Related Experiment Videos
Improving breast cancer diagnosis with computer-aided diagnosis
Y Jiang1, R M Nishikawa, R A Schmidt
1Kurt Rossmann Laboratories for Radiologic Image Research, Department of Radiology, University of Chicago, IL 60637, USA.
Academic Radiology
|January 19, 1999
Summary
Computer-aided diagnosis (CAD) significantly enhances radiologists' ability to detect breast cancer. This tool improves accuracy in identifying malignant microcalcifications, leading to better diagnostic performance.
Area of Science:
- Radiology
- Medical Imaging
- Oncology
Background:
- Accurate breast cancer diagnosis is crucial for effective treatment.
- Microcalcifications are common mammographic findings requiring careful interpretation.
- Radiologist performance can be influenced by various factors, including image complexity.
Purpose of the Study:
- To evaluate the impact of computer-aided diagnosis (CAD) on radiologist performance in breast cancer detection.
- To assess whether CAD systems can improve the accuracy of interpreting mammographic microcalcifications.
Main Methods:
- A computer classification scheme was developed using eight features extracted from mammograms to estimate malignancy likelihood for microcalcifications.
- 104 histologically verified cases (46 malignant, 58 benign) were analyzed.
- 10 radiologists' performance was compared with and without CAD assistance, using receiver operating characteristic (ROC) analysis and biopsy recommendations.
Main Results:
- The average area under the ROC curve (Az) improved from 0.61 without CAD to 0.75 with CAD (P < .0001).
- CAD use led to a significant increase in sensitivity (73.5% to 87.4%) and specificity (31.6% to 41.9%).
- CAD assistance resulted in more appropriate biopsy recommendations, increasing positive biopsy yield from 46% to 55%.
Conclusions:
- Computer-aided diagnosis (CAD) demonstrably improves radiologists' diagnostic performance in breast cancer detection.
- CAD systems enhance the accuracy of identifying malignant microcalcifications, leading to better patient outcomes.
- The integration of CAD technology is a valuable advancement in mammographic interpretation for breast cancer diagnosis.