AI-based lumbar central canal stenosis classification on sagittal MR images is comparable to experienced radiologists using axial images

Abstract

Objectives: The assessment of lumbar central canal stenosis (LCCS) is crucial for diagnosing and planning treatment for patients with low back pain and neurogenic pain. However, manual assessment methods are time-consuming, variable, and require axial MRIs. The aim of this study is to develop and validate an AI-based model that automatically classifies LCCS using sagittal T2-weighted MRIs.

Methods: A pre-existing 3D AI algorithm was utilized to segment the spinal canal and intervertebral discs (IVDs), enabling quantitative measurements at each IVD level. Four musculoskeletal radiologists graded 683 IVD levels from 186 LCCS patients using the 4-class Lee grading system. A second consensus reading was conducted by readers 1 and 2, which, along with automatic measurements, formed the training dataset for a multiclass (grade 0-3) and binary (grade 0-1 vs. 2-3) random forest classifier with tenfold cross-validation.

Results: The multiclass model achieved a Cohen's weighted kappa of 0.86 (95% CI: 0.82-0.90), comparable to readers 3 and 4 with 0.85 (95% CI: 0.80-0.89) and 0.73 (95% CI: 0.68-0.79) respectively. The binary model demonstrated an AUC of 0.98 (95% CI: 0.97-0.99), sensitivity of 93% (95% CI: 91-96%), and specificity of 91% (95% CI: 87-95%). In comparison, readers 3 and 4 achieved a specificity of 98 and 99% and sensitivity of 74 and 54%, respectively.

Conclusion: Both the multiclass and binary models, while only using sagittal MR images, perform on par with experienced radiologists who also had access to axial sequences. This underscores the potential of this novel algorithm in enhancing diagnostic accuracy and efficiency in medical imaging.

Key points: Question How can the classification of lumbar central canal stenosis (LCCS) be made more efficient? Findings Multiclass and binary AI models, using only sagittal MR images, performed on par with experienced radiologists who also had access to axial sequences. Clinical relevance Our AI algorithm accurately classifies LCCS from sagittal MRI, matching experienced radiologists. This study offers a promising tool for automated LCCS assessment from sagittal T2 MRI, potentially reducing the reliance on additional axial imaging.

Keywords: Deep learning; Lumbar central canal stenosis; MRI; Machine learning; Spine.

Fig. 1

Flow diagram of MRI study selection

Fig. 2

An MRI study paired with its automatically generated segmentation mask is presented. a Depicts the mid-sagittal slice from a series of sagittal T2-weighted MRI images. b Illustrates an axial slice at the L4-L5 intervertebral disc level. c Displays a reconstructed axial view, derived from the sagittal MRI series, corresponding to the level of image b

Fig. 3

Visualization of automatically generated measurements, which includes the dural sac antero-posterior diameter (APD) in green and the dural sac cross-sectional area (CSA) in red

Fig. 4

Visualization of the angulations of all measurements. Angulation was determined for each intervertebral disc, with absolute measurements extracted from the green volumes, utilizing only the measurements from the most stenotic plane within that volume. These absolute measurements were then compared to the cranially adjacent mid-vertebral measurements (presented by red lines) to derive the relative measurement values (the ratio-based values)

Fig. 5

Confusion matrices of the multiclass random forest model, Reader 3, and Reader 4. The results are compared to the ground truth, which is the consensus reading

Fig. 6

Receiver operating characteristic (ROC) curves of all binary models with different input metrics. The performances of Reader 3 and 4 are also shown, compared to the consensus. CSA, cross-sectional area; APD, antero-posterior diameter; FSL, fluid signal loss; AUC, area under the curve