Cobb Angle Measurement of Scoliosis with Reduced Variability
Cobb angle, which is a measure of spinal curvature is the standard method for quantifying the magnitude of Scoliosis related to spinal deformity in orthopedics. Determining the Cobb angle through manual process is subject to human errors. In this work, we propose a methodology to measure the magnitude of Cobb angle, which appreciably reduces the variability related to its measurement compared to the related works. The proposed methodology is facilitated by using a suitable new improved version of Non-Local Means for image denoisation and Otsus automatic threshold selection for Canny edge detection. We have selected NLM for preprocessing of the image as it is one of the fine states of art for image denoisation and helps in retaining the image quality. Trimmedmean, median are more robust to outliners than mean and following this concept we observed that NLM denoising quality performance can be enhanced by using Euclidean trimmed-mean replacing the mean. To prove the better performance of the Non-Local Euclidean Trimmed-mean denoising filter, we have provided some comparative study results of the proposed denoising technique with traditional NLM and NonLocal Euclidean Medians. The experimental results for Cobb angle measurement over intra observer and inter observer experimental data reveals the better performance and superiority of the proposed approach compared to the related works. MATLAB2009b image processing toolbox was used for the purpose of simulation and verification of the proposed methodology.