CORRECTION OF INTENSITY VARIATIONS IN MR-IMAGES FOR COMPUTER-AIDED TISSUE CLASSIFICATION

A number of supervised and unsupervised pattern recognition techniques have been proposed in recent years for the segmentation and the quant itative analysis of MR images. However, the efficacy of these techniqu es is affected by acquisition artifacts such as inter-slice, intra-sli ce, and inter-patient intensity variations. In this paper a new approa ch to the correction of intra-slice intensity variations is presented. Results demonstrate that the correction process enhances the performa nce of backpropagation neural network classifiers designed for the seg mentation of the images. Two slightly different versions of the method are presented. The first version fits an intensity correction surface directly to reference points selected by the user in the images. The second version fits the surface to reference points obtained by an int ermediate classification operation. Qualitative and quantitative evalu ation of both methods reveals that the first one leads to a better cor rection of the images than the second but that it is more sensitive to operator errors.