Contrast Enhancement of 3D X-ray Microtomography Using CLAHE for Trabecular Bone Segmentation

X-ray microtomography $(\mu \mathrm{CT})$ is a nondestructive and high-resolution method of microstructural analysis. The $\mu C T$ is compounded by a succession of 2D grayscale digital images, in this work a three stages methodology is proposed to enhance these images: image acquisition, image processing, and image post-processing. In the image processing stage an image integration, an anisotropic diffusion filter, the CLAHE algorithm, a morphological reconstruction, and a segmentation with the OTSU method are applied. SSIM is used to measure the structural similarity between the image after the CLAHE and the image next to the morphological reconstruction. This is used to determine a disk as a structural element with a radius of 1, with these values in the morphological reconstruction, SSIM greater than 0.95 is achieved. Another comparison is made between the original $\mu \mathrm{CT}$ image and the image after the morphological reconstruction to measure the similarity, obtaining an SSIM greater than 0.90. This index is used to determine the clip value for the CLAHE, the clip range is 0.085 - 0.1. After the application of the proposed methodology, the results are compared using the Jaccard index obtaining a value greater than 0.90. The segmented images can be used to obtain bone morphometry and know the bone quality.