Microsoft researchers, in cooperation with the University of Washington and the Fred Hachnson Cancer Center, has developed a new model of artificial intelligence that promises a revolution in early detection of breast cancer
Microsoft today announced the results of a study published in the radiology magazine, which provides a new approach to examining breast cancer through artificial intelligence. The project, which was developed at Ai for Good Lab in partnership with the University of Washington and the Fred Hitchinson Center for Tema, offers an innovative model called FCDD.
FCDD is characterized by its ability to discover abnormal conditions by learning the level of natural mammary tests, rather than trying to determine different types of cancer. This methodology was particularly effective in the tests conducted with more than 9,700 MRI, with only 1.85 % of cases had cancer.
The results showed that the model was able to double the positive predictive value compared to traditional systems, reducing the wrong warnings by more than 25 %. FCDD also generates visually determining the suspicious site of the tumor, reaching a resolution of 92 % compared to the evaluation of expert radiologists.
Savana Bartridge, a professor of x -ray at Washington University and study author, notes that the model can be applied to both the abbreviated MRI tests and full diagnostic protocols, which may reduce exam times and interpretation.
Microsoft has publicly made the project symbol and methodology available through GitHub, allowing access to the scientific community. The model will undergo additional tests in broader and varied clinical groups to verify its effectiveness in various contexts.
PR/hn
Other articles with interest:
“Writer. Analyst. Avid travel maven. Devoted twitter guru. Unapologetic pop culture expert. General zombie enthusiast.”
