Seminar Date: Tuesday, January 20, 2026
Time: 11:00 AM PT
Location: 67-3111 & Zoom
Talk Title: Advanced cryo-electron tomography via nonlinear phase retrieval
Zoom link
Add this event to your calendar
Abstract:
Cryogenic electron tomography (cryo-ET) enables three-dimensional imaging of radiation-sensitive materials at nano- or atomic resolution. However, conventional cryo-ET techniques face fundamental resolution limits due to nonlinear image contrast and multiple electron scattering, challenges that become severe for thick or compositionally heterogeneous specimens. To overcome these limitations, we developed an advanced cryo-ET reconstruction framework based on nonlinear phase retrieval and its pipeline. This approach explicitly models the full three-dimensional electron scattering process using a multislice method, accurately accounting for dynamical scattering within the specimen. We further incorporate Bayesian optimization to refine tilt-series alignment and correct optical aberrations, enabling highly precise reconstruction. This talk presents the basic principles of the framework and its experimental demonstrations. Using standard cryo-TEM instrumentation, we achieved a three-dimensional resolution of 1.6 Å, surpassing the limits of conventional tomography methods. The resolution is validated by validated by multislice simulations, and the method is extended to biological specimens such as HIV-1 particles, yielding improved reconstruction quality and reduced reconstruction artifacts. This framework provides a practical route for high-resolution 3D imaging of beam-sensitive, heterogeneous, or isolated systems, opening new opportunities for atomic-scale studies across materials and biological sciences.
Bio:
Juhyeok Lee is a postdoctoral researcher co-advised by Prof. Mary Scott and Dr. Michael Whittaker. He received his BS degree in Physics from Korea University in 2017 and his Ph.D. in Physics at the Korea Advanced Institute of Science and Technology (KAIST) in 2023 under the supervision of Prof. Yongsoo Yang. His doctoral work primarily focuses on development of deep learning based atomic resolution electron microscopy and algorithm development of 4D-STEM-based electron tomography. His broader research interests include tomography algorithm development, TEM/STEM measurement, and the applications of deep learning for image refinement and reconstruction. Currently, he is working on advancing HRTEM based tomography algorithms development and their experimental implementation for high-resolution 3D imaging of nanomaterials and beam-sensitive system.
