Objective The purpose of this study was to evaluate the clinical value of 3D-CPA in differentiating epithelial ovarian tumors by studying their vascular pathology.Methods 67 epithelial ovarian tumors(40 ovarian carcinomas, 27 benign epithelial ovarian tumors) were examined preoperatively with 3D-CPA. The vascular pattern and vascularity index of tumor were assessed respectively. VEGF protein and microvessels were stained immunohistochemically by using antiVEGF and anti-CD_(34) antibody respectively after operation. Both the microvessel density(MVD) and the intensity of VEGF staining were analyzed.Results The MVD in ovarian carcinomas(38.8±5.9/400×magnified field) was significantly higher than that in benign epithelial ovarian tumors(15.5±5.0/400×magnified field) (P<0.05). The intensity of VEGF staining in ovarian carcinomas was significantly stronger than that in benign ovarian tumors (P<0.05). The vascular pattern of 3D-CPA Ⅲ was shown in 33 (82.5%) ovarian carcinomas. None of 27 benign epithelial ovarian tumors had the vascular pattern of 3DCPAⅢ. Using the vascular pattern of 3D-CPAⅢ, the sensitivity in predicting ovarian carcinoma was 82.5% and specificity was 100%. When 0.02 counts/cm~3 was considered as the cutoff of vascularity index, the sensitivity and specificity of vaseularity index in detecting ovarian carcinoma were 77.5% and 81.5%, respectively. The vascularity index was correlated with MVD in ovarian carcinomas.Conclusion 3D-CPA may be used to evaluate the angiogenesis in ovarian carcinoma before operation and could provide valuable information for the differential diagnosis of epithelial ovarian tumors. Objective The purpose of this study was to evaluate the clinical value of 3D-CPA in differentiating epithelial ovarian tumors by studying their vascular pathology. Methods 67 prosial ovarian tumors (40 ovarian carcinomas, 27 benign epithelial ovarian tumors) were examined preoperatively with 3D-CPA . The vascular pattern and vascularity index of tumor were assessed respectively. VEGF protein and microvessels were stained immunohistochemically by using antiVEGF and anti-CD (34) antibody respectively after operation. Both the microvessel density (MVD) and the intensity of VEGF . Results The MVD in ovarian carcinomas (38.8 ± 5.9 / 400 × magnified field) was significantly higher than that in benign epithelial ovarian tumors (15.5 ± 5.0 / 400 × magnified field) (P <0.05). The intensity of VEGF staining in ovarian carcinomas was significantly stronger than that in benign ovarian tumors (P <0.05). The vascular pattern of 3D-CPA III was shown in 33 (82.5%) ovarian carcinomas. None of 2 The benign epithelial ovarian tumors had the vascular pattern of 3DCPA III. Using the vascular pattern of 3D-CPA III, the sensitivity in predicting ovarian carcinoma was 82.5% and specificity was 100%. When 0.02 counts / cm ~ 3 was considered as the cutoff of vascularity index, the sensitivity and specificity of vaseularity index in detecting ovarian carcinoma were 77.5% and 81.5%, respectively. The vascularity index was correlated with MVD in ovarian carcinomas. Confusion 3D-CPA may be used to evaluate the angiogenesis in ovarian carcinoma before operation and could provide valuable information for the differential diagnosis of epithelial ovarian tumors.