Differentiating magnetic resonance images of pyogenic spondylitis and spinal Modic change using an artificial intelligence

<p>Study Design: Retrospective analysis of magnetic resonance imaging (MRI).</p><p>Introduction: Discriminating between spinal pyogenic spondylitis and Modic change is crucial. A deep-learning approach based on convolutional neural networks (CNNs) is attracting attention in the medical imaging field. The aim of this study was to evaluate the performance of our CNN in differentiating between spinal Pyogenic spondylitis and Modic change on MRI. We compared the performance of the CNN and that of four specialists.</p><p>Materials: Data from patients with spinal pyogenic spondylitis and Modic change who had undergone MRI. There were 50 patients with Modic change and 50 patients with pyogenic spondylitis. Sagittal T1-weighted MRI (T1WI) and sagittal T2-weighted MRI (T2WI) and short TI inversion recovery (STIR) were used for the CNN training and validation. The deep learning framework Tensorflow was used to construct the CNN architecture. To evaluate the performance of the CNN, the receiver operating characteristic curve (ROC) was plotted, and the area under the curve (AUC) was caluculated. We calculated and compared the accuracy, sensitivity, and specificity of the diagnosis by the CNN and a radiologist and a spine surgeon, and two orthopedic surgeons were compared.</p><p>Results: The CNN-based AUCs of the ROC from the T1WI, T2WI, and STIR were 0.95, 0.94, and 0.95, respectively. The accuracy of the CNN was significantly better than that of four specialists on T1WI and STIR (p<.05), and better than a radiologist and one orthopedic surgeon on the T2WI (p<.05). The sensitivity was significantly better than that of four specialists on T1WI and STIR (p<.05), and better than a radiologist and one orthopedic surgeon on the T2WI (p<.05). The specificity was significantly better than one orthopedic surgeon on T1WI and T2WI (p<.05) and better than both orthopedic surgeons on STIR (p<.05).</p><p>Conclusions: Spinal pyogenic spondylitis and Modic change were successfully differentiated using the CNN with high diagnostic accuracy comparable to that of four specialists.</p>