CASUS Institute Seminar, Rui Li, PhD candidate, Center for Advanced Systems Understanding CASUS, Helmholtz-Zentrum Dresden-Rossendorf (HZDR)

Abstract of the talk// Understanding biological phenomena heavily relies on 3D information. While microscopy offers valuable insights, it predominantly focuses on 2D images. Some 3D imaging techniques (e.g. tomograph, confocal, etc.) often are characterized by low yielded resolution, high time-consumption, and complicated hardware. Rui in his PhD project focused on advancing 3D imaging (rapid and high-quality) by light-(LM) and cryo-electron microscopy (EM) using machine learning. Depth information restoration invokes the prominent discourse in the domain of computer vision. Nowadays, deep learning contributes significantly to systems biology studies. In this project, the 3D restoration in microscopy was formulated as an inverse problem. In the imaging process, microscopy inherently convolves the 3D information of targets into 2D images. Yet, a single 2D image corresponds to a multitude of potential 3D information configurations due to non-linearity. This makes the process an ill-posed inverse problem. Consequently, the 3D information recovery from 2D images involves solving the ill-posed inverse problem, which entails intricate aspects such as deconvolution and imaging artifacts. Solving the inverse problem using deep neural networks (DNN) involves learning the data manifold structure and the data distribution on that manifold. In this project, Rui and his colleagues present new ideas in formulating 3D microscopy imaging as ill-posed inverse problems. Adopting DNN to optical molecular tomography, they validated their approach in both use cases: LM and cryo-LM.

Rui will be talking live in Görlitz. However, as the event is organized in a hybrid format that includes a videoconferencing tool by Zoom Inc., people interested in the topic have the chance to also join the talk remotely. Please ask for the login details via