The aim of the present paper is to find analytical parameters to distinguish severe dysplasia from cancer. A population of 150 nuclei was selected randomly from: 5 cases of large bowel carcinoma, 5 cases of adenomatous polyps with severe dysplasia--according to Kozuka's criteria--(without any reference to the architecture: tubular, tubulo-villous and villous) and 5 cases of normal mucosa as control. The files of the coordinates of the nuclear contours were submitted to the S.A.M. (Shape Analytical Morphometry) work-station, which produces quantitative parameters (area, perimeter, maximum diameter and Roundness Factor) as well as analytical shape-descriptors. The latter were extracted by the S.A.E. (Shape Asymmetry Evaluator) procedure, which evaluates the symmetry of each contour studied, and by Fourier harmonic analysis which breaks up each curve into a defined set of harmonics which are then used to perform statistical tests. In the comparison between normal and severe dysplastic nuclei, normal and carcinomatous nuclei and severe dysplastic and carcinomatous nuclei univariate statistical evaluation (Student t test) proved to be highly significant for the most of the analytical parameters.

Download full-text PDF

Source

Publication Analysis

Top Keywords

analytical morphometry
8
distinguish severe
8
severe dysplasia
8
large bowel
8
bowel carcinoma
8
analytical parameters
8
severe dysplastic
8
carcinomatous nuclei
8
analytical
5
severe
5

Similar Publications

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!