Hairline breakage detection in X-ray images using data fusion

This paper deals with identification of hairline breakage in the X-Ray images. The crack in the X-Ray images can be missed due to absence of sharp edges and the intensity inhomogeneity. This work is carried out in two phases. In the first phase the preprocessing step is done using anisotropic diffusion filter and wavelet to preserve the edges and fine details. In the second phase Expectation Maximization (EM) algorithm is used for segmenting the image. The mask produced from the EM algorithm separates the bone region. The intensity variation calculation is performed over the selected region to detect the cracks. The performance of the proposed work is calculated using the parameters sensitivity and accuracy. This new approach is experimented with ten patient’s data and validated by Radiologists. The performance of the proposed work is compared with recent works. This work greatly improves the accuracy of the segmentation on medical images and the overall accuracy is about 98%.