/DCE_MRI_Preproc

This is the Matlab code that pre-process a series of DCE MRI images of the breast. It first does image registration and then image classification using the Kinetic information of the curves.

Primary LanguageMATLAB

DCE-MRI Image preprocessing for breast lesion segmentation

This code contains a set of Matlab scripts to perform the preprocessing of DCE-MRI images of the breast. The preprocessing has two steps, a registration step using affine intensity based transformations; and an image enhancement step using machine learning.

Requirements

This software uses the NIfTIi and ANALYZE image toolbox available at: http://www.mathworks.com/matlabcentral/fileexchange/8797-tools-for-nifti-and-analyze-image

It also assumes that you have access to a group of DCE-MRI images stored somewhere in separate folders. These are the two paths that are defined in most of the matlab files: one for the nifti library, one for the DCE-MRI image folders.

Description

DCE-MRI images are registered using an intensity based rigid transformation algorithm based on gradient descent or a genetic algorithm. After the registration, voxels that correspond to breast lesions are enhanced using the Naive Bayes machine learning classifier or a Regression tree. The classifier is trained to identify four different classes in breast images: lesion, normal tissue, chest, and background. Training is performed by manually selecting 150 voxels for each of the four classes from images where breast lesions have been confirmed by an expert in the field. Thirteen attributes obtained from the kinetic curves of the selected voxels are later used to train the classifier. Finally, the classifier is used to increase the intensity values of voxels labeled as lesions and decrease the intensities of all the other voxels. The post-processed images are used for
volume segmentation of breast lesions using a level set method based on the Active Contours Without Edges algorithm.