Adaptive Computer-Assisted Mammography Training for Improved Breast Cancer Screening.

摘要

In this project, we propose to research the methodology for constructing adaptive computer-aided education systems for mammography. Improved mammography education could lead to improved benefit of mammography to breast cancer care and in turn to decreased mortality from the disease. The project includes: Observer studies to collect reading data from radiology trainees; Extraction of image features (human- and computer- based); Statistical modeling of the reader data to discover patterns in their error making; Development of methodology for adaptive training that utilizes the constructed models. The proposed adaptive system could improve education in mammography. This may in turn result in improved benefit of mammography in breast cancer detection and lower mortality associated the disease.