Massachusetts Institute of Technology MIT along with a team of researchers was working on an Artificial Intelligence on deep learning Platform. The platform is highly successful in detecting the risk of breast cancer.


From the mammogram records present in Massachusetts General Hospital (MGH), chances of developing breast cancer were more.

The platform was helpful in identifying women with a larger risk of evolving breast cancer. The study was over almost 90,000 screening mammograms at Massachusetts General Hospital (MGH).

Because of the learning model, 31% of patients were identified. Earlier the results were approximately 18% with the present Tyrer-Cuzick model. The research is indicative of a deep learning mammography-based model, providing more accurate results.

The mammographic density recovers the precision of models for breast cancer risk. However, the use of breast density is depending on variation across radiologists, subjective assessment, and restricted data. The team was working on the machine learning model rather than manually identifying image patterns.

With the help of a deep learning model on the basis of mammography, the team was able to provide better results and predicting risks.

The research teams were able to deliver a more precise risk forecast, and their hybrid DL model is similarly correct for both African American and white women — another improvement over the Tyrer-Cuzick model.


Women participating in the study had either a previous diagnosis of breast cancer. The growing risk factors present in the fusion deep learning model contain the weight, height, age, menopausal status. The report is also consisting of breast and ovarian cancer history present in the family.

The editorial in Radiology, by saying that assessing cancer risk with the help of mammograms deep learning will likely have larger risks.

To summarize the study, all AI systems are in the developing process to further improve the reconstructing process of medical images.