Simple exmaples of Scikit-image for medical images.
Language: Python
File name is histogram2d.py. Original scikit-image example is here. The figure on the left is CT value histogram of a non-contrast chest sagittal CT image. The lower and upper limits are set to -200 and 200. The figure on the right is a histogram of the rescaled intensity.
File name is hessian2d.py. Usage of hessian_matrix function is here. The function was applied to non-contrast axial, sagittal, and coronal CT images. The parameter sigma was set to 1.0 as follows.
H_elems = hessian_matrix(input_img, sigma=1.0, order='rc')
File name is skeleton2d.py. This example consists of five steps. (1)Thresholding of CT value for coarse bone segmentation from a non-contrast CT image. (2)Applying a morphological closing for filling the holes in the bone regions, using binary_closing. (3)Applying a morphological opening for remove small regions, using binary_opening. (4)Convex hull to the bone regions, using convex_hull_image. (5)Skeletonizing the bone regions, using skeletonize. Finally, because of the difficulty of observing skeletonized thin lines, morphological dilation was applied.
File name is ridge2d.py. Original scikit-image example is here. Four ridge filters were applied to non-contrast chest CT image.
Hidenobu SUZUKI
"AppSciImg" is under MIT license.