Computer-aided detection (CAD) in mammography: Does it help the junior or the senior radiologist?
Balleyguier C, Kinkel K, Fermanian J, Malan S, Djen G, Taourel P, Helenon O.
OBJECTIVES:: To evaluate the impact of a computer-aided detection (CAD) system on the ability of a junior and senior radiologist to detect breast cancers on mammograms, and to determine the potential of CAD as a teaching tool in mammography.
METHODS:: Hundred biopsy-proven cancers and 100 normal mammograms were randomly analyzed by a CAD system. The sensitivity (Se) and specificity (Sp) of the CAD system were calculated. In the second phase, to simulate daily practice, 110 mammograms (97 normal or with benign lesions, and 13 cancers) were examined independently by a junior and a senior radiologist, with and without CAD. Interpretations were standardized according to BI-RADS classification. Sensitivity, Specificity, positive and negative predictive values (PPV, NPV) were calculated for each session.
RESULTS:: For the senior radiologist, Se slightly improved from 76.9 to 84.6% after CAD analysis (NS) (one case of clustered microcalcifications case overlooked by the senior radiologist was detected by CAD). Sp, PPV and PNV did not change significantly. For the junior radiologist, Se improved from 61.9 to 84.6% (significant change). Three cancers overlooked by the junior radiologist were detected by CAD. Sp was unchanged.
OBJECTIVES:: To evaluate the impact of a computer-aided detection (CAD) system on the ability of a junior and senior radiologist to detect breast cancers on mammograms, and to determine the potential of CAD as a teaching tool in mammography.
METHODS:: Hundred biopsy-proven cancers and 100 normal mammograms were randomly analyzed by a CAD system. The sensitivity (Se) and specificity (Sp) of the CAD system were calculated. In the second phase, to simulate daily practice, 110 mammograms (97 normal or with benign lesions, and 13 cancers) were examined independently by a junior and a senior radiologist, with and without CAD. Interpretations were standardized according to BI-RADS classification. Sensitivity, Specificity, positive and negative predictive values (PPV, NPV) were calculated for each session.
RESULTS:: For the senior radiologist, Se slightly improved from 76.9 to 84.6% after CAD analysis (NS) (one case of clustered microcalcifications case overlooked by the senior radiologist was detected by CAD). Sp, PPV and PNV did not change significantly. For the junior radiologist, Se improved from 61.9 to 84.6% (significant change). Three cancers overlooked by the junior radiologist were detected by CAD. Sp was unchanged.
CONCLUSION:: CAD mammography proved more useful for the junior than for the senior radiologist, improving sensitivity. The CAD system may represent a useful educational tool for mammography.
Eur J Radiol. 2005 Apr;54(1):90-6.
Eur J Radiol. 2005 Apr;54(1):90-6.
Computer-aided detection (CAD) in mammography: Does it help the junior or the senior radiologist?
![]()
Reviewed by Sumer Sethi
on
Wednesday, April 06, 2005
Rating:
Sumer Sethi
Unique blend of academic excellence and entrepreneurship, heading leading firms in India- Teleradiology Providers, pioneering company providing teleradiology services and DAMS (Delhi Academy of Medical Sciences) Premier test preparation institute in India for MD/MS/MCI preparation. He has also been an invited faculty member at various conferences, including Teleradiology in IRIA 2008 and 2011, Hospital Build Middle East, Congress of the Brain Tumor Radiology in Neuro-oncology Society. Dr. Sethi is Editor-in-Chief of Internet Journal of Radiology. He has a keen interest in Web 2.0 technologies and in maintaining his famous radiology blog, which has been featured in multiple international journals.
No comments:
Post a Comment