| 摘要 |
Disclosed is a computer-implemented method that allows automatic discrimination between two tissues of interest: extrinsic tissue and intrinsic tissue, from a plurality of images, allowing a quantitative valuation of each of said tissues to be obtained without requiring the intervention of an expert. Said method involves: defining, using a parametric contour, a discrimination area in images obtained from a medical image acquisition device; and subsequently discriminating and quantifying the tissues on the basis of the photometric features of the different tissues observed in the images, evaluating the local area of each voxel belonging to the discrimination area previously defined within the plurality of images. The disclosed method considerably increases precision in the discrimination and quantification of tissues, while the percentage error shown is considered tolerable for diagnostic purposes. |
