(she/her/hers)
UC Berkeley
medical imaging, MRI, machine learning, biomedical engineering
Efrat Shimron is a postdoctoral fellow at the Department of Electrical Engineering and Computer Sciences (EECS) at UC Berkeley, working with Prof. Michael (Miki) Lustig. Her research interests are in developing machine learning techniques for rapid and robust medical imaging, focusing on Magnetic Resonance Imaging (MRI). Previously, she obtained a PhD in biomedical engineering from the Technion – Israel Institute of Technology. Efrat's research work received over 20 international excellence awards. Most recently, she was named as an EECS Rising Star, received an Outstanding Emerging Investigator award, and received the 2022-2024 Weizmann Institute Career in Science Award. Her work on identifying “data crimes” in medical AI has been published in the Proceedings of the National Academy of Sciences (PNAS) journal and received wide media coverage.
AI-powered Computational MRI for Next-generation Healthcare
Although Machine learning (ML) algorithms have recently made a huge impact on medical imaging, their development for clinical applications must be conducted carefully. My research interests are in the development of ML algorithms for medical imaging. In this seminar I will describe my work in this field, which reveals both potential pitfalls and enormous benefits offered by ML techniques. In the first part I will describe our recent study, titled “Implicit Data Crimes”, which showed that naïve use training of ML algorithms using open-access databases can lead to biased, overly optimistic results. Moreover, this could lead to algorithmic failure for clinical real-world data. While this phenomenon is general, examples will focus on algorithms developed for magnetic resonance imaging (MRI) reconstruction. In the second part I will focus on areas where ML can be impactful and introduce a novel technique for dynamic (video-generating) MRI, named BladeNet, which addresses some of the current barriers in abdominal pediatric imaging. BladeNet not only accelerates the data sampling process, but also removes motion-blurring from the videos. It therefore offers fast, motion-robust, high-resolution MRI protocols that are highly suitable for children. Finally, I will share my vision of how computational ML techniques can be impactful for a range of future cutting-edge healthcare applications.