“African American women continue to have breast cancer at a younger age and often in later stages,” Salewai Oseni, a breast surgeon at Massachusetts General Hospital, said in a recent press release. “This, coupled with the higher incidence of triple negative breast cancer in this group, has led to an increased mortality from breast cancer.”
For the past two years, researchers from MIT CSAIL and the Abdul Latif Jameel Clinic for Machine Learning in Health have been working to develop a new deep-learning system that can predict a patient’s cancer risk using only the person’s mammograms, which are said to be equally work effectively, regardless of race or ethnicity.
This algorithm, called ‘Mirai’ (not to be confused with Toyota’s Fuel Cell EV), is reportedly capable of ‘modeling a patient’s risk over different future times’, while keeping small deviations in mind. taken, according to the mammogram machine of the clinic. used, according to a Wednesday release from MIT. Its predictions can be further optimized if other clinical risk factors – such as age or family history – are available.
The CSAIL team initially trained Mirai in a data set of 200,000 Massachusetts General Hospital (MGH) exams before confirming the predictive results on additional sets from the Karolinska Institute in Sweden and the Chang Gung Memorial Hospital in Taiwan. So far, the results are very encouraging, with results indicating that according to the release, Mirai is ‘significantly more accurate’, predicting the risks of patients in all three data groups and being able to predict almost twice as many potential cancer cases among high-risk groups. identify as the diagnostic Tyrer-Cuzick model currently used throughout the course of the study.
To ensure that Mirai’s recommendations are consistent, the CSAIL team biased the algorithm by running it through a contrast network to distinguish between aspects of the mammogram that are important and those caused by minor random environmental deviations (such as the make / model of the mammogram machine.).
“Improved risk models for breast cancer enable targeted screening strategies to facilitate earlier detection, and less screening damage than existing guidelines,” said Adam Yala, lead author of CSAIL of the forthcoming Science Translational Medicine study, said in a statement. “Our goal is to make this progress part of the standard of care.”
MIT
It can advance the state of oncological science. Modern mammograms still suffer from reliability issues, even now 60 years after the technology was widely used. Experts still disagree on how often women should be screened, and some argue in favor of more aggressive strategies to catch cancer growth as early as possible, while others argue for longer gaps between routine tests around the rate of false positive aspects to reduce (as well as medical costs for patients down). Mirai will be used to help physicians determine which patients will benefit most (and most fairly) from complementary imaging and MRIs, based on both the mammogram and other factors such as the person’s age, genetics, family history, and breast cancer. .
“We know that MRI can detect cancer earlier than mammography, and that earlier detection improves patient outcomes,” Yala explained. ‘But for patients at low risk for cancer, the risk of false positive benefits may outweigh. With improved risk models, we can design more nuanced risk screening guidelines that provide more sensitive screening, such as MRI, to patients who will develop cancer, to get better results, while reducing unnecessary screening and overtreatment for the rest. ”
Mirai also considers risk factors that do not necessarily appear in the mammogram, such as the patient’s age, hormone levels, and menopausal status. These factors were incorporated during the training phase, enabling the model to predict them based on the given mammogram image, even if the clinician did not provide the information manually.
In the future, Mirai may find use in other medical applications for the benefit of the community. Although the system may not be able to interpret the history of a patient’s existing imaging results and integrate it into the assessment, it may build on any additional X-rays / MRIs provided to him. The team is also considering integrating tomosynthesis techniques to further increase Mirai’s statistical aptitude. The CSAIL team also collaborated with researchers at Emory University to further validate the model.