Possibilities of artificial intelligence in classification of spinal pathologies at the present stage of development: a systematic review

INTRODUCTION: A large amount of information, including a growing population and the number of studies significantly increases the workload of a doctor of any specialty, including a radiologist. Digitalization of research makes it possible to facilitate diagnostic processes, because competent software helps a specialist to quickly detect a pathological process, and therefore — to start muchneeded treatment faster. One of the most developing and important software models for these purposes is neural networks. 

OBJECTIVE: To determine the capabilities of neural networks in vertebrology at the current stage of their development in the field of such a neural network function as classification.

MATERIALS AND METHODS: Using the PRISMA protocol, a search was performed in the Pubmed database for the period from January 2017 to December 31, 2023 using keywords. 

RESULTS: For the systematic review, 34 articles were selected that described such a neural network function as classification. 

DISCUSSION: Based on the analysis of literature sources, conclusions were drawn about the usefulness of using artificial intelligence at the current stage of development in vertebrology in such a function as classification. 

CONCLUSION: The classification function of neural networks for pathological changes in the spine is not so clear-cut. Good results in detecting degenerative changes are primarily limited to central spinal stenosis, while the results for detecting lateral recess and foraminal stenosis are inconclusive. For spinal fractures, GoogleNet has emerged as the clear leader. Two networks follow GoogleNet: a proprietary DCNN (Germann C. et al.) and a combination of a multilayer perceptron and 3D radiomics techniques (Chiari-Correia N. S. et al.). Only one network, ResNet 50, excels in the classification of spinal tumors, while the results of other architectures lag significantly behind and require further refinement.

1. Jamshidi M., Rajabian M., Avery M.B. et al. A novel self-expanding primarily bioabsorbable braided flow-diverting stent for aneurysms: initial safety results // Journal of neurointerventional surgery . 2020. Vol. 12, No. 7. P. 700–705. doi: 10.1136/neurintsurg-2019-015555.

2. Swart N., Maroni R., Muldrew B. et al. Economic evaluation of Cytosponge®-trefoil factor 3 for Barrett esophagus: A cost-utility analysis of randomised controlled trial data // EClinicalMedicine . 2021. Vol. 37. doi: 10.1016/j.eclinm.2021.100969.

3. Albers G.W., Marks M.P., Kemp S. et al. Thrombectomy for Stroke at 6 to 16 Hours with Selection by Perfusion Imaging: Albers G.W., Marks M.P., Kemp S. et al. // N. Engl. J. Med . 2018. Vol. 378. Р. 708–718 // Journal of Emergency Medicine . 2018. Vol. 55, No. 1. P. 151. doi: 10.1056/NEJMoa1713973

4. Thong L.T., Chou H.S., Chew H.S.J., Lau Y. Diagnostic test accuracy of artificial intelligence-based imaging for lung cancer screening: A systematic review and metaanalysis // Lung Cancer . 2023. Vol. 176. P. 4–13. doi: 10.1016/j.lungcan.2022.12.002.

5. Gautam N., Saluja P., Malkawi A. et al. Current and future applications of artificial intelligence in coronary artery disease // Healthcare. MDPI . 2022. Vol. 10, No. 2. P. 232. doi: 10.3390/healthcare10020232.

6. Roberts M., Driggs D., Thorpe M. et al. Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans // Nature Machine Intelligence . 2021. Vol. 3, No. 3. P. 199–217. doi: 10.1038/s42256-021-00307-0.

7. Wang Z., Liu X., Hu X. et al. Age-related changes in cervical sagittal alignment: based on 625 Chinese asymptomatic subjects // European Spine Journal . 2023. Vol. 32, No. 5. P. 1607–1615. doi: 10.1007/s00586-023-07632-5.

8. Tricco A.C., Lillie E., Zarin W., O’Brien K.K., Colquhoun H., Levac. PRISMA extension for scoping reviews (PRISMA-ScR): checklist and explanation // Annals of internal medicine . 2018. Vol. 169, No. 7. P. 467–473. doi: 10.7326/M18-0850.

9. Liawrungrueang W., Kim P., Kotheeranurak V., Jitpakdee K., Sarasombath Automatic detection, classification, and grading of lumbar intervertebral disc degeneration using an artificial neural network model // Diagnostics . 2023. Vol. 13, No. 4. P. 663. doi: 10.3390/diagnostics13040663.

10. Mastouri R., Khlifa N., Neji H., Hantous-Zannad S. A bilinear convolutional neural network for lung nodules classification on CT images // International Journal of Computer Assisted Radiology and Surgery . 2021. Vol. 16. P. 91–101. doi: 10.1007/s11548-020-02283-z.

11. Merali Z., Wang J.Z., Badhiwala J.H. et al. A deep learning model for detection of cervical spinal cord compression in MRI scans // Scientific reports . 2021. Vol. 11, No. 1. P. 1–11. doi: 10.1038/s41598-021-89848-3.

12. Kumar K.K., Gopal T.V. A novel approach to self order feature reweighting in CBIR to reduce semantic gap using Relevance Feedback // 2014 International Conference on Circuits, Power and Computing Technologies [ICCPCT-2014] . 2014. P. 1437–1442. doi: 10.1109/ICCPCT.2014.7054919.

13. Abdullina K.M., Spivak A.I. Application of the U-Net neural network architecture for the problem of localizing integrity violations of digital images. Scientific and Technical Bulletin of Information Technologies, Mechanics and Optics , 2020, Vol. 20, No. 3, рр. 425–431 (In Russ.). doi: 10.17586/2226-1494-2020-20-3-425-431.

14. Hallinan J.T.P.D., Zhu L., Yang K… et al. Deep learning model for automated detection and classification of central canal, lateral recess, and neural foraminal stenosis at lumbar spine MRI // Radiology . 2021. Vol. 300, No. 1. P. 130–138. doi: 10.1148/radiol.2021204289.

15. Lehnen N.C., Haase R., Faber J. et al. Detection of degenerative changes on MR images of the lumbar spine with a convolutional neural network: a feasibility study // Diagnostics . 2021. Vol. 11, No. 5. P. 902. doi: 10.3390/diagnostics11050902.

16. Jamaludin A., Lootus M., Kadir T. et al. ISSLS prize in bioengeneering science 2017: Automation of reading of radiological features from magnetic resonance images (MRIs) of the lumbar spine without human intervention is comparable with an expert radiologist // European spine journal . 2017. Vol. 26, No. 5. P. 1374–1383. doi: 10.1007/s00586-017-4956-3.

17. Lu J. T., Pedemonte S., Bizzo B. et al. Deep Spine: Automated lumbar vertebral segmentation, disc-level designation, and spinal stenosis grading using deep learning // Machine Learning for Healthcare Conference // PMLR . 2018. P. 403–419. doi: 10.48550/arXiv.1807.10215.

18. Won D., Lee H.J., Lee S.J., Park S.H. Spinal stenosis grading in magnetic resonance imaging using deep convolutional neural networks // Spine . 2020. Vol. 45, No. 12. P. 804–812. doi: 10.1097/BRS.0000000000003377.

19. Han Z., Wei B., Leung S., Chung J., Li S. Towards automatic report generation in spine radiology using weakly supervised framework // Medical Image Computing and Computer Assisted Intervention–MICCAI . 2018: 21[st] International Conference, Granada, Spain, September 16–20, 2018, Proceedings, Part IV 11. Springer International Publishing, 2018. P. 185–193. doi: 10.1007/978-3-030-00937-3_22.

20. Merali Z., Wang J.Z., Badhiwala J.H. et al. A deep learning model for detection of cervical spinal cord compression in MRI scans // Scientific reports . 2021. Vol. 11, No. 1. P. 1–11. doi: 10.1038/s41598-021-89848-3.

21. Lewandrowski K.U., Muraleedharan N., Eddy S.A. et al. Feasibility of deep learning algorithms for reporting in routine spine magnetic resonance imaging // International Journal of Spine Surgery. 2020. Vol. 14, No. s3. P. S86-S97. doi: 10.14444/7131.

22. Ghauri M.S., Reddy A.J., Tak N. et al. Utilizing deep learning for X-ray imaging: detecting and classifying degenerative spinal conditions // Cureus . 2023. Vol. 15, No. 7. doi: 10.7759/cureus.41582.

23. Gao K. T., Tibrewala R., Hess M., Bharadwaj U.U., Inamdar G. et al. Automatic detection and voxel‐wise mapping of lumbar spine Modic changes with deep learning // JOR spine . 2022. Vol. 5, No. 2. P. e1204. doi: 10.1002/jsp2.1204.

24. Zhang W., Chen Z., Su Z. et al. Deep learning‐based detection and classification of lumbar disc herniation on magnetic resonance images // JOR spine . 2023. Vol. 6, No. 3. P. e1276. doi: 10.1002/jsp2.1276.

25. Bharadwaj U.U., Christine M., Li S. et al. Deep learning for automated, interpretable classification of lumbar spinal stenosis and facet arthropathy from axial MRI // European radiology . 2023. Vol. 33, No. 5. P. 3435–3443. doi: 10.1007/s00330-023-09483-6.

26. Liawrungrueang W., Kim P., Kotheeranurak V., Jitpakdee K., Sarasombath P. Automatic detection, classification, and grading of lumbar intervertebral disc degeneration using an artificial neural network model // Diagnostics . 2023. Vol. 13, No. 4. P. 663. doi: 10.3390/diagnostics13040663.

27. Dong Q. Deep Learning Classification of Spinal Osteoporotic Compression Fractures on Radiographs using an Adaptation of the Genant Semiquantitative Criteria // Academic radiology . 2022. Vol. 29, No. 12. P. 1819–1832. doi: 10.1016/j.acra.2022.02.020.

28. Germann C., Meyer A. N., Staib M., Sutter R., Fritz B. Performance of a deep convolutional neural network for MRI-based vertebral body measurements and insufficiency fracture detection // European Radiology . 2023. Vol. 33, No. 5. P. 3188–3199. doi: 10.1007/s00330-022-09354-6.

29. Chiari-Correia N.S., Nogueira-Barbosa M.H., Chiari-Correia R.D., Azevedo-Marques P.M. A 3D radiomics-based artificial neural network model for benign versus malignant vertebral compression fracture classification in MRI // Journal of Digital Imaging . 2023. Vol. 36, No. 4. P. 1565–1577. doi: 10.1007/s10278-023-00847-4.

30. Doerr S. A., Weber-Levine C., Hersh A.M. et al. Automated prediction of the Thoracolumbar Injury Classification and Severity Score from CT using a novel deep learning algorithm // Neurosurgical focus . 2022. Vol. 52, No. 4. P. E5. doi: 10.3171/2022.1.FOCUS21745.

31. Bar A., Wolf L., Amitai O.B., Toledano E., Elnekave E. Compression fractures detection on CT // Medical imaging 2017: computer-aided diagnosis // SPIE . 2017. Vol. 10134. P. 1036–1043. doi: 10.1117/12.2249635.

32. Small J.E., Osler P., Paul A.B., Kunst M. CT cervical spine fracture detection using a convolutional neural network // American Journal of Neuroradiology . 2021. Vol. 42, No. 7. P. 1341–1347. doi: 10.3174/ajnr.A7094.

33. Tomita N., Cheung Y.Y., Hassanpour S. Deep neural networks for automatic detection of osteoporotic vertebral fractures on CT scans // Computers in biology and medicine . 2018. Vol. 98. P. 8–15. doi: 10.1016/j.compbiomed.2018.05.011.

34. Derkatch S., Kirby C., Kimelman D. et al. Identification of vertebral fractures by convolutional neural networks to predict nonvertebral and hip fractures: a registrybased cohort study of dual X-ray absorptiometry // Radiology . 2019. Vol. 293, No. 2. P. 405–411. doi: 10.1148/radiol.2019190201.

35. Zhang J., Liu F., Xu J. et al. Automated detection and classification of acute vertebral body fractures using a convolutional neural network on computed tomography // Frontiers in endocrinology . 2023. Vol. 14. P. 1132725. doi: 10.3389/fendo.2023.1132725.

36. Dong Q., Luo G., Lane N.E. et al. Generalizability of deep learning classification of spinal osteoporotic compression fractures on radiographs using an adaptation of the modified-2 algorithm-based qualitative criteria // Academic radiology . 2023. Vol. 30, No. 12. P. 2973–2987. doi: 10.1016/j.acra.2023.04.023.

37. Boonrod A. et al. Diagnostic accuracy of deep learning for evaluation of C-spine injury from lateral neck radiographs // Ш. 2022. Vol. 8, No. 8. doi: 10.1016/j.heliyon.2022.e10372.

38. Chen X.G., Liu Y.A. classification method for thoracolumbar vertebral fractures due to basketball sports injury based on deep learning // Computational and mathematical methods in medicine . 2022. Vol. 2022, No. 1. P. 8747487. doi: 10.1155/2022/8747487.

39. Yeh L.R., Zhang Y., Chen J.H. et al. A deep learning-based method for the diagnosis of vertebral fractures on spine MRI: retrospective training and validation of ResNet // European Spine Journal . 2022. P. 1–9. doi: 10.1007/s00586-022-07121-1.

40. Liu B., Jin Y., Feng S. et al. Benign vs malignant vertebral compression fractures with MRI: a comparison between automatic deep learning network and radiologist’s assessment // European Radiology . 2023. Vol. 33, No. 7. P. 5060–5068. doi: 10.1007/s00330-023-09713-x.

41. Xu F., Xiong Y., Ye G. et al. Deep learning-based artificial intelligence model for classification of vertebral compression fractures: A multicenter diagnostic study // Frontiers in Endocrinology . 2023. Vol. 14. P. 1025749. doi: 10.3389/fendo.2023.1025749.

42. Lang N., Zhang Y., Zhang E. et al. Differentiation of spinal metastases originated from lung and other cancers using radiomics and deep learning based on DCE-MRI // Magnetic resonance imaging . 2019. Vol. 64. P. 4. doi: 10.1016/j.mri.2019.02.013.

43. Chen K., Cao J., Zhang X. et al. Differentiation between spinal multiple myeloma and metastases originated from lung using multi-view attention-guided network // Frontiers in Oncology . 2022. Vol. 12. P. 981769. doi: 10.3389/fonc.2022.981769.

44. Hallinan J.T.P.D., Zhu L., Zhang W. et al. Deep learning model for classifying metastatic epidural spinal cord compression on MRI // Frontiers in Oncology . 2022. Vol. 12. P. 849447. doi: 10.3389/fonc.2022.849447.

45. Xiong X., Wang J., Hu S. et al. Differentiating between multiple myeloma and metastasis subtypes of lumbar vertebra lesions using machine learning-based radiomics // Frontiers in Oncology . 2021. Vol. 11. P. 601699. doi: 10.3389/fonc.2021.601699.

46. Hallinan J.T.P.D., Zhu L., Zhang W. et al. Deep learning assessment compared to radiologist reporting for metastatic spinal cord compression on CT // Frontiers in Oncology . 2023. Vol. 13. P. 1151073. doi: 10.3389/fonc.2023.1151073.

47. Ito S., Nakashima H., Segi N. et al. Automated detection and diagnosis of spinal schwannomas and meningiomas using deep learning and magnetic resonance imaging // J. Clin. Med . 2023. Vol. 12, No. 15. P. 5075. doi: 10.3390/jcm12155075.