Objective To detect the features of breast carcinoma with nipple discharge using selective galactography and evaluate the value of this method. Methods The galactography findings of 39 cases of pathology proved breast carcinoma with nipple discharge were analyzed retrospectively. Results Among the 39 cases, the pathology diagnosis included 7 cases of intraductal papillary carcinoma, 13 cases of ductal carcinoma in situ, and 19 cases of invasive ductal carcinoma. The galactographic features included: duct filling defect (24 cases, 61.5%), ductal obstruction (15 cases, 38.5%), rigidity and disorder (31 cases, 79.5%), wormy erosion sign (18 cases, 46.2%), pond lake sign (4 cases, 10.3%), ductal interruption (17 cases, 43.6%), ductal stenosis (24 casesm, 61.5%), ductal dilatation (22 cases, 56.4%). Among all of the cases, the accuracy rate of mammography and galactography was 64.1% (25/39) and 87.2% (34/39), respectively (P＜0.05). Among the 13 cases of ductal carcinoma in situ, the accuracy rate of mammography and galactography was 61.5% (8/13) and 100% (13/13), respectively (P＜0.05). Conclusion Selective galactography could improve the detection rate of breast carcinoma, and has great value in the early diagnosis of ductal carcinoma in situ.
This study aims to explore the clinical value of the computer-aided diagnosis (CAD) system for early detection of the pulmonary nodules on digital chest X-ray. A total of 100 cases of digital chest radiographs with pulmonary nodules of 5-20 mm diameter were selected from Pictures Archiving and Communication System (PACS) database in West China Hospital of Sichuan University were enrolled into trial group, and other 200 chest radiographs without pulmonary nodules as control group. All cases were confirmed by CT examination. Firstly, these cases were diagnosed by 5 different-seniority doctors without CAD, and after three months, these cases were re-diagnosed by the 5 doctors with CAD. Subsequently, the diagnostic results were analyzed by using SPSS statistical methods. The results showed that the sensitivity and specificity for detecting pulmonary nodules tended to be improved by using the CAD system, especially for specificity, but there was no significant difference before and after using CAD system.