MORPHOMETRIC ANALYSIS OF WHITE-MATTER LESIONS IN MR-IMAGES - METHOD AND VALIDATION

The analysis of MR images is evolving from qualitative to quantitative . More and more, the question asked by clinicians is how much and wher e, rather than a simple statement on the presence or absence of abnorm alities. This paper presents a study in which the results obtained wit h a semiautomatic, multispectral segmentation technique are quantitati vely compared to manually delineated regions. The core of the semiauto matic image analysis system is a supervised artificial neural network classifier augmented with dedicated pre- and postprocessing algorithms , including anisotropic noise filtering and a surface-fitting method f or the correction of spatial intensity variations. The study was focus ed on the quantitation of white matter lesions in the human brain. A t otal of 36 images from sis brain volumes was analyzed twice by each of two operators, under supervision of a neuroradiologist. Both the intr a- and interrater variability of the methods were studied in terms of the average tissue area detected per slice, the correlation coefficien ts between area measurements, and a measure of similarity derived from the kappa statistic. The results indicate that, compared to a manual method, the use of the semiautomatic technique not only facilitates th e analysis of the images, but also has similar or lower intra- and int errater variabilities.