Researchers at MIT recently demonstrated a risk-prediction model for breast cancer patients. The model placed 31 percent women in high-risk category, accurately. Traditional risk-prediction model only identified 18% of such patients.
The AI program received training using the records of over 60,000 patients. This training resulted in advanced capabilities of the program to detect subtle nuances. These subtle differences will empower physicians to detect breast cancer as early as five years in advance.
Additionally, the new model based on Artificial Intelligence (AI) also promises better results for black women. Earlier, black women were at a much higher risk of dying from breast cancer. According to experts, the risk was 42% higher as compared to white women. The new model promises equally accurate results for various races. While this may not result in lower deaths due to unequal access to health care, it does promise better results overall.
AI Technology a Need of the Hour for Breast Cancer Patients
According to Regina Barzilay, MIT Professor the program offers immense hope for personalization of breast cancer diagnosis. This is an important field and involves extreme uncertainty at the moment. For example, the US Preventative Task Force advises starting screening at the age of 50, in every two years. On the other hand, the American Cancer Society recommends women to be more cautions after the age of 45.
The research was a collaboration between MIT and Massachusetts general hospital.
Moreover, most breast cancer patients are still taken back by surprise by detection. This is despite some major advances in modern imaging and valuable input on genetics. Late diagnosis results in higher costs, aggressive treatments, and more uncertainity for patients. Many fail to get treated despite access to the best of modern healthcare. Hence, this MIT research can prove to a bridge to a healthy life for many women worldwide.