Bronstein Group

Bronstein Group

The rapid advances in machine learning have revolutionized nearly every domain of science and engineering. At the Bronstein Group, we develop computational data-driven methods to tackle complex challenges in natural and life sciences and engineering. We embrace interdisciplinary collaboration and actively invite students, researchers, and principal investigators across campus to engage with us.

One of our key research areas is computational imaging, a field that blends optics design, image sensor technology, and computational systems to produce highly refined images. Unlike traditional photography, where the image captured by the sensor is immediately interpretable, computational imaging captures data that may not be human-readable but contains all the information needed for advanced algorithms to reconstruct a meaningful image. Our group was among the pioneers in using machine-learning techniques to co-design both the hardware and software components of imaging systems, optimizing them for specific tasks. We have successfully applied these methods across diverse imaging modalities, including medical ultrasound, magnetic resonance imaging (MRI), radar, and electric impedance tomography. We are now expanding into molecular imaging, exploring protein structures using X-ray crystallography, and cryo-electron microscopy and tomography. More broadly, we aim to extend these innovations to the design of general sensing systems and scientific experiments.

Our second core focus is structural biology, where we leverage machine learning to model, analyze, and design protein structures, dynamics, and functions. We are especially interested in the structural implications of synonymous genetic coding in proteins—an area that challenges long-held dogmas in molecular biology, which traditionally suggested synonymous mutations had no structural impact.