Image-based risk assessment to identify women at high-risk for breast cancer

7. PROJECT SUMMARY Breast cancer is the most common cancer worldwide and the most common cancer diagnosed in American women. While there has been good progress regarding detection and treatment methods, breast cancer remains the primary cause of death from malignant tumors. Hence, there is a critical need for the development of novel

predictive and prognostic factors. Risk assessments are currently performed by medical professionals to identify women that could benefit from enhanced breast surveillance or risk reduction methods. Unfortunately, most diagnosed cases do not have an identifiable risk factor, making it a challenge to identify high risk women prior to

onset using classical risk assessments. This medical difficulty has resulted in the development of several artificial intelligence and machine learning approaches being applied to screening mammograms to identify breast cancer earlier. However, these approaches search for abnormalities that indicate an existing cancer and have been

found to not be generalizable to the entire screening population. It is becoming more common for younger women to be diagnosed with breast cancer, and the cancers tend to be more aggressive. This Phase I proposes to create a risk assessment product for mammography that is not based on machine learning but rather a novel

measurement of risky dense tissue. Alteration in the architecture and composition of microenvironment is a well- recognized component of breast pathologies and some changes may occur prior to tumor onset. WAVED Medical’s measurement is sensitive to these alternations in identifying areas of dense tissue that is tumor prone.

This feasibility study seeks to demonstrate that the novel measurement of risky dense breast tissue has the potential to be implemented into classical risk models. Phase I specific aims are to 1) improve efficiency in identifying risky dense tissue on mammograms by creating a secure database that contains preprocessed data

for optimized analysis, and 2) establish risky dense tissue as a better predictor of breast cancer than traditional mammographic percent density (MPD), by showing risky dense tissue is more accurate in predicting breast cancer than MPD. Follow-on Phase II efforts will include developing a platform and integrating WAVED into

hospital infrastructure for evaluating mammograms. These improvements will create a risk assessment product that increases the accuracy of medical professionals at identifying high-risk patients and ensures patients are receiving additional medical care, such as supplemental screening or risk reduction methods, to prevent invasive

cancer. Successful completion of the project has potential to advance state-of-the-art breast cancer assessments to provide quantification of risky dense tissue to identify high-risk patients needing preventive care.