Breast Cancer Research
The CAD-AI Research Laboratory is developing CAD (computer-aided diagnosis) tools for detection and characterization of breast cancer in mammography, digital breast tomosynthesis (DBT), 2D and 3D ultrasound, and breast magnetic resonance imaging (MR). Chan et al. developed the first CAD system for detection of microcalcifications on mammograms and conducted the first observer study to demonstrate that CAD can improve radiologists’ detection of subtle breast cancer manifested as microcalcifications. The CAD-AI Lab first applied convolutional neural network (CNN) to the detection of microcalcifications in 1993 and the detection of breast masses in 1994. In recent years, CNN has emerged as the main machine learning approach for CAD and artificial intelligence tools in medical imaging. The CAD-AI Lab continues to develop various machine learning and deep CNN methods for various CAD applications. Click the links below to learn more about our areas of interest for breast cancer research.
History of CAD-AI Breast Cancer Research
Convolutional Neural Network (CNN) architecture used in the 1990’s
CNN for computer-aided detection of microcalcifications
Chan H-P, LoSCB, Helvie MA, et al. Radiolofy 1993, 189(P): 318
Chan H-P, LoSCB, Sahiner B, et al. Medical Physics 1995; 22: 1555
CNN for computer-aided detection of masses
Chan H-P, Sahiner B, LoSCB, et al. Medical Physics 1994; 21: 875
Sahiner B, Chan H-P, Petrick N, et al. IEEE Trans Medical Imaging 1996; 15:598