Segmentation and Quantification of Breast Arterial Calcifications - Xiaoyuan Guo

Explore a comprehensive lecture on the segmentation and quantification of breast arterial calcifications (BAC) in mammograms. Delve into the development of a lightweight fine vessel segmentation method called Simple Context U-Net (SCU-Net) for accurate BAC detection. Learn about five quantitative metrics proposed to measure BAC progression, including Sum of Mask Probability, Sum of Mask Area, and Sum of Mask Intensity. Discover how these metrics perform in longitudinal studies and compare to breast CT measurements. Gain insights into the potential of BAC measurements for personalized cardiovascular disease risk assessment in women.

Introduction
What is a mammogram
Challenges
Goals
Annotations
Segmentation
Patchwise Segmentation
Results
Semantic Segmentation
Dilated Convolution
Comparison
Patchwork Results
Quantification
Accuracy
BC Quantification
Evaluation
Conclusion
Question for Sophie