Screening Mammography-Detected Cancers: The Sensitivity of the Computer-aided Detection System as Applied to Full-field Digital Mammography.
10.3348/jkrs.2006.54.4.327
- Author:
Sang Kyu YANG
1
;
Nariya CHO
;
Eun Sook KO
;
Do Yeon KIM
;
Woo Kyung MOON
Author Information
1. Department of Radiology, College of Medicine Seoul National University and The Institute of Radiation Medicine, Seoul National University Medical Research Center, Korea. moonwk@radcom.snu.ac.kr
- Publication Type:Original Article
- Keywords:
Breast neoplasms;
Computers;
Diagnostic Aid;
Digital radiography
- MeSH:
Breast;
Breast Neoplasms;
Carcinoma, Ductal;
Carcinoma, Intraductal, Noninfiltrating;
Carcinoma, Lobular;
Humans;
Mammography*;
Mass Screening*;
Prospective Studies;
Radiographic Image Enhancement
- From:Journal of the Korean Radiological Society
2006;54(4):327-332
- CountryRepublic of Korea
- Language:Korean
-
Abstract:
PURPOSE: We wanted to evaluate the sensitivity of the computer-aided detection (CAD) system for performing full-field digital mammography (FFDM) on the breast cancers that were originally detected by screening mammography. MATERIALS AND METHODS: The CAD system (Image Checker v3.1, R2 Technology, Los Altos, Calif.) together with a full-field digital mammography system (Senographe 2000D, GE Medical Systems, Buc, France) was prospectively applied to the mammograms of 70 mammographically detected breast cancer patients (age range, 37-69; median age, 51 years) who had negative findings on their clinical examinations. The sensitivity of the CAD system, according to histopathologic findings and radiologic primary features (i.e., mass, microcalcifications or mass with microcalcifications) and also the false-positive marking rate were then determined. RESULTS: The CAD system correctly depicted 67 of 70 breast cancer lesions (95.7%). The CAD system marked 29 of 30 breast cancers that presented with microcalcifications only (sensitivity 96.7%) and all 18 breast cancers that presented with mass together with microcalcifications (sensitivity 100%). Twenty of the 22 lesions that appeared as a mass only were marked correctly by the CAD system (sensitivity 90.9%). The CAD system correctly depicted all 22 lesions of ductal carcinoma in situ (sensitivity: 100%), all 13 lesions of invasive ductal carcinoma with ductal carcinoma in situ (sensitivity: 100%) and the 1 lesion of invasive lobular carcinoma (sensitivity: 100%). Thirty one of the 34 lesions of invasive ductal carcinoma were marked correctly by the CAD system (sensitivity: 91.8%). The rate of false-positive marks was 0.21 mass marks per image and 0.16 microcalcification marks per image. The overall rate of false-positive marks was 0.37 per image. CONCLUSION: The CAD system using FFDM is useful for the detection of asymptomatic breast cancers, and it has a high overall tumor detection rate. The false negative cases were found in relatively small invasive ductal carcinoma.
