Abstract:

Objective To set up CT imaging diagnostic comprehensive model for the diagnosis of malignant peritoneal mesothelioma, peritoneal metastatic carcinoma, and tuberculous peritonitis by Bayes discriminant analysis and to explore its value in the differential diagnosis of peritoneal diseases. Methods CT images of 147 patients with peritoneal thickening were retrospectively analyzed. Images of lesions of pleural and peritoneal were observed, Chi-square test between groups were used to screen out the variables that could be grouped, 117 cases(79.6%) were grouped experimental group, 30 cases(20.4%) in test group, and Bayes discriminant analysis was used to establish a comprehensive diagnosis model. Results By comparison of CT imaging manifestations of three peritoneal diseases between groups, peritoneum thickening, greater omentum thickening, intestinal wall fixation, organ involvement(organ infiltration or metastasis), and pleural plaque were regarded as the grouped variables finally. CT imaging diagnostic models of three peritoneal diseases were obtained by Bayes discriminant analysis. In the diagnosis of the comprehensive diagnosis model, there was the accuracy rate(89.7%) and the misjudgment rate(10.3%) in experimental group, and the accuracy rate (86.7%) and the misjudgment rate(13.3%) in test group; pleural plaques(20.4) contributed the most to the diagnosis of malignant peritoneal mesothelioma. Conclusion CT imaging diagnostic model has certain clinical value in the differential diagnosis of peritoneal diseases.

Key words: peritoneal diseases, ascites, mesothelioma, peritonitis, tuberculous, peritoneal metastatic carcinoma