"Han","year":"2016"},{"key":"10.7717\/peerj-cs.3567\/ref-7","article-title":"Learning both weights and connections for efficient neural network","author":"Han","year":"2015"},{"key":"10.7717\/peerj-cs.3567\/ref-8","doi-asserted-by":"publisher","first-page":"4340","DOI":"10.1109\/cvpr.2019.00447","article-title":"Filter pruning via geometric median for deep convolutional neural networks acceleration","author":"He","year":"2020","journal-title":"IEEE Transactions on Neural Networks and Learning Systems"},{"key":"10.7717\/peerj-cs.3567\/ref-9","article-title":"Channel pruning for accelerating very deep neural networks","author":"He","year":"2019"},{"key":"10.7717\/peerj-cs.3567\/ref-10","article-title":"Adversarial feature map pruning for backdoor defense in deep neural networks","author":"Huang","year":"2024"},{"key":"10.7717\/peerj-cs.3567\/ref-11","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2404.15009","article-title":"The brain tumor segmentation in pediatrics (BraTS-PEDs) challenge: focus on pediatrics (CBTN-CONNECT-DIPGR-ASNR-MICCAI BraTS-PEDs)","author":"Kazerooni","year":"2024a"},{"key":"10.7717\/peerj-cs.3567\/ref-12","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2407.08855","article-title":"BraTS-PEDs: results of the multi-consortium international pediatric brain tumor segmentation challenge 2023","author":"Kazerooni","year":"2024b"},{"issue":"2","key":"10.7717\/peerj-cs.3567\/ref-13","doi-asserted-by":"publisher","first-page":"100156","DOI":"10.1016\/j.neuri.2024.100156","article-title":"Brain tumor segmentation with advanced nnU-Net: pediatrics and adults tumors","volume":"4","author":"Kharaji","year":"2024","journal-title":"Neuroscience Informatics"},{"key":"10.7717\/peerj-cs.3567\/ref-14","first-page":"2554","article-title":"Fast ConvNets using group-wise brain damage","author":"Lebedev","year":"2016"},{"key":"10.7717\/peerj-cs.3567\/ref-15","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2305.09011","article-title":"The brain tumor segmentation (BraTS) challenge 2023: brain MR image synthesis for tumor segmentation (BraSyn)","author":"Li","year":"2024a"},{"key":"10.7717\/peerj-cs.3567\/ref-16","article-title":"Pruning filters for efficient ConvNets","author":"Li","year":"2016"},{"key":"10.7717\/peerj-cs.3567\/ref-17","article-title":"Model pruning based on quantified similarity of neurons","author":"Li","year":"2024b"},{"key":"10.7717\/peerj-cs.3567\/ref-18","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2206.14413","article-title":"The lighter the better: rethinking transformers in medical image segmentation through adaptive pruning","author":"Lin","year":"2022"},{"key":"10.7717\/peerj-cs.3567\/ref-19","doi-asserted-by":"crossref","DOI":"10.1109\/ICCV.2017.298","article-title":"Learning efficient convolutional networks through network slimming","author":"Liu","year":"2017"},{"key":"10.7717\/peerj-cs.3567\/ref-20","doi-asserted-by":"publisher","first-page":"16","DOI":"10.1109\/tim.2024.3413130","article-title":"TransSea: hybrid CNN\u2013transformer with semantic awareness for 3-D brain tumor segmentation","volume":"73","author":"Liu","year":"2024b","journal-title":"IEEE Transactions on Instrumentation and Measurement"},{"key":"10.7717\/peerj-cs.3567\/ref-21","doi-asserted-by":"publisher","first-page":"1199","DOI":"10.1109\/TIP.2024.3359815","article-title":"Multimodal brain tumor segmentation boosted by monomodal normal brain images","volume":"33","author":"Liu","year":"2024a","journal-title":"IEEE Transactions on Image Processing"},{"key":"10.7717\/peerj-cs.3567\/ref-22","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1906.10771","article-title":"Importance estimation for neural network pruning","author":"Molchanov","year":"2019"},{"key":"10.7717\/peerj-cs.3567\/ref-23","article-title":"Pruning convolutional neural networks for resource efficient transfer learning","author":"Molchanov","year":"2021"},{"key":"10.7717\/peerj-cs.3567\/ref-24","doi-asserted-by":"publisher","first-page":"2204","DOI":"10.1109\/WACV57701.2024.00220","article-title":"Towards better structured pruning saliency by reorganizing convolution","author":"Sun","year":"2024","journal-title":"IEEE Transactions on Pattern Analysis and Machine Intelligence"},{"key":"10.7717\/peerj-cs.3567\/ref-25","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2209.13590","article-title":"Sauron U-Net: simple automated redundancy elimination in medical image segmentation via filter pruning","author":"Valverde","year":"2022"},{"issue":"1","key":"10.7717\/peerj-cs.3567\/ref-26","doi-asserted-by":"publisher","first-page":"103654","DOI":"10.1016\/j.media.2025.103654","article-title":"Diff-UNet: a diffusion embedded network for robust 3D medical image segmentation","volume":"105","author":"Xing","year":"2025a","journal-title":"Medical Image Analysis"},{"issue":"1","key":"10.7717\/peerj-cs.3567\/ref-27","doi-asserted-by":"publisher","first-page":"4","DOI":"10.1109\/tmi.2025.3589797","article-title":"SegMamba-V2: long-range sequential modeling mamba for general 3D medical image segmentation","volume":"45","author":"Xing","year":"2025b","journal-title":"IEEE Transactions on Medical Imaging"},{"key":"10.7717\/peerj-cs.3567\/ref-28","first-page":"578","article-title":"SegMamba: long-range sequential modeling mamba for 3D medical image segmentation","author":"Xing","year":"2024"},{"key":"10.7717\/peerj-cs.3567\/ref-29","first-page":"123","article-title":"NestedFormer: nested modality-aware transformer for brain tumor segmentation","author":"Xing","year":"2022"},{"issue":"4","key":"10.7717\/peerj-cs.3567\/ref-30","doi-asserted-by":"publisher","first-page":"2115","DOI":"10.1109\/jbhi.2024.3360239","article-title":"Hybrid masked image modeling for 3D medical image segmentation","volume":"28","author":"Xing","year":"2023","journal-title":"IEEE Journal of Biomedical and Health Informatics"},{"key":"10.7717\/peerj-cs.3567\/ref-31","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.2403.15735","article-title":"TransUNet for BraTs 2023","author":"Yang","year":"2024"},{"issue":"1","key":"10.7717\/peerj-cs.3567\/ref-32","doi-asserted-by":"publisher","first-page":"e855","DOI":"10.1186\/s12880-022-00931-1","article-title":"Brain tumour segmentation based on an improved U-Net","volume":"22","author":"Zheng","year":"2022","journal-title":"BMC Medical Imaging"}],"container-title":["PeerJ Computer Science"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/peerj.com\/articles\/cs-3567.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/peerj.com\/articles\/cs-3567.xml","content-type":"application\/xml","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/peerj.com\/articles\/cs-3567.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/peerj.com\/articles\/cs-3567.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,3,3]],"date-time":"2026-03-03T08:08:29Z","timestamp":1772525309000},"score":1,"resource":{"primary":{"URL":"https:\/\/peerj.com\/articles\/cs-3567"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3,3]]},"references-count":32,"alternative-id":["10.7717\/peerj-cs.3567"],"URL":"https:\/\/doi.org\/10.7717\/peerj-cs.3567","archive":["CLOCKSS","LOCKSS","Portico"],"relation":{},"ISSN":["2376-5992"],"issn-type":[{"value":"2376-5992","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3,3]]},"article-number":"e3567"}}