Imaging throughput of compact handheld microscopes for quantitative single cell studies

Sophie Bulloch, Tienan Xu, David Herrmann, Paul Timpson, Tri Giang Phan, Yean Jin Lim, Woei Ming Lee

Research output: Working paperPreprint

Abstract

Incuscopes, incubator-compatible microscopes, are crucial for live single-cell imaging studies that spans several hours to days. However, traditional microscopy prioritize high-resolution imaging performance over throughput, neglecting efficient live-cell image sampling. This study challenges existing spatial bandwidth product (field of view/optical resolution) criteria for image sampling for live cells. We demonstrate that imaging throughput is fundamentally determined by the minimal pixel count necessary to adequately resolve single cells across the field of view, not spatial bandwidth product. Using an off-the-shelf handheld microscope (5MP, ~0.03 NA) and a scientific microscope (8MP, 4x, 0.4 NA), we revealed a striking disparity. Contrary to expectations, the handheld microscope exhibited ~4-fold higher imaging throughput, highlighting oversampling inherent in many scientific microscopes. This efficiency stems from a more optimized pixel-to-cell ratio for throughput. We validated this concept by deploying the handheld microscopes within a compact 30-liter incubator, enabling continuous imaging over 40 hours using open-source Micro-Manager. A series of experiments, including cell counting, tracking, division, migration, and spheroid dynamics, were successfully performed. The handheld microscope's compactness, ease of use, and cost effectiveness render it a compelling alternative to high-grade incubator microscopes for routine, non-fluorescence cell culture studies, offering a paradigm shift towards pixel optimized imaging throughput.
Original languageEnglish
Number of pages18
DOIs
Publication statusPublished - 2025

Fingerprint

Dive into the research topics of 'Imaging throughput of compact handheld microscopes for quantitative single cell studies'. Together they form a unique fingerprint.

Cite this