{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T15:02:31Z","timestamp":1789743751790,"version":"4.0.1"},"reference-count":41,"publisher":"Frontiers Media SA","license":[{"start":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T00:00:00Z","timestamp":1789689600000},"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. Radiol."],"abstract":"<jats:sec>\n                    <jats:title>Background<\/jats:title>\n                    <jats:p>The increasing demand for lumbar spine MRI exams, combined with radiologist shortages, has underscored the need for automated tools to assist in diagnosing degenerative spinal conditions. Despite the proliferation of AI tools in radiology, no CE or FDA-approved applications currently focus on pathologies in the lumbar spine. This study aimed to develop a pipeline for automatic identification of intervertebral disc herniation, contributing to more efficient clinical decision-making.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Methods<\/jats:title>\n                    <jats:p>A dataset of 484 lumbar spine MRI examinations acquired from four manufacturers was used to develop a multi-stage deep learning pipeline. An overview of the proposed automated lumbar spine MRI analysis framework is presented, illustrating the main stages of the pipeline including dataset preparation, disc segmentation, feature extraction, and classification. A subset of 165 studies (5,200 sagittal slices) was manually annotated by radiologists and used for training and validation of vertebrae and intervertebral disc segmentation models, including U-Net, DeepLabV3+, and MA-Net with EfficientNet-B2 backbones. Anatomically informed post-processing was applied to assign disc-level labels and extract disc-centered volumes. Disc-level binary classification (herniated vs. non-herniated) was then performed using 3D patch-based models, including ResNet, SENet, EfficientNet-B2, and Vision Transformer (ViT) architectures, with all dataset splits performed at the patient level.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Results<\/jats:title>\n                    <jats:p>The best segmentation performance was achieved by MA-Net with EfficientNet-B2, yielding a Dice Score of 0.898 for three-class segmentation (background, vertebrae, discs). For classification, the evaluated models achieved F1-scores between 0.66 and 0.73. The best performance was obtained with a ResNet18-based model, indicating that once anatomical localization and disc extraction are addressed by the segmentation stage, lower-capacity architectures can be effective for binary classification. The simplified three-class segmentation improved model robustness, and post-processing enabled disc-level precision essential for downstream classification, which achieved consistent performance across models under the available data regime.<\/jats:p>\n                  <\/jats:sec>\n                  <jats:sec>\n                    <jats:title>Conclusion<\/jats:title>\n                    <jats:p>The proposed AI pipeline successfully automates the identification of lumbar disc herniation in MRI exams. This approach has the potential to enhance diagnostic efficiency, reduce radiologist workload, and improve reproducibility in identifying disc pathologies. While limited by sample size and class imbalance, the methodology lays a strong foundation for scalable annotation and integration of AI in routine spinal diagnostics.<\/jats:p>\n                  <\/jats:sec>","DOI":"10.3389\/fradi.2026.1722312","type":"journal-article","created":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T14:05:16Z","timestamp":1789740316000},"update-policy":"https:\/\/doi.org\/10.3389\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["The application of deep neural networks to the detection of lumbar hernias: results from the approaches based on convolutional networks and transformers"],"prefix":"10.3389","volume":"6","author":[{"given":"Ant\u00f3nio","family":"Fernandes","sequence":"first","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]},{"name":"ALGORITMI, LASI, Universidade do Minho","place":["Braga, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jo\u00e3o Pedro Mazuco","family":"Rodriguez","sequence":"additional","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Susana","family":"Moleirinho","sequence":"additional","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Irina","family":"Trofimenko","sequence":"additional","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ekaterina","family":"Guseva","sequence":"additional","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Alexander","family":"Martinovich","sequence":"additional","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Ilzane Maria","family":"de Oliveira Morais","sequence":"additional","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Louise","family":"Bisolo","sequence":"additional","affiliation":[{"name":"SliceD Group Research Center","place":["Lisbon, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Lu\u00eds","family":"Mendes Gomes","sequence":"additional","affiliation":[{"name":"ALGORITMI, LASI, Universidade do Minho","place":["Braga, Portugal"]},{"name":"Faculdade de Ci\u00eancias e Tecnologia, Universidade dos A\u00e7ores","place":["Ponta Delgada, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Jos\u00e9","family":"Machado","sequence":"additional","affiliation":[{"name":"ALGORITMI, LASI, Universidade do Minho","place":["Braga, Portugal"]}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"1965","published-online":{"date-parts":[[2026,9,18]]},"reference":[{"key":"B1","year":"2023"},{"key":"B2","doi-asserted-by":"publisher","DOI":"10.1186\/s13244-023-01541-3","article-title":"Developing, purchasing, implementing and monitoring AI tools in radiology: practical considerations. A multi-society statement from the ACR, CAR, ESR, RANZCR & RSNA","volume":"15","author":"Brady","year":"2024","journal-title":"Insights Imaging"},{"key":"B3","doi-asserted-by":"publisher","DOI":"10.1148\/rg.2017160130","article-title":"Machine learning for medical imaging","volume":"37","author":"Erickson","year":"2017","journal-title":"Radiographics"},{"key":"B4","year":"2024"},{"key":"B5","doi-asserted-by":"publisher","first-page":"2356","DOI":"10.1016\/S0140-6736(18)30480-X","article-title":"What low back pain is and why we need to pay attention","volume":"391","author":"Hartvigsen","year":"2018","journal-title":"Lancet"},{"key":"B6","doi-asserted-by":"crossref","first-page":"299","DOI":"10.21037\/atm.2020.02.175","article-title":"Global low back pain prevalence and years lived with disability from 1990 to 2017: estimates from the global burden of disease study 2017","volume":"8","author":"Wu","year":"2020","journal-title":"Ann Transl Med"},{"key":"B7","doi-asserted-by":"publisher","DOI":"10.7759\/cureus.47908","article-title":"Management of lumbar disc herniation: a systematic review","volume":"15","author":"Awadalla","year":"2023","journal-title":"Cureus"},{"key":"B8","doi-asserted-by":"publisher","first-page":"811","DOI":"10.3174\/ajnr.A4173","article-title":"Systematic literature review of imaging features of spinal degeneration in asymptomatic populations","volume":"36","author":"Brinjikji","year":"2015","journal-title":"Am J Neuroradiol"},{"key":"B9","doi-asserted-by":"publisher","first-page":"11","DOI":"10.1186\/s41747-023-00408-y","article-title":"Image annotation and curation in radiology: an overview for machine learning practitioners","volume":"8","author":"Galbusera","year":"2024","journal-title":"Eur Radiol Exp"},{"key":"B10","doi-asserted-by":"publisher","first-page":"2368","DOI":"10.1016\/S0140-6736(18)30489-6","article-title":"Prevention and treatment of low back pain: evidence, challenges, and promising directions","volume":"391","author":"Foster","year":"2018","journal-title":"Lancet"},{"key":"B11","doi-asserted-by":"publisher","first-page":"11708","DOI":"10.3390\/ijerph191811708","article-title":"Artificial intelligence in spinal imaging: current Status and future directions","volume":"19","author":"Cui","year":"2022","journal-title":"Int J Environ Res Public Health"},{"key":"B12","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":"B13","doi-asserted-by":"publisher","first-page":"102444","DOI":"10.1016\/j.media.2022.102444","article-title":"Recent advances and clinical applications of deep learning in medical image analysis","volume":"79","author":"Chen","year":"2022","journal-title":"Med Image Anal"},{"key":"B14","doi-asserted-by":"publisher","first-page":"954","DOI":"10.1038\/s41591-019-0447-x","article-title":"End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography","volume":"25","author":"Ardila","year":"2019","journal-title":"Nat Med"},{"key":"B15","doi-asserted-by":"publisher","first-page":"102","DOI":"10.1016\/j.zemedi.2018.11.002","article-title":"An overview of deep learning in medical imaging focusing on MRI","volume":"29","author":"Lundervold","year":"2019","journal-title":"Z Med Phys"},{"key":"B16","doi-asserted-by":"publisher","first-page":"102166","DOI":"10.1016\/j.media.2021.102166","article-title":"VERSE: a vertebrae labelling and segmentation benchmark for multi-detector CT images","volume":"73","author":"Sekuboyina","year":"2021","journal-title":"Med Image Anal"},{"key":"B17","doi-asserted-by":"publisher","first-page":"119","DOI":"10.1038\/s41746-020-00323-1","article-title":"The future of digital health with federated learning","volume":"3","author":"Rieke","year":"2020","journal-title":"NPJ Digit Med"},{"key":"B18","doi-asserted-by":"publisher","first-page":"2","DOI":"10.1038\/s41592-023-02151-z","article-title":"Metrics reloaded: pitfalls and recommendations for image analysis validation","volume":"21","author":"Maier-Hein","year":"2024","journal-title":"Nat Methods"},{"key":"B19","doi-asserted-by":"publisher","first-page":"101693","DOI":"10.1016\/j.media.2020.101693","article-title":"Embracing imperfect datasets: a review of deep learning solutions for medical image segmentation","volume":"63","author":"Tajbakhsh","year":"2020","journal-title":"Med Image Anal"},{"key":"B20","doi-asserted-by":"publisher","first-page":"E14","DOI":"10.1038\/s41586-020-2766-y","article-title":"Transparency and reproducibility in artificial intelligence","volume":"586","author":"Haibe-Kains","year":"2020","journal-title":"Nature"},{"key":"B21","doi-asserted-by":"publisher","first-page":"100279","DOI":"10.1016\/j.wnsx.2024.100279","article-title":"Lumbar disc herniation: epidemiology, clinical and radiologic diagnosis WFNS spine committee recommendations","volume":"22","author":"Pojskic","year":"2024","journal-title":"World Neurosurg X"},{"key":"B22","doi-asserted-by":"publisher","first-page":"1267","DOI":"10.1016\/S0033-8389(08)70006-X","article-title":"Classification, diagnostic imaging, and imaging characterization of a lumbar herniated disk","volume":"38","author":"Milette","year":"2000","journal-title":"Radiol Clin N Am"},{"key":"B23","volume-title":"Lumbar Disk Herniation","author":"Al Qaraghli","year":"2023"},{"key":"B24","author":"Tan","year":"2019"},{"key":"B25","article-title":"Very deep convolutional networks for large-scale image recognition","author":"Simonyan","year":"2015"},{"key":"B26","doi-asserted-by":"crossref","DOI":"10.1109\/CVPR.2016.90","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"B27","doi-asserted-by":"publisher","first-page":"109131","DOI":"10.1016\/j.patcog.2022.109131","article-title":"Mix-ViT: mixing attentive vision transformer for ultra-fine-grained visual categorization","volume":"135","author":"Yu","year":"2023","journal-title":"Pattern Recognit"},{"key":"B28","author":"Ronneberger","year":"2015"},{"key":"B29","author":"Zhou","year":"2018"},{"key":"B30","first-page":"833","article-title":"Encoder-Decoder with atrous separable convolution for semantic image segmentation BT - computer vision \u2013 ECCV 2018","author":"Chen","year":"2018"},{"key":"B31","first-page":"12077","article-title":"Segformer: simple and efficient design for semantic segmentation with transformers","volume":"15","author":"Xie","year":"2021","journal-title":"Adv Neural Inf Process Syst"},{"key":"B32","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1109\/TGRS.2021.3093977","article-title":"Multiattention network for semantic segmentation of fine-resolution remote sensing images","volume":"60","author":"Li","year":"2022","journal-title":"IEEE Trans Geosci Remote Sens"},{"key":"B33","article-title":"Adam: a method for stochastic optimization","author":"Kingma","year":"2015"},{"key":"B34","article-title":"Exploration of vision transformer models in medical images synthesis","author":"Chen","year":"2022"},{"key":"B35","author":"Karimi","year":"2021"},{"key":"B36","author":"Valanarasu","year":"2022"},{"key":"B37","doi-asserted-by":"publisher","first-page":"4","DOI":"10.1148\/radiol.2020192224","article-title":"Preparing medical imaging data for machine learning","volume":"295","author":"Willemink","year":"2020","journal-title":"Radiology"},{"key":"B38","first-page":"0","article-title":"Deep medicine - how artificial intelligence can make healthcare human again","volume":"53","author":"Topol","year":"2019","journal-title":"J Chem Inf Model"},{"key":"B39","doi-asserted-by":"publisher","first-page":"4793","DOI":"10.1109\/TNNLS.2020.3027314","article-title":"A survey on explainable artificial intelligence (XAI): toward medical XAI","volume":"32","author":"Tjoa","year":"2021","journal-title":"IEEE Trans Neural Netw Learn Syst"},{"key":"B40","doi-asserted-by":"publisher","first-page":"63","DOI":"10.1016\/j.media.2017.07.002","article-title":"Spinenet: automated classification and evidence visualization in spinal MRIs","volume":"41","author":"Jamaludin","year":"2017","journal-title":"Med Image Anal"},{"key":"B41","doi-asserted-by":"publisher","first-page":"495","DOI":"10.1007\/s00330-019-06313-4","article-title":"Automatic detection of lumbar disc herniation from MRI using deep neural networks","volume":"30","author":"Chen","year":"2020","journal-title":"Eur Radiol"}],"container-title":["Frontiers in Radiology"],"original-title":[],"link":[{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2026.1722312\/full","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,18]],"date-time":"2026-09-18T14:05:17Z","timestamp":1789740317000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.frontiersin.org\/articles\/10.3389\/fradi.2026.1722312\/full"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,18]]},"references-count":41,"alternative-id":["10.3389\/fradi.2026.1722312"],"URL":"https:\/\/doi.org\/10.3389\/fradi.2026.1722312","relation":{},"ISSN":["2673-8740"],"issn-type":[{"value":"2673-8740","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,18]]},"article-number":"1722312"}}