Skip to main content Skip to secondary navigation
Main content start

Artificial Intelligence Methods Bridge the Gap between Radiology and Pathology To Improve the Radiology Image Interpretation - Mirabela Rusu, PhD

Event Details:

Friday, March 24, 2023
12:00pm - 1:00pm PDT

Location

Virtual

This event is open to:

General Public
Mirabela Rusu, PhD
Assistant Professor of Radiology
Stanford University

This event is open to all. JOIN HERE

Abstract: Artificial Intelligence methods, specifically Machine Learning, have been extensively used in medical imaging for a variety of tasks, e.g., segmenting organs, detecting and localizing diseases, and many more. There are many challenges associated with training such methods, associated with (1) availability of medical data, (2) availability of accurate labels, and (3) generalization to other cohorts beyond the training data. In this lecture, I will present my team’s efforts to address these challenges by bridging the gap between radiology and pathology images. We develop artificial intelligence methods focused on bringing information from pathology into radiology images, either as labels (through multi-modal registration) or radiology imaging signatures (through multi-modal correlation learning). Along with the data, the labels and imaging signatures are used to train models to detect cancer and distinguish aggressive from indolent cancers in different types of radiology images. Most of the talk will be focused on our studies in prostate cancer using MRI or b-mode ultrasound images, with some examples on renal and breast cancers.

Explore More Events