RESUMO
BACKGROUND: Artificial intelligence is gaining traction in automated medical imaging analysis. Development of more accurate magnetic resonance imaging (MRI) predictors of successful clinical outcomes is necessary to better define indications for surgery, improve clinical outcomes with targeted minimally invasive and endoscopic procedures, and realize cost savings by avoiding more invasive spine care. OBJECTIVE: To demonstrate the ability for deep learning neural network models to identify features in MRI DICOM datasets that represent varying intensities or severities of common spinal pathologies and injuries and to demonstrate the feasibility of generating automated verbal MRI reports comparable to those produced by reading radiologists. METHODS: A 3-dimensional (3D) anatomical model of the lumbar spine was fitted to each of the patient's MRIs by a team of technicians. MRI T1, T2, sagittal, axial, and transverse reconstruction image series were used to train segmentation models by the intersection of the 3D model through these image sequences. Class definitions were extracted from the radiologist report for the central canal: (0) no disc bulge/protrusion/canal stenosis, (1) disc bulge without canal stenosis, (2) disc bulge resulting in canal stenosis, and (3) disc herniation/protrusion/extrusion resulting in canal stenosis. Both the left and right neural foramina were assessed with either (0) neural foraminal stenosis absent, or (1) neural foramina stenosis present. Reporting criteria for the pathologies at each disc level and, when available, the grading of severity were extracted, and a natural language processing model was used to generate a verbal and written report. These data were then used to train a set of very deep convolutional neural network models, optimizing for minimal binary cross-entropy for each classification. RESULTS: The initial prediction validation of the implemented deep learning algorithm was done on 20% of the dataset, which was not used for artificial intelligence training. Of the 17,800 total disc locations for which MRI images and radiology reports were available, 14,720 were used to train the model, and 3560 were used to validate against. The convergence of validation accuracy achieved with the deep learning algorithm for the foraminal stenosis detector was 81% (sensitivity = 72.4.4%, specificity = 83.1%) after 25 complete iterations through the entire training dataset (epoch). The accuracy was 86.2% (sensitivity = 91.1%, specificity = 82.5%) for the central stenosis detector and 85.2% (sensitivity = 81.8%, specificity = 87.4%) for the disc herniation detector. CONCLUSIONS: Deep learning algorithms may be used for routine reporting in spine MRI. There was a minimal disparity among accuracy, sensitivity, and specificity, indicating that the data were not overfitted to the training set. We concluded that variability in the training data tends to reduce overfitting and overtraining as the deep neural network models learn to focus on the common pathologies. Future studies should demonstrate the accuracy of deep neural network models and the predictive value of favorable clinical outcomes with intervention and surgery. LEVEL OF EVIDENCE: 3. CLINICAL RELEVANCE: Feasibility, clinical teaching, and evaluation study.
RESUMO
BACKGROUND: Identifying pain generators in multilevel lumbar degenerative disc disease is not trivial but is crucial for lasting symptom relief with the targeted endoscopic spinal decompression surgery. Artificial intelligence (AI) applications of deep learning neural networks to the analysis of routine lumbar MRI scans could help the primary care and endoscopic specialist physician to compare the radiologist's report with a review of endoscopic clinical outcomes. OBJECTIVE: To analyze and compare the probability of predicting successful outcome with lumbar spinal endoscopy by using the radiologist's MRI grading and interpretation of the radiologic image with a novel AI deep learning neural network (Multus Radbot™) as independent prognosticators. METHODS: The location and severity of foraminal stenosis were analyzed using comparative ordinal grading by the radiologist, and a contiguous grading by the AI network in patients suffering from lateral recess and foraminal stenosis due to lumbar herniated disc. The compressive pathology definitions were extracted from the radiologist lumbar MRI reports from 65 patients with a total of 383 levels for the central canal - (0) no disc bulge/protrusion/canal stenosis, (1) disc bulge without canal stenosis, (2) disc bulge resulting in canal stenosis, and (3) disc herniation/protrusion/extrusion resulting in canal stenosis. Both neural foramina were assessed with either - (0) neural foraminal stenosis absent, or (1) neural foramina are stenosis present. Reporting criteria for the pathologies at each disc level and, when available, the grading of severity were extracted and assigned into two categories: "Normal," and "Stenosis." Clinical outcomes were graded using dichotomized modified Macnab criteria considering Excellent and Good results as "Improved," and Fair and Poor outcomes as "Not Improved." Binary logistic regression analysis was used to predict the probability of the AI- and radiologist grading of stenosis at the 88 foraminal decompression sites to result in "Improved" outcomes. RESULTS: The average age of the 65 patients was 62.7 +/- 12.7 years. They consisted of 51 (54.3%) males and 43 (45.7%) females. At an average final follow-up of 57.4 +/- 12.57, Macnab outcome analysis showed that 86.4% of the 88 foraminal decompressions resulted in Excellent and Good (Improved) clinical outcomes. The stenosis grading by the radiologist showed an average severity score of 4.71 +/- 2.626, and the average AI severity grading was 5.65 +/- 3.73. Logit regression probability analysis of the two independent prognosticators showed that both the grading by the radiologist (86.2%; odds ratio 1.264) and the AI grading (86.4%; odds ratio 1.267) were nearly equally predictive of a successful outcome with the endoscopic decompression. CONCLUSIONS: Deep learning algorithms are capable of identifying lumbar foraminal compression due to herniated disc. The treatment outcome was correlated to the decompression of the directly visualized corresponding pathology during the lumbar endoscopy. This research should be extended to other validated pain generators in the lumbar spine. LEVEL OF EVIDENCE: 3. CLINICAL RELEVANCE: Validity, clinical teaching, evaluation study.
RESUMO
BACKGROUND: Artificial intelligence could provide more accurate magnetic resonance imaging (MRI) predictors of successful clinical outcomes in targeted spine care. OBJECTIVE: To analyze the level of agreement between lumbar MRI reports created by a deep learning neural network (RadBot) and the radiologists' MRI reading. METHODS: The compressive pathology definitions were extracted from the radiologist lumbar MRI reports from 65 patients with a total of 383 levels for the central canal: (0) no disc bulge/protrusion/canal stenosis, (1) disc bulge without canal stenosis, (2) disc bulge resulting in canal stenosis, and (3) disc herniation/protrusion/extrusion resulting in canal stenosis. For both, neural foramina were assessed with either (0) neural foraminal stenosis absent or (1) neural foramina stenosis present. Reporting criteria for the pathologies at each disc level and, when available, the grading of severity were extracted, and the Natural Language Processing model was used to generate a verbal and written report. The RadBot report was analyzed similarly as the MRI report by the radiologist. MRI reports were investigated by dichotomizing the data into 2 categories: normal and stenosis. The quality of the RadBot test was assessed by determining its sensitivity, specificity, and positive and negative predictive value as well as its reliability with the calculation of the Cronbach alpha and Cohen kappa using the radiologist MRI report as a gold standard. RESULTS: The authors found a RadBot sensitivity of 73.3%, a specificity of 88.4%, a positive predictive value of 80.3%, and a negative predictive value of 83.7%. The reliability analysis revealed the Cronbach alpha as 0.772. The highest individual values of the Cronbach alpha were 0.629 and 0.681 when compared to the MRI report by the radiologist, rending values of 0.566 and 0.688, respectively. Analysis of interobserver reliability rendered an overall kappa for the RadBot of 0.627. Analysis of receiver operating characteristics (ROC) showed a value of 0.808 for the area under the ROC curve. CONCLUSIONS: Deep learning algorithms, when used for routine reporting in lumbar spine MRI, showed excellent quality as a diagnostic test that can distinguish the presence of neural element compression (stenosis) at a statistically significant level (P < .0001) from a random event distribution. This research should be extended to validated and directly visualized pain generators to improve the accuracy and prognostic value of the routine lumbar MRI scan for favorable clinical outcomes with intervention and surgery. LEVEL OF EVIDENCE: 3. CLINICAL RELEVANCE: Validity, clinical teaching, and evaluation study.