Learning-Based Coronal Spine Alignment Prediction Using Smartphone-Acquired Scoliosis Radiograph Images

DICOM X-rays are not easily accessible for telemedicine, and existing learning-based automated Cobb angle (CA) predictions are not accurate on suboptimal X-ray images. To develop an automated CA prediction system irrespective of image quality, with no restrictions on curve patterns, 367 consecutive patients attending our scoliosis clinic were recruited and their coronal X-rays were re-captured using mobile phones. Five-fold cross-validation was conducted (each with 294 randomly selected images for training a neural network SpineHRNet to detect endplate landmarks and end-vertebrae, and the remaining 73 images for testing). The predicted heatmaps of vertebral landmarks were visualized to enhance interpretability of the SpineHRNet. Per-landmark Euclidean distance (L2) errors and recall of landmark detection were calculated to assess the accuracy of the predicted landmarks. Further computed CAs were quantitatively compared with spine-specialists measured ground truth (GT). The average L2 error and the recall of the detected endplates landmarks were 2.8 pixels and 0.99 respectively. The predicted CAs were all significantly correlated with GT (
$\text{p} < 0.01$
). Compared with GT, the mean absolute error was 3.73-4.15° and standard deviation was 0.8-1.7° for the predicted CAs at different spinal regions. This is the first study on non-original X-rays to automatically and accurately predict endplate landmarks of the scoliotic spine and compute the CAs at different regions of the spine, irrespective of image qualities. SpineHRNet’s applicability is evidenced by five-fold cross-validations, which may be used with telemedicine to facilitate fast and reliable auto-diagnosis and follow-up.