This thesis demonstrates that artificial intelligence (AI) can contribute to improved and more personalized breast cancer screening. AI supports radiologists by enabling more accurate and faster detection of breast cancer and helps identify clinically relevant cancers at an earlier stage. When AI and human readers work together, they complement each other, reducing the chance that cancers are missed. In addition, AI can support more targeted selection of women for supplemental MRI screening. This allows women at higher risk to receive additional imaging, while unnecessary MRI examinations and follow-up procedures can be reduced. The research also shows that some women develop breast cancer after discontinuation of MRI screening, highlighting the importance of ongoing attention to individual risk profiles and the need for further personalization of screening strategies. By combining AI with personalized risk-based approaches, breast cancer screening can become more accurate, accessible, and cost-effective, ultimately improving early detection and quality of care for women.
Suzanne van Winkel (Amsterdam, 1988) is a clinical epidemiologist. She obtained her Master’s degree in Evidence Based Practice (Amsterdam UMC, 2018), alongside her work as youth health care nurse (Municipality of Utrecht). She started a PhD on AI and personalized breast cancer screening (Radboudumc, 2020), followed by postdoctoral research on AI-supported clinical decision-making (Amsterdam UMC). She currently coordinates the Drugs Incident Monitor (Trimbos Institute).