Dr. Loïc Royer: Future of Bioimaging: Next Generation Instruments & Artificial Intelligence

Recording of Loïc Royer (Chan Zuckerberg BioHub) talk on October 19, 2023, at the EPFL Seminar Series in Imaging.
Abstract:
Imaging and comprehending life's processes, from the localization of proteins in cells to the complex cellular choreography in developing embryos, requires the unique synergies that arise from an interdisciplinary approach. From its inception, my team's strategy has been to drive imaging innovations on three fronts by: (i) developing novel state-of-the-art fluorescence imaging instruments that can capture in ever-increasing spatiotemporal resolution and scale the complex dance of cells in large living specimens, (ii) leveraging self-supervised deep-learning approaches to process and analyze bioimage data, (iii) making an effort to widely disseminate our algorithmic approaches via open-source and easy-to-use packages. In this talk, I will delve into the array of hardware and algorithmic technologies we've engineered, including the development of advanced light-sheet microscopes (DaXi), novel methods for self-supervised learning for bioimaging (CytoSelf), fast and scalable image processing for light-sheet time-lapse data (DEXP), high-accuracy cell tracking at scale in developing embryos (ultrack), fast, robust, and user-friendly image denoising (Aydin, Noise2Self), and using napari and large language models for accelerating the dissemination of bioimaging skills and methods (Omega). I will also share my vision of the future of bioimage analysis and what I believe the future holds.