Project Spotlight: Improving Breast Cancer Diagnosis with Interpretable Machine Learning

The team’s goal aims to increase the effectiveness of breast cancer screenings by using AI to enhance current processes for continually training radiologists using clinical data. To motivate one possible avenue to improve care, the team notes the well-documented and significant inconsistencies in how various doctors evaluate mammogram screenings. Recall rates (the fraction of patients asked to return for more targeted diagnostic testing) vary widely across doctors. The team believes that deep learning can enhance the selection process for review in the paraclinical setting, helping to increase cancer detection rates and reduce false positives, leading to a more effectively treated patient population.

Please share a little about your background and your research experiences.

Adam Perer, PhD, is an Assistant Research Professor in the Human-Computer Interaction Institute at Carnegie Mellon University (CMU). His work on human-centered data science, data visualization, and interpretable machine learning is often inspired by challenges in healthcare, which is a domain that is becoming increasingly transformed by data and algorithmic advances. The PHDA provides a unique opportunity to connect the computer scientists at CMU with access to domain experts and data in the medical research community.

Walk us through your project.

Our goal is to increase the effectiveness of breast cancer screenings by using AI to enhance current processes for continually training radiologists using clinical data. By using thousands of mammogram screenings and modeling not only ground truth labels, but also the tendencies of each radiologist, our AI solution will identify ideal training cases for review. The solution will utilize state-of-the-art computer vision techniques together with human-centered visualizations and design principles, to make the results of the AI interpretable to medical experts with limited AI expertise.

What led you to the PHDA?

I learned about the PHDA during my faculty interviews at CMU in 2018. As I was presenting my healthcare-related research, I was fortunate enough to hear about the exciting medical research possibilities now made easier at CMU thanks to the alliance between Pitt, UPMC, and CMU.

What are your project’s next steps?

I am most excited about studying how radiologists and AI can collaborate to improve patient outcomes. There is a lot of speculation on how AI will impact radiology, but our research will help us understand the complementary strengths of human and machine perception in radiology, and of how these strengths might be effectively combined in practice.

Subscribe to our newsletter to get the latest Alliance news, event invitations and updates, and behind-the-scenes spotlights on our researchers, projects, and team members.
Subscribe

By clicking Submit, I agree to the Pittsburgh Health Data Alliance's Terms of Use and Privacy Policy.