ºÚÁÏÍø

Abstract

Background: The problem of Breast Cancer (BC) is one of the most important in modern oncology, due to its large spread in many countries of the world, where it took first place in the structure of malignancies in women. Aim: The aim of our work was to create an expert system for the diagnosis of breast cancer (separate morphological form) based on aggregate quantitative traits atypia cells. Materials & Methods: The object of the study was the nucleus and cells of glandular epithelium from punctured material of a ductal breast cancer for cytology. Light morphometry was performed by computer analyzer color images based on global microscope «Bresser» made in Germany, digital camera Leica DC 200 frames and grips (Leica, Germany) lens x 40. For morphometric study used a software tool AMS. It was used for measuring the area of cells, their nuclei and cytoplasm, calculating nuclear-cytoplasmic ratio. The measurement results are transferred to the metric system by aligning digital camera. 50 Nucleus were calculated in each of cytological preparations, and then a mean value of individual parameters was defined. The total sample size was approximately 5000 nuclei. Each kariometric parameter was tested by the expert system on cytologic specimens with breast cancer. We tested 70 cytological preparations with confirmed histologic diagnosis. These punctate swabs selected for testing were processed by computer image analysis. Then on the basis of cytometrix data for each case were calculated quantitative parameters, which values were compared with deciding criteria which are specific to breast cancer. The test results were determined informativeness of each parameter. Evaluation of informativeness was performed by calculating accuracy, sensitivity and specificity. Express-system, which included standard matrix (S-matrix) â?? the basis of diagnostic criteria for breast cancer, has been made to automate the process of diagnosing. Benchmarks and mathematical transformation (calculation of average values, plotting histograms, the distribution area of the nuclei and cytoplasm) were transformed into quantitative parameters that adequately describe the abnormal cell nuclei. The parameters which were obtained have been automatically introduced in the X-matrix constructed in the same format as the S-matrix. Then the computer program in real time comparing elements (quantitative parameters) obtained by X-matrix with elements S-matrix. The program is logged by coincidence x element of the corresponding s element. When the value of the X matrix fell in the range of S matrix elements coefficients corresponding cytometryx parameters were match. Diagnostic index was calculated as the sum of the coefficients cytometryx parameters that fell into the range of values for the S-matrix. In our cases, the clinical trials of the expert system, it was found that out of 14 cases of malignant disease of the breast diagnostic index was 70%.

Biography

Petro O Vavrukh has completed his PhD from the Danylo Halytsky Lviv National Medical University (Ukraine). He is the Associate Professor of Pathological Anatomy, Dissection Course and Forensic Medicine Departments of I. Ya. Horbachevsky Ternopil State Medical University (Ukraine). He is a Pathologist in Ternopil Regional Bureau of Pathologists, Ternopil (Ukraine). He has published more than 20 papers in reputed journals and Member of the European Society of Pathology and Association of Pathologists of Ukraine.