[BK세미나] 3/13(금) Kevin Tsia(The University of Hong Kong) "Toward Petabyte-scale Microscopy - from Brain Imaging to Imagin

기계공학부 구성원들의 많은 관심과 참여 부탁드립니다.

▣ 주 제: Toward Petabyte-scale Microscopy - from Brain Imaging to Imaging Cytometry

▣ 연 사: Prof. Kevin Tsia

▣ 소 속: The University of Hong Kong

▣ 일 시: 2026. 3. 13.(금) 13:30

▣ 장 소: 제4공학관 D404호

▣ 초 록

High-throughput microscopy is redefining how we interrogate complex biological systems, from millisecond neural computation to large-scale single-cell phenotyping. I will introduce ultrafast two-photon fluorescence microscopes that combine all-optical laser scanning with submicron resolution to achieve kilohertz imaging across large fields and volumes in vivo. These platforms capture supra- and subthreshold neural activity at up to 3,000 frames per second to 345 μm depth in awake mice, enabling simultaneous voltage recordings from ~200 neurons and calcium imaging from >14,000 neurons in mouse visual cortex, as well as volumetric calcium imaging across the larval zebrafish brain. By pairing megahertz line-scan rates with optimized optics and computation, we preserve sensitivity while scaling both speed and coverage.

Complementing these instruments, I will present deep learning–powered, high-throughput imaging cytometry pipelines that unify biophysical and biochemical assays on a single platform. These systems deliver million-cell phenotyping within minutes, approaching petabyte-scale data generation while enhancing content, specificity, and sensitivity—particularly in biophysical and mechanical phenotyping regimes previously considered impractical. Integrated analytics convert raw image deluges into actionable features for discovery and decision-making.

Together, these advances chart a practical route toward petabyte-scale microscopy that links mechanism-rich imaging with scalable computation. We highlight applications spanning cancer and immune-cell subtyping, targeted drug sensitivity prediction, and image-based genetic screens, illustrating how unified, ultrafast imaging and AI-driven analysis can impact basic science and translational biomedicine.