> Our Work > The Work We Fund

2024 The John & Daria Barry Foundation – PCF Young Investigator Award

Artificial Intelligence to Augment MRI-Based Identification of Extraprostatic Extension and Predictive Models for Prostate Cancer

Hiten Patel, MD, MPH
Northwestern University

Mentors: Ashley Ross, MD, PhD; Ulas Bagci, PhD; Adam Murphy, MD, MBA

Description:

  • Prostate MRI has improved detection and management of prostate cancer over the past decade. However, radiologists often fail to identify the high-risk feature of extraprostatic extension on MRI, with 57% of cases found on surgical pathology not identified by MRI. There remains a need to quantify the variation in radiologist interpretation of extraprostatic extension on MRIand correlation with surgical pathology and to develop improved methods for interpreting such high-risk features on MRI. 
  • Dr. Patel’s project will quantify radiologist variation and develop artificial intelligence (AI)-driven applications for localized prostate cancer staging through machine learning using prostate MRI.
  • The variation will be quantified between radiologists and clinical practices in identifying extraprostatic extension on prostate MRI and the association with extraprostatic extension on surgical pathology.
  • AI-driven applications will be developed for localized prostate cancer staging through machine learning of prostate MRIs and the performance will be compared to traditional tools and radiologist interpretation.
  • If successful, this project will help identify radiologist MRI interpretation practices that can be improved and develop AI tools for improved MRI interpretations that will aid radiologists, inform surgical and radiation planning, and impact downstream oncologic and functional outcomes.

What this means to patients: Extraprostatic extension is a high-risk feature in prostate cancer that is difficult to detect by radiologists on MRI. Better methods to identify extraprostatic extension will improve patient management, treatment, and outcomes. Dr. Patel’s project will identify interventions to improve accuracy in MRI interpretation by radiologists and develop AI-driven prediction models that will outperform traditional methods of estimating the risk of extraprostatic extension. These models will be studied in randomized trials to assess impact on treatment planning and outcomes.