{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,5,5]],"date-time":"2026-05-05T02:56:00Z","timestamp":1777949760557,"version":"3.51.4"},"reference-count":36,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T00:00:00Z","timestamp":1777852800000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["frontiersin.org"],"crossmark-restriction":true},"short-container-title":["Front. Med."],"abstract":"<jats:sec>\n                    <jats:title>Introduction<\/jats:title>\n                    <jats:p>Brain tumor lesions in MRI images are characterized by low contrast, indistinct boundaries, and irregular morphology, posing substantial challenges for automated detection. RT-DETR provides an end-to-end, NMS-free detection paradigm, but its original design is optimized for natural images and exhibits three domain-specific limitations in brain tumor MRI detection: insufficient local texture perception of low-contrast lesion boundaries, neglect of feature magnitude differences in linear attention, and weak modeling of spatial continuity for morphologically irregular tumors. This study proposes CMRA-DETR (CSP-MambaOut with Retention and Magnitude-Aware Attention for Real-Time Detection Transformer), a lightweight yet high-accuracy detection framework built upon RT-DETR-R18, to address these limitations through targeted architectural adaptations.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>CMRA-DETR introduces three improvements: (1) a CSP-MambaOut backbone that enhances local texture perception of low-contrast lesion boundaries via gated feature selection; (2) an AIFI-MALA module that introduces magnitude-aware linear attention to correct the distributional smoothing deficiency of standard linear attention; and (3) a RetBlockC3 module that incorporates a Manhattan distance decay-based spatial retention mechanism to improve modeling of spatial continuity for morphologically irregular tumors. The model was trained and evaluated on an internal independent test set of 5,731 MRI images covering four categories (no tumor, meningioma, glioma, and pituitary adenoma), and further assessed on an external independent test set from the BRISC dataset without fine-tuning. Its performance was benchmarked against Faster R-CNN, YOLOv5n, YOLOv8n, YOLO11n, YOLO12n, and the RT-DETR-R18 baseline.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results and discussion<\/jats:title>\n                    <jats:p>\n                      On the internal independent test set, CMRA-DETR achieved achieves\n                      <jats:italic>p<\/jats:italic>\n                      \u202f=\u202f95.5%,\n                      <jats:italic>R<\/jats:italic>\n                      \u202f=\u202f95.7%, mAP@50 = 97.9%, and mAP@50-95 = 82.6%, while reducing parameters and GFLOPs by 37.7% and 30.9%, respectively, relative to the baseline, achieving competitive or superior performance compared to the evaluated baseline models under the specified experimental conditions. On the external independent test set from the BRISC dataset, the model achieved mAP@50 = 96.6% and mAP@50-95 = 79.8% without fine-tuning, demonstrating robust cross-dataset generalization. These findings indicate that CMRA-DETR achieves a favorable balance among detection accuracy, model lightweightness, and inference efficiency, demonstrating practical potential for AI-assisted brain tumor detection on resource-constrained clinical devices.\n                    <\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fmed.2026.1839485","type":"journal-article","created":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T04:25:33Z","timestamp":1777868733000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["CMRA-DETR: a lightweight and high-accuracy detection framework for MRI-based brain tumor identification"],"prefix":"10.3389","volume":"13","author":[{"given":"Cai","family":"Weng","sequence":"first","affiliation":[{"name":"The Second Affiliated Hospital of Fujian University of Traditional Chinese Medicine","place":["Fuzhou, Fujian, China"]},{"name":"Fujian University of Traditional Chinese Medicine","place":["Fuzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Bowei","family":"Huang","sequence":"additional","affiliation":[{"name":"Fujian University of Traditional Chinese Medicine","place":["Fuzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Jinghui","family":"Chen","sequence":"additional","affiliation":[{"name":"The First Clinical Medical College, The Affiliated People\u2019s Hospital of Fujian University of Traditional Chinese Medicine","place":["Fuzhou, Fujian, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wei","family":"Hu","sequence":"additional","affiliation":[{"name":"The Third Affiliated People\u2019s Hospital of Fujian University of Traditional Chinese Medicine","place":["Fuzhou, Fujian, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Zhiqing","family":"Huang","sequence":"additional","affiliation":[{"name":"The Second Affiliated Hospital of Fujian University of Traditional Chinese Medicine","place":["Fuzhou, Fujian, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Punan","family":"Weng","sequence":"additional","affiliation":[{"name":"Fujian University of Traditional Chinese Medicine","place":["Fuzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Hongjia","family":"Zhao","sequence":"additional","affiliation":[{"name":"Fujian University of Traditional Chinese Medicine","place":["Fuzhou, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Minqin","family":"Zheng","sequence":"additional","affiliation":[{"name":"Fujian University of Traditional Chinese Medicine","place":["Fuzhou, China"]},{"name":"Fuzhou Hospital of Traditional Chinese Medicine","place":["Fuzhou, Fujian, China"]}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1965","published-online":{"date-parts":[[2026,5,4]]},"reference":[{"key":"ref1","first-page":"143","volume-title":"Epidemiology and Outcome of Glioblastoma","author":"Tamimi","year":"2017"},{"key":"ref2","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1186\/s40001-023-01011-y","article-title":"The global prevalence of primary central nervous system tumors: a systematic review and meta-analysis","volume":"28","author":"Salari","year":"2023","journal-title":"Eur J Med Res"},{"key":"ref3","first-page":"211","article-title":"Primary brain tumors in adults: diagnosis and treatment","volume":"93","author":"Perkins","year":"2016","journal-title":"Am Fam Physician"},{"key":"ref4","volume-title":"Holland-Frei Cancer Medicine 8","author":"Holland","year":"2010"},{"key":"ref5","doi-asserted-by":"publisher","first-page":"284","DOI":"10.1186\/s12951-022-01479-6","article-title":"Recent development of contrast agents for magnetic resonance and multimodal imaging of glioblastoma","volume":"20","author":"Zhuang","year":"2022","journal-title":"J Nanobiotechnol"},{"key":"ref6","doi-asserted-by":"publisher","first-page":"1188","DOI":"10.1093\/neuonc\/nov095","article-title":"Consensus recommendations for a standardized brain tumor imaging protocol in clinical trials","volume":"17","author":"Ellingson","year":"2015","journal-title":"Neurooncology"},{"key":"ref7","doi-asserted-by":"publisher","first-page":"e305","DOI":"10.1016\/S1470-2045(20)30064-4","article-title":"Response assessment in paediatric low-grade glioma: recommendations from the response assessment in Pediatric neuro-oncology (Rapno) working group","volume":"21","author":"Fangusaro","year":"2020","journal-title":"Lancet Oncol"},{"key":"ref8","doi-asserted-by":"publisher","first-page":"2650","DOI":"10.3390\/diagnostics13162650","article-title":"Brain tumor segmentation from MRI images using handcrafted convolutional neural network","volume":"13","author":"Ullah","year":"2023","journal-title":"Diagnostics"},{"key":"ref9","doi-asserted-by":"publisher","first-page":"181","DOI":"10.1007\/s12553-023-00737-3","article-title":"A survey of deep learning for MRI brain tumor segmentation methods: trends, challenges, and future directions","volume":"13","author":"Krishnapriya","year":"2023","journal-title":"Health Technol"},{"key":"ref10","doi-asserted-by":"publisher","first-page":"34","DOI":"10.1016\/j.compmedimag.2019.05.001","article-title":"Brain tumor classification for Mr images using transfer learning and fine-tuning","volume":"75","author":"Swati","year":"2019","journal-title":"Comput Med Imaging Graph"},{"key":"ref11","doi-asserted-by":"publisher","first-page":"103345","DOI":"10.1016\/j.compbiomed.2019.103345","article-title":"Brain tumor classification using deep CNN features via transfer learning","volume":"111","author":"Deepak","year":"2019","journal-title":"Comput Biol Med"},{"key":"ref12","doi-asserted-by":"publisher","first-page":"133","DOI":"10.1002\/jemt.23597","article-title":"Microscopic brain tumor detection and classification using 3d Cnn and feature selection architecture","volume":"84","author":"Rehman","year":"2021","journal-title":"Microsc Res Tech"},{"key":"ref13","doi-asserted-by":"publisher","first-page":"174","DOI":"10.1016\/j.jocs.2018.12.003","article-title":"Multi-grade brain tumor classification using deep CNN with extensive data augmentation","volume":"30","author":"Sajjad","year":"2019","journal-title":"J Comput Sci"},{"key":"ref14","volume-title":"World Congress on Medical Physics and Biomedical Engineering 2018","author":"Abiwinanda","year":"2018"},{"key":"ref15","doi-asserted-by":"publisher","first-page":"2664","DOI":"10.1038\/s41598-024-52823-9","article-title":"Brain tumor detection from images and comparison with transfer learning methods and 3-layer Cnn","volume":"14","author":"Khaliki","year":"2024","journal-title":"Sci Rep"},{"key":"ref16","doi-asserted-by":"publisher","first-page":"2976","DOI":"10.3390\/electronics14152976","article-title":"Evaluation of vision transformers for multi-organ tumor classification using MRI and CT imaging","volume":"14","author":"Mart\u00edn","year":"2025","journal-title":"Electronics"},{"key":"ref17","doi-asserted-by":"publisher","first-page":"7913","DOI":"10.3390\/s23187913","article-title":"Advancing brain tumor classification through fine-tuned vision transformers: a comparative study of pre-trained models","volume":"23","author":"Asiri","year":"2023","journal-title":"Sensors"},{"key":"ref18","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Zhao","year":"2024"},{"key":"ref19","doi-asserted-by":"publisher","first-page":"364","DOI":"10.3390\/biomedicines11020364","article-title":"Magnetic resonance imaging of primary adult brain Tumors: state of the art and future perspectives","volume":"11","author":"Martucci","year":"2023","journal-title":"Biomedicine"},{"key":"ref20","doi-asserted-by":"publisher","first-page":"S107","DOI":"10.1259\/bjr\/65711810","article-title":"Conventional MRI evaluation of gliomas","volume":"84","author":"Upadhyay","year":"2011","journal-title":"Br J Radiol"},{"key":"ref21","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition Workshops","author":"Wang","year":"2020"},{"key":"ref22","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer Vision and Pattern Recognition","author":"Yu","year":"2025"},{"key":"ref23","volume-title":"Proceedings of the IEEE\/CVF Conference on Computer vision and Pattern Recognition","author":"Fan","year":"2024"},{"key":"ref24","doi-asserted-by":"publisher","DOI":"10.34740\/KAGGLE\/DSV\/14832123","article-title":"Brain tumor MRI dataset","author":"Nickparvar","year":"2026","journal-title":"Kaggle"},{"key":"ref25","doi-asserted-by":"publisher","DOI":"10.34740\/KAGGLE\/DSV\/12745533","article-title":"Brain tumor classification (MRI)","author":"Bhuvaji","year":"2025","journal-title":"Kaggle"},{"key":"ref26","doi-asserted-by":"publisher","DOI":"10.1038\/s41597-026-06753-y","article-title":"Brisc: annotated dataset for brain tumor segmentation and classification","volume":"13","author":"Fateh","year":"2026","journal-title":"Sci Data"},{"key":"ref27","article-title":"Transformers are RNNS: fast autoregressive transformers with linear attention","volume-title":"International Conference on Machine Learning","author":"Katharopoulos","year":"2020"},{"key":"ref28","doi-asserted-by":"publisher","first-page":"60","DOI":"10.1016\/j.media.2017.07.005","article-title":"A survey on deep learning in medical image analysis","volume":"42","author":"Litjens","year":"2017","journal-title":"Med Image Anal"},{"key":"ref29","doi-asserted-by":"publisher","first-page":"70","DOI":"10.1007\/s12530-025-09699-5","article-title":"Comparing the efficiency of Yolo-M for face recognition in images and videos degraded by compression artifacts","volume":"16","author":"Ferreira","year":"2025","journal-title":"Evol Syst"},{"key":"ref30","doi-asserted-by":"publisher","first-page":"22037","DOI":"10.1007\/s00521-024-10351-7","article-title":"Exploring the potential of Yolov8 in hybrid models for facial mask identification in diverse environments","volume":"36","author":"Ferreira","year":"2024","journal-title":"Neural Comput & Applic"},{"key":"ref31","doi-asserted-by":"publisher","first-page":"167","DOI":"10.1007\/s13721-025-00669-0","article-title":"Explainable machine learning framework integrating Pppm for automated cytological triage and personalized breast cancer risk stratification","volume":"14","author":"Ferreira","year":"2025","journal-title":"Netw Model Anal Health Inform Bioinform"},{"key":"ref32","volume-title":"European Conference on Computer Vision","author":"Carion","year":"2020"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1910.04597","article-title":"Machine learning with multi-site imaging data: an empirical study on the impact of scanner effects","author":"Glocker","year":"2019","journal-title":"arXiv preprint arXiv:191004597"},{"key":"ref34","doi-asserted-by":"publisher","first-page":"1231","DOI":"10.1093\/neuonc\/noab106","article-title":"The 2021 who classification of Tumors of the central nervous system: a summary","volume":"23","author":"Louis","year":"2021","journal-title":"Neuro-Oncology"},{"key":"ref35","doi-asserted-by":"publisher","first-page":"101796","DOI":"10.1016\/j.media.2020.101796","article-title":"Bias: transparent reporting of biomedical image analysis challenges","volume":"66","author":"Maier-Hein","year":"2020","journal-title":"Med Image Anal"},{"key":"ref36","doi-asserted-by":"publisher","first-page":"170117","DOI":"10.1038\/sdata.2017.117","article-title":"Advancing the Cancer genome atlas glioma Mri collections with expert segmentation labels and radiomic features","volume":"4","author":"Bakas","year":"2017","journal-title":"Sci Data"}],"container-title":["Frontiers in Medicine"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2026.1839485\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,5,4]],"date-time":"2026-05-04T04:25:36Z","timestamp":1777868736000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fmed.2026.1839485\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,5,4]]},"references-count":36,"alternative-id":["10.3389\/fmed.2026.1839485"],"URL":"https:\/\/doi.org\/10.3389\/fmed.2026.1839485","relation":{},"ISSN":["2296-858X"],"issn-type":[{"value":"2296-858X","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,5,4]]},"article-number":"1839485"}}