Robust Artificial Intelligence Tools May Predict Future Breast Cancer
Researchers at Massachusetts Institute of Technology’s (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) and Massachusetts General Hospital (MGH) created a new deep-learning artificial intelligence model that can better predict if a patient is likely to develop breast cancer from a mammogram, as much as five years in the future. This AI has been named “Mirai”.
This AI program is specifically trained on mammograms and their known outcomes using data from over 60,000 patients. Using the data, the software is able to discern even subtle patterns in breast tissue that are potential precursors to malignancy.
Mirai is able to model a patient’s risk of breast cancer through the use of multiple sets of information, and create an accurate predictive model across multiple future time points by using clinical risk factors such as age or family history.
The algorithm used is very agile and adaptive and can produce consistent predictions even with variations in clinical environments and quality of mammogram.
So far, Mirai has shown to be significantly more accurate than prior algorithms at detecting breast cancer risk. It has shown to be consistent and accurate in its predictions across patients of different races, age groups, and breast density categories as well as cancer subtypes.
Source link:https://news.mit.edu/2021/robust-artificial-intelligence-tools-predict-future-cancer-0128
Know the cancer – Breast cancer treatment
Previous :Postmenopausal breast cancer survivors should focus on osteoporosis prevention