Use of computerized interactive morphometry in the diagnosis of mammary adenoma and adenocarcinoma in dogs

We attempted to define quantitative objective criteria for diagnosis of mammary adenoma and adenocarcinoma in dogs. To that end, correlation between diagnosis made by conventional histologic examination and morphometric descriptors, obtained by computerized histologic analysis, was assessed in biopsy specimens of mammary tissue from 63 dogs. We used 2 interactive computer procedures: one assessed mean nuclear perimeter, mean nuclear area, and circularity factor; and the other evaluated nuclear stratification and cell crowding. A data base was generated with 11 specimens from normal mammary tissue, 17 specimens from mammary adenoma, 18 specimens from low-grade mammary adenocarcinoma, and 15 specimens from mammary adenocarcinoma. The mean values of nuclear perimeter, nuclear area, nuclei per millimeter of basement membrane, and minimal distances from cells to basement membrane gradually increased from normal to high-grade malignancy. Distributions of nuclear areas and of minimal distances from cells to basement membrane were shifted in specimens from malignant tumors. Multivariate analysis confirmed the homogeneity of the diagnostic groups.