Breast cancer detection and diagnosis is challenging for medical professionals. Different imaging modalities play a role, while existing technologies and procedures are continuously improved. This talk focuses on development of machine learning applications in breast imaging. Topics include computer aided detection in mammography, fully automated reading of mammograms and breast tomosynthesis, whole breast ultrasound, and diagnostic support in digital pathology.
The presentation will highlight changes the field as undergone over the past years from traditional machine learning to deep learning, both from a research and industry perspective. Reader study results will be presented which illustrate that deep learning based systems are rapidly approaching the level of human readers and that there is potential to increase the quality of image interpretation beyond the human expert level.
Where: Wranne, CMIV
When: 14:00, 24 oktober
Who: Nico Karssemeijer