{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T02:57:09Z","timestamp":1790650629501,"version":"4.1.0"},"reference-count":25,"publisher":"Springer Science and Business Media LLC","license":[{"start":{"date-parts":[[2026,9,28]],"date-time":"2026-09-28T00:00:00Z","timestamp":1790553600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,9,28]],"date-time":"2026-09-28T00:00:00Z","timestamp":1790553600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Phys Eng Sci Med"],"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    Lumbar spine degeneration (LSD) is a major contributor to lower back pain (LBP) worldwide, typically diagnosed using magnetic resonance imaging (MRI). Manual interpretation of MRI scans is time-consuming and prone to variability, creating an urgent need for automated solutions. This study introduces a deep learning (DL) framework for classifying three key degenerative conditions, neural foraminal narrowing, subarticular stenosis, and spinal canal stenosis, across five lumbar disc levels. We evaluated multiple architectures, including VGG-16, EfficientNetV2-S, and YOLOv5, and enhanced feature extraction using the Convolutional Block Attention Module (CBAM) while improving classification through a Multilayer Perceptron (MLP). The dataset comprises 48,657 multi-centre MRI images and was partitioned using a stratified 80:20 training\u2013testing split. Class imbalance was addressed using Random Oversampling (ROS), Synthetic Minority Over-sampling Technique (SMOTE), and Generative Adversarial Network (GAN)-based augmentation. Model development incorporated validation monitoring and early stopping during training, while final performance was assessed on an independent testing set using accuracy, precision, recall, F1-score, confusion matrices, and receiver operating characteristic (ROC) analysis. The proposed hybrid model, VGG\u201116+CBAM+MLP, achieved 94% accuracy, with precision, recall, and F1\u2011score exceeding 91%. Paired statistical tests confirmed these improvements were significant (\n                    <jats:italic>p<\/jats:italic>\n                    \u202f&lt;\u202f0.05), reinforcing the robustness of the approach. These results demonstrate the potential of Artificial Intelligence (AI)-driven tools to support lumbar spine MRI assessment. Further external validation is required before clinical deployment.\n                  <\/jats:p>","DOI":"10.1007\/s13246-026-01808-1","type":"journal-article","created":{"date-parts":[[2026,9,28]],"date-time":"2026-09-28T13:55:14Z","timestamp":1790603714000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Hybrid attention and MLP-based deep learning for MRI classification of lumbar spine degeneration"],"prefix":"10.1007","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7065-9914","authenticated-orcid":false,"given":"Bhuvendhraa","family":"Rudrusamy","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yin Shao","family":"Ng","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Heng Kar","family":"Lau","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,9,28]]},"reference":[{"issue":"4","key":"1808_CR1","doi-asserted-by":"publisher","first-page":"611","DOI":"10.1007\/s13244-018-0639-9","volume":"9","author":"R Yamashita","year":"2018","unstructured":"Yamashita R, Nishio M, Do RKG, Togashi K (2018) Convolutional neural networks: an overview and application in radiology. Insights Imaging 9(4):611\u2013629. https:\/\/doi.org\/10.1007\/s13244-018-0639-9","journal-title":"Insights Imaging"},{"key":"1808_CR2","doi-asserted-by":"publisher","DOI":"10.3389\/fbioe.2025.1526478","author":"A Wang","year":"2025","unstructured":"Wang A, Wang T, Liu X, Fan N, Yuan S, Du P et al (2025) Automated diagnosis and grading of lumbar intervertebral disc degeneration based on a modified YOLO framework. Front Bioeng Biotechnol. https:\/\/doi.org\/10.3389\/fbioe.2025.1526478","journal-title":"Front Bioeng Biotechnol"},{"key":"1808_CR3","doi-asserted-by":"publisher","DOI":"10.1016\/j.inat.2020.100837","volume":"22","author":"W Mbarki","year":"2020","unstructured":"Mbarki W, Bouchouicha M, Frizzi S, Tshibasu F, Farhat LB, Sayadi M (2020) Lumbar spine discs classification based on deep convolutional neural networks using axial view MRI. Interdiscip Neurosurg 22:100837. https:\/\/doi.org\/10.1016\/j.inat.2020.100837","journal-title":"Interdiscip Neurosurg"},{"issue":"10","key":"1808_CR4","doi-asserted-by":"publisher","DOI":"10.3390\/bioengineering11101021","volume":"11","author":"J Lin","year":"2024","unstructured":"Lin J, Zhang H, Shang H (2024) Convolutional neural network incorporating multiple attention mechanisms for MRI classification of lumbar spinal stenosis. Bioengineering (Basel) 11(10):1021. https:\/\/doi.org\/10.3390\/bioengineering11101021","journal-title":"Bioengineering (Basel)"},{"key":"1808_CR5","doi-asserted-by":"publisher","DOI":"10.1016\/j.imavis.2025.105440","author":"A Dastgir","year":"2025","unstructured":"Dastgir A, Bin W, Saeed MU, Sheng J, Saleem S (2025) MAFMv3: an automated multi-scale attention-based feature fusion MobileNetv3 for spine lesion classification. Image Vis Comput. https:\/\/doi.org\/10.1016\/j.imavis.2025.105440","journal-title":"Image Vis Comput"},{"key":"1808_CR6","doi-asserted-by":"publisher","DOI":"10.1016\/j.bspc.2024.105943","author":"MU Saeed","year":"2024","unstructured":"Saeed MU, Bin W, Sheng J, Albarakati HM, Dastgir A (2024) MSFF: an automated multi-scale feature fusion deep learning model for spine fracture segmentation using MRI. Biomed Signal Process Control. https:\/\/doi.org\/10.1016\/j.bspc.2024.105943","journal-title":"Biomed Signal Process Control"},{"key":"1808_CR7","doi-asserted-by":"publisher","DOI":"10.1155\/2022\/7459260","author":"D Saravagi","year":"2022","unstructured":"Saravagi D, Agrawal S, Saravagi M, Chatterjee JM, Agarwal M (2022) Diagnosis of lumbar spondylolisthesis using optimized pretrained CNN models. Comput Intell Neurosci. https:\/\/doi.org\/10.1155\/2022\/7459260","journal-title":"Comput Intell Neurosci"},{"key":"1808_CR8","doi-asserted-by":"publisher","DOI":"10.1016\/j.mlwa.2024.100572","volume":"17","author":"AA Neloy","year":"2024","unstructured":"Neloy AA, Turgeon M (2024) A comprehensive study of auto-encoders for anomaly detection: efficiency and trade-offs. Mach Learn Appl 17:100572. https:\/\/doi.org\/10.1016\/j.mlwa.2024.100572","journal-title":"Mach Learn Appl"},{"key":"1808_CR9","doi-asserted-by":"publisher","DOI":"10.1016\/j.ecoinf.2023.102449","volume":"80","author":"KA Gibb","year":"2024","unstructured":"Gibb KA, Eldridge A, Sandom CJ, Simpson IJA (2024) Towards interpretable learned representations for ecoacoustics using variational auto-encoding. Ecol Inform 80:102449. https:\/\/doi.org\/10.1016\/j.ecoinf.2023.102449","journal-title":"Ecol Inform"},{"issue":"11","key":"1808_CR10","doi-asserted-by":"publisher","first-page":"2911","DOI":"10.1007\/s10439-025-03840-w","volume":"53","author":"M Perrone","year":"2025","unstructured":"Perrone M, Moore DM, Ukeba D, Martin JT (2025) A CNN autoencoder for learning latent disc geometry from segmented lumbar spine MRI. Ann Biomed Eng 53(11):2911\u20132919. https:\/\/doi.org\/10.1007\/s10439-025-03840-w","journal-title":"Ann Biomed Eng"},{"key":"1808_CR11","doi-asserted-by":"publisher","unstructured":"Pikulkaew K. Enhancing brain tumor detection with gradient-weighted class activation mapping and deep learning techniques. In: 2023 20th international joint conference on computer science and software engineering (JCSSE) 2023. p. 339\u201344. https:\/\/doi.org\/10.1109\/JCSSE58229.2023.10202020","DOI":"10.1109\/JCSSE58229.2023.10202020"},{"issue":"1","key":"1808_CR12","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-023-50885-9","volume":"14","author":"S Park","year":"2024","unstructured":"Park S, Kim J-H, Ahn Y, Lee C-H, Kim Y-G, Yuh WT et al (2024) Multi-pose-based convolutional neural network model for diagnosis of patients with central lumbar spinal stenosis. Sci Rep 14(1):203. https:\/\/doi.org\/10.1038\/s41598-023-50885-9","journal-title":"Sci Rep"},{"issue":"13","key":"1808_CR13","doi-asserted-by":"publisher","DOI":"10.3390\/math12131955","volume":"12","author":"M Haider","year":"2024","unstructured":"Haider M, Hashmi MSA, Raza A, Ibrahim M, Fitriyani NL, Syafrudin M et al (2024) Novel ensemble learning algorithm for early detection of lower back pain using spinal anomalies. Mathematics (Basel) 12(13):1955. https:\/\/doi.org\/10.3390\/math12131955","journal-title":"Mathematics (Basel)"},{"key":"1808_CR14","doi-asserted-by":"publisher","unstructured":"Mousavi A. Lumbar spine degenerative classification using YOLO v8 and DeepScoreNet. medRxiv. 2024:2024.12.06.24318595. https:\/\/doi.org\/10.1101\/2024.12.06.24318595.","DOI":"10.1101\/2024.12.06.24318595"},{"issue":"4","key":"1808_CR15","doi-asserted-by":"publisher","first-page":"1095","DOI":"10.2214\/ajr.09.2772","volume":"194","author":"S Lee","year":"2010","unstructured":"Lee S, Lee JW, Yeom JS, Kim K-J, Kim H-J, Chung SK et al (2010) A practical MRI grading system for lumbar foraminal stenosis. Am J Roentgenol 194(4):1095\u20131098. https:\/\/doi.org\/10.2214\/ajr.09.2772","journal-title":"Am J Roentgenol"},{"key":"1808_CR16","doi-asserted-by":"publisher","unstructured":"Sethi D, Nagaraj N, B HN. Neurochaos feature transformation and classification for imbalanced learning. arXiv [csNE]. 2022. https:\/\/doi.org\/10.48550\/arXiv.2205.06742.","DOI":"10.48550\/arXiv.2205.06742"},{"key":"1808_CR17","doi-asserted-by":"publisher","first-page":"30655","DOI":"10.1109\/ACCESS.2022.3158977","volume":"10","author":"A Sharma","year":"2022","unstructured":"Sharma A, Singh PK, Chandra R (2022) SMOTified-GAN for class imbalanced pattern classification problems. IEEE Access 10:30655\u201330665. https:\/\/doi.org\/10.1109\/ACCESS.2022.3158977","journal-title":"IEEE Access"},{"issue":"11","key":"1808_CR18","doi-asserted-by":"publisher","DOI":"10.3390\/app14114570","volume":"14","author":"A Nikolopoulos","year":"2024","unstructured":"Nikolopoulos A, Karalis VD (2024) Implementation of a generative AI algorithm for virtually increasing the sample size of clinical studies. Appl Sci 14(11):4570. https:\/\/doi.org\/10.3390\/app14114570","journal-title":"Appl Sci"},{"key":"1808_CR19","doi-asserted-by":"publisher","first-page":"3","DOI":"10.1007\/978-981-97-0037-0_1","volume-title":"Data engineering and applications","author":"SS Rawat","year":"2024","unstructured":"Rawat SS, Mishra AK (2024) Review of methods for handling class imbalance in classification problems. In: Agrawal J, Shukla RK, Sharma S, Shieh C-S (eds) Data engineering and applications. Springer Nature Singapore, Singapore, pp 3\u201314. https:\/\/doi.org\/10.1007\/978-981-97-0037-0_1"},{"issue":"1","key":"1808_CR20","doi-asserted-by":"publisher","DOI":"10.1038\/s41598-024-84786-2","volume":"15","author":"A Salehi","year":"2025","unstructured":"Salehi A, Khedmati M (2025) Hybrid clustering strategies for effective oversampling and undersampling in multiclass classification. Sci Rep 15(1):3460. https:\/\/doi.org\/10.1038\/s41598-024-84786-2","journal-title":"Sci Rep"},{"key":"1808_CR21","doi-asserted-by":"publisher","DOI":"10.1456\/IJACSA.2021.0120670","author":"\u017d\u00d0 Vujovic","year":"2021","unstructured":"Vujovic \u017d\u00d0 (2021) Classification model evaluation metrics. Int J Adv Comput Sci Appl. https:\/\/doi.org\/10.1456\/IJACSA.2021.0120670","journal-title":"Int J Adv Comput Sci Appl"},{"key":"1808_CR22","doi-asserted-by":"publisher","unstructured":"Sathyanarayanan S, Roopashri Tantri B (2024) Confusion matrix-based performance evaluation metrics. Afr J Biomed Res. 27(4S):4023\u201331. https:\/\/doi.org\/10.53555\/AJBR.v27i4S.4345.","DOI":"10.53555\/AJBR.v27i4S.4345"},{"key":"1808_CR23","doi-asserted-by":"publisher","unstructured":"Jeny AA, Junayed MS, Ahmed I, Habib MT, Rahman MR. FoNet - local food recognition using deep residual neural networks. In: 2019 international conference on information technology (ICIT) 2019. p. 184\u20139. https:\/\/doi.org\/10.1109\/ICIT48102.2019.00039","DOI":"10.1109\/ICIT48102.2019.00039"},{"issue":"7","key":"1808_CR24","doi-asserted-by":"publisher","first-page":"1145","DOI":"10.1016\/S0031-3203(96)00142-2","volume":"30","author":"AP Bradley","year":"1997","unstructured":"Bradley AP (1997) The use of the area under the ROC curve in the evaluation of machine learning algorithms. Pattern Recognit 30(7):1145\u20131159. https:\/\/doi.org\/10.1016\/S0031-3203(96)00142-2","journal-title":"Pattern Recognit"},{"key":"1808_CR25","doi-asserted-by":"publisher","unstructured":"Mannor S, Peleg D, Rubinstein R. The cross entropy method for classification. In: Proceedings of the 22nd international conference on Machine learning. Bonn, Germany: Association for Computing Machinery; 2005. p. 561\u20138. https:\/\/doi.org\/10.1145\/1102351.1102422","DOI":"10.1145\/1102351.1102422"}],"container-title":["Physical and Engineering Sciences in Medicine"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s13246-026-01808-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s13246-026-01808-1","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s13246-026-01808-1.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,9,29]],"date-time":"2026-09-29T02:02:06Z","timestamp":1790647326000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s13246-026-01808-1"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,9,28]]},"references-count":25,"alternative-id":["1808"],"URL":"https:\/\/doi.org\/10.1007\/s13246-026-01808-1","relation":{},"ISSN":["2662-4729","2662-4737"],"issn-type":[{"value":"2662-4729","type":"print"},{"value":"2662-4737","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,9,28]]},"assertion":[{"value":"10 March 2026","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"17 September 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"28 September 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"The authors declare that they have no competing interests.","order":1,"name":"Ethics","label":"Conflict of interest","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"This study used a fully anonymised, publicly available dataset (RSNA Lumbar Spine Competition). No identifiable human data were used; therefore, institutional ethics approval was not required.","order":2,"name":"Ethics","label":"Ethical approval","group":{"name":"EthicsHeading","label":"Declarations"}}]}}