THE COMBINATION OF A DECISION TREE TECHNIQUE WITH THE COMPUTER-ASSISTED MICROSCOPE ANALYSIS OF FEULGEN-STAINED NUCLEI TO ASSESS AGGRESSIVENESS IN LIPOMATOUS AND SMOOTH-MUSCLE TUMORS

The present study describes a computer-assisted methodology whose purp ose is to I educe the degree of subjectivity in the diagnosis of soft tissue tumors. This methodology associates three complementary techniq ues, namely digital cell image analysis, the discretisation of numeric al data and a Decision TI ee technique (DT). The first technique relie s on the use of the digital cell image analysis of Feulgen-stained nuc lei, a technique which makes possible a quantitative and thus objectiv e description of nuclei with the help of 24 numerical parameters (15 m olphonuclear and 9 DNA content- (ploidy level and proliferation activi ty) related). The second technique transforms each numerical parameter into an ordinal one with a small number of values (2 to 4) so that on ly the relevant physical significance of the parameters is retained Th e Decision Tree technique generates classification rules on the basis of the discretised parameters quoted above. This methodology was appli ed to 53 human soft tissue tumors which included 26 lipomatous tumors (13 malignant liposarcomas and 13 benign lipomas) and 27 smooth muscle tumors (11 malignant leiomyosarcomas and 16 benign leiomyomas). The r esults show that a distinction between benign (lipoma) and malignant ( liposarcoma) lipomatous tumors can easily be made by means of simple l ogical rules depending on only foul discretised cytological parameters (two ploidy- and two morphonuclear-related). In contrast, no stable o r predictive characterisation can be obtained with respect to the diff erence between leiomyosarcomas and the leiomyomas. Hence, while lipoma s and liposarcomas appeared to be two completely distinct biological e ntities, leiomyomas and leiomyosarcomas seem to involve a continuous b iological process.