Machine Learning Pipeline for Predicting Bone Marrow Edema Along the Sacroiliac Joints on Magnetic Resonance Imaging.
Female
Humans
Sacroiliac Joint
/ diagnostic imaging
Bone Marrow
/ diagnostic imaging
Artificial Intelligence
Spondylarthritis
/ pathology
Bone Marrow Diseases
/ diagnostic imaging
Inflammation
/ pathology
Magnetic Resonance Imaging
/ methods
Edema
/ diagnostic imaging
Machine Learning
Sacroiliitis
/ pathology
Journal
Arthritis & rheumatology (Hoboken, N.J.)
ISSN: 2326-5205
Titre abrégé: Arthritis Rheumatol
Pays: United States
ID NLM: 101623795
Informations de publication
Date de publication:
Dec 2023
Dec 2023
Historique:
revised:
22
06
2023
received:
19
12
2022
accepted:
26
06
2023
medline:
30
11
2023
pubmed:
6
7
2023
entrez:
6
7
2023
Statut:
ppublish
Résumé
We aimed to develop and validate a fully automated machine learning (ML) algorithm that predicts bone marrow edema (BME) on a quadrant level in sacroiliac (SI) joint magnetic resonance imaging (MRI). A computer vision workflow automatically locates the SI joints, segments regions of interest (ilium and sacrum), performs objective quadrant extraction, and predicts presence of BME, suggestive of inflammatory lesions, on a quadrant level in semicoronal slices of T1/T2-weighted MRI scans. Ground truth was determined by consensus among human readers. The inflammation classifier was trained using a ResNet18 backbone and five-fold cross-validated on scans of patients with spondyloarthritis (SpA) (n = 279), postpartum individuals (n = 71), and healthy subjects (n = 114). Independent SpA patient MRI scans (n = 243) served as test data set. Patient-level predictions were derived from aggregating quadrant-level predictions, ie, at least one positive quadrant. The algorithm automatically detects the SI joints with a precision of 98.4% and segments ilium/sacrum with an intersection over union of 85.6% and 67.9%, respectively. The inflammation classifier performed well in cross-validation: area under the curve (AUC) 94.5%, balanced accuracy (B-ACC) 80.5%, and F1 score 64.1%. In the test data set, AUC was 88.2%, B-ACC 72.1%, and F1 score 50.8%. On a patient level, the model achieved a B-ACC of 81.6% and 81.4% in the cross-validation and test data set, respectively. We propose a fully automated ML pipeline that enables objective and standardized evaluation of BME along the SI joints on MRI. This method has the potential to screen large numbers of patients with (suspected) SpA and is a step closer towards artificial intelligence-assisted diagnosis and follow-up.
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
2169-2177Subventions
Organisme : Flanders AI Research Program
Informations de copyright
© 2023 American College of Rheumatology.
Références
Mandl P, Navarro-Compan V, Terslev L, et al. EULAR recommendations for the use of imaging in the diagnosis and management of spondyloarthritis in clinical practice. Ann Rheum Dis 2015;74:1327-1339.
OECD Data. Magnetic resonance imaging (MRI) exams (indicator). 2022. Accessed June 27, 2022. https://data.oecd.org/healthcare/magnetic-resonance-imaging-mri-exams.htm
De Winter J, de Hooge M, van de Sande M, et al. Magnetic resonance imaging of the sacroiliac joints indicating sacroiliitis according to the Assessment of SpondyloArthritis International Society definition in healthy individuals, runners, and women with postpartum back pain. Arthritis Rheumatol 2018;70:1042-1048.
Renson T, Depicker A, De Craemer AS, et al. High prevalence of spondyloarthritis-like MRI lesions in postpartum women: a prospective analysis in relation to maternal, child and birth characteristics. Ann Rheum Dis 2020;79:929-934.
Maksymowych WP, Inman RD, Salonen D, et al. Spondyloarthritis Research Consortium of Canada magnetic resonance imaging index for assessment of sacroiliac joint inflammation in ankylosing spondylitis. Arthritis Rheum 2005;53:703-709.
Maksymowych WP, Lambert RG, Baraliakos X, et al. Data-driven definitions for active and structural MRI lesions in the sacroiliac joint in spondyloarthritis and their predictive utility. Rheumatology (Oxford) 2021;60:4778-4789.
Landewé RB, Hermann KG, van der Heijde DM, et al. Scoring sacroiliac joints by magnetic resonance imaging. A multiple-reader reliability experiment. J Rheumatol 2005;32:2050-2055.
Zhao SS, Hong C, Cai T, et al. Incorporating natural language processing to improve classification of axial spondyloarthritis using electronic health records. Rheumatology (Oxford) 2020;59:1059-1065.
Shenkman Y, Qutteineh B, Joskowicz L, et al. Automatic detection and diagnosis of sacroiliitis in CT scans as incidental findings. Med Image Anal 2019;57:165-175.
Bressem KK, Vahldiek JL, Adams L, et al. Deep learning for detection of radiographic sacroiliitis: achieving expert-level performance. Arthritis Res Ther 2021;23:106.
Faleiros MC, Nogueira-Barbosa MH, Dalto VF, et al. Machine learning techniques for computer-aided classification of active inflammatory sacroiliitis in magnetic resonance imaging. Adv Rheumatol 2020;60:25.
Lee KH, Choi ST, Lee GY, et al. Method for diagnosing the bone marrow edema of sacroiliac joint in patients with axial spondyloarthritis using magnetic resonance image analysis based on deep learning. Diagnostics (Basel) 2021;11:1156.
Rzecki K, Kucybała I, Gut D, et al. Fully automated algorithm for the detection of bone marrow oedema lesions in patients with axial spondyloarthritis - feasibility study. Biocybern Biomed Eng 2021;41:833-853.
Kucybała I, Tabor Z, Polak J, et al. The semi-automated algorithm for the detection of bone marrow oedema lesions in patients with axial spondyloarthritis. Rheumatol Int 2020;40:625-633.
De Craemer AS, Renson T, Deroo L, et al. Peripheral manifestations are major determinants of disease phenotype and outcome in new onset spondyloarthritis. Rheumatology (Oxford) 2021;61:3279-3288.
Renson T, de Hooge M, De Craemer AS, et al. Progressive increase in sacroiliac joint and spinal MRI lesions in healthy individuals in relation to age. Arthritis Rheumatol 2022;74:1506-1514.
Weber U, Maksymowych WP, Chan SM, et al. Does evaluation of the ligamentous compartment enhance diagnostic utility of sacroiliac joint MRI in axial spondyloarthritis? Arthritis Res Ther 2015;17:246.
Jans L, Coeman L, Van Praet L, et al. How sensitive and specific are MRI features of sacroiliitis for diagnosis of spondyloarthritis in patients with inflammatory back pain? JBR-BTR 2014;97:202-205.
Tan M, Pang R, Le Q. EfficientDet: scalable and efficient object detection. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition. IEEE Computer Society; 2020:10778-10787.
Berg S, Kutra D, Kroeger T, et al. ilastik: interactive machine learning for (bio)image analysis [review]. Nat Methods 2019;16:1226-1232.
Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation. In: Navab N, Hornegger J, Wells WM, et al, eds. Medical Image Computing and Computer-Assisted Intervention - MICCAI 2015. Springer; 2015:234-241.
Mirikharaji Z, Yan Y, Hamarneh G. Learning to segment skin lesions from noisy annotations. arXiv doi:10.48550/arXiv.1906.03815. Preprint posted online June 10, 2019.
Yu S, Chen M, Zhang E, et al. Robustness study of noisy annotation in deep learning based medical image segmentation. Phys Med Biol 2020;65:175007.
He K, Zhang X, Ren S, et al. Deep residual learning for image recognition. In: 2016 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE Computer Society; 2016:770-778.
Lambert RG, Bakker PA, van der Heijde D, et al. Defining active sacroiliitis on MRI for classification of axial spondyloarthritis: update by the ASAS MRI working group [review]. Ann Rheum Dis 2016;75:1958-1963.
Zhou B, Khosla A, Lapedriza A, et al. Learning deep features for discriminative localization. In: 2016 IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computer Society; 2016:2921-2929.
Selvaraju RR, Cogswell M, Das A, et al. Grad-CAM: visual explanations from deep networks via gradient-based localization. Int J Comput Vision 2020;128:336-359.
Sundararajan M, Taly A, Yan Q. Axiomatic attribution for deep networks. arXiv Preprint posted online June 13, 2017. https://doi.org/10.48550/arXiv.1703.01365
Wang Y, Zhang J, Kan M, et al. Self-supervised equivariant attention mechanism for weakly supervised semantic segmentation. In: 2020 IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computer Society; 2020:12272-12281.
Krizhevsky A, Sutskever I, Hinton G. ImageNet classification with deep convolutional neural networks. In: Pereira F, Burges CJ, Bottou L, et al. Advances in Neural Information Processing Systems 25 (NIPS 2012). Curran Associates; 2012:1097-1105.
Xie S, Girshick R, Dollár P, et al. Aggregated residual transformations for deep neural networks. In: 2017 IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computing Society; 2017:5987-5995.
Huang G, Liu Z, Weinberger KQ. Densely connected convolutional networks. In: 2017 IEEE Conference on Computer Vision and Pattern Recognition. IEEE Computing Society; 2017:2261-2269.
Lin KY, Peng C, Lee KH, et al. Deep learning algorithms for magnetic resonance imaging of inflammatory sacroiliitis in axial spondyloarthritis. Rheumatology (Oxford) 2022;61:4198-4206.