R. Ricco

Differential diagnosis between thyroid follicular adenoma and carcinoma. Analytic morphometric approach.

In this study some nuclear dimensional and analytical parameters were evaluated in order to distinguish follicular atypical adenoma from follicular carcinoma of the thyroid. Eighty nuclei from carcinomas, 80 from adenomas and 80 from normal thyroid were studied. Analytical parameters obtained by the nuclear shape study (by S.A.M. system) as well as dimensional parameters were submitted to univariate statistical analysis. On the ground of our results atypical adenoma could be considered as an intermediate aspect of a progressive change from benign to malignant even if they are closer to normal thyroid than to carcinoma.

Nuclear Morphometry in Squamous Cell Carcinoma (SCC) of the Tongue

In the histological grading of oral squamous cell carcinoma (OSCC) nuclear features are very important. Nevertheless evaluation is usually performed in a subjective and not highly reproducible way. The aim of this work was to investigate the relationship between nuclear shape and survival in 30 cases of carcinoma of the tongue. All the patients were divided into two groups: short-term survival and long-term survival. Twenty nuclei for each tumour were submitted to a morphometrical study by the shape analytical morphometry (SAM) software system. It was thus possible to evaluate not only nuclear dimensions but also nuclear contour irregularities and nuclear shape asymmetries. Multivariate discriminant analysis (MDA) of the quantitative parameters obtained by the morphometrical study distinguished the patients of the two groups with only a 10% error; moreover successful cluster analysis was performed by using Fourier parameters. Both these sets of results were achieved mainly owing to the parameters for contour irregularities.

Non-Hodgkin's lymphoma diagnosis aided by the S.A.M. system.

In the attempt to discriminate between centroblasts and immunoblasts, an analytical morphometrical procedure was adopted, considering the nuclei of a randomly selected centroblastic and immunoblastic population from non-Hodgkins lymphomas and the nucleoli of a more restricted number of the same neoplastic population. All the size-independent shape-descriptor parameters extracted for each step of the mathematical analyses used were submitted to Hotellings multivariate discriminant analysis that enabled us to achieve a good distinction (3% error) between centroblasts and immunoblasts when nuclear and nucleolar parameters were used together.

Morphometrical investigation of medulloblastoma nuclei by S.A.M. (Shape Analytical Morphometry) software system

In order to characterize medulloblastomas and to get over the difficulties sometimes encountered in differential diagnosis, a double morphometric procedure has been applied to its nuclei. The first consisted of size measurements (maximum diameter, area and perimeter), the latter is represented by S.A.M. (Shape Analytical Morphometry) software-system specifically implemented to describe shape of biological structure by analytical parameters. Analytical and dimensional parameters submitted to Hotellings multivariate discriminant analysis gave the best results when used together in convenient discriminant subsets, thereby allowing a good distinction between medulloblastoma in comparison with neuroblastoma, Ewings tumor, lymphoblastic and lymphocytic lymphoma. These results underline the usefulness of morphometric characterization also for practical diagnostic purposes.

Hand-Mirror (HM) Cells in Acute Lymphoblastic Leukaemia (A.L.L.): An Attempt at Cellular Classification by Morphometric Analytical Parameters Using the S.A.M. (Shape Analytical Morphometry) System

A morphometric study of the nuclear and cytoplasmic shape of a blastic population in a case of HM-A.L.L. was performed by comparing the two differently shaped populations with and without HM configuration. The results obtained using analytical size-independent parameters created by the S.A.M. work-station enable us to characterize the shape of both blastic cell populations quantitatively, and strongly suggest the existence of shape modulation from one cellular type toward the other. Thus a possible sequence from blastic cells (having regular, rather round nuclei) to HM cells (characterized by high distortion of both nucleus and cytoplasm) was proposed.

A new approach to the histologic study of ovarian tumors by analytical morphometry

In recent years, borderline epithelial tumors of the ovary have been investigated by morphometric techniques to allow for a differential diagnosis from benign and malignant neoplasms. In order to enhance this discriminant power, we have applied a new analytical procedure to the evaluation of the nuclear shape in epithelial ovarian tumors. Sixty nuclei of benign ovarian serous neoplasms, 60 nuclei of serous borderline tumors and 60 nuclei of serous carcinomas (18 cases in all) were examined using the software system SAM (Shape Analytical Morphometry). The morphometric procedure consisted of three different phases: (i) extraction of nuclear fundamental curve: this is a function curve giving the smoothing of the original contour by two parametric equations (separately for x and y values as dependent variables); (ii) evaluation of nuclear contour irregularities by Fourier analysis; (iii) evaluation of nuclear shape asymmetry by SAE (Shape Asymmetry Evaluator): this is the ratio between the length of a segment of a parabola interpolating the original curve points and a straight line joining its extremities for a 180° barycentric rotation according 10° steps. All the obtained independent parameters were submitted to statistical analysis. Nuclei of borderline tumors showed dimensional parameters which were intermediate between those of benign and malignant tumors. Both the asymmetry and the irregularities of nuclear contour were greatest in carcinomas.