On the 96-well plate coverglass tilt and curvature suppression in 96-camera imaging system

TL;DR

This paper presents a method to suppress tilt and curvature variations of cover glasses in a 96-camera imaging system, enhancing its extended depth of focus capabilities through customized mounts and calibration procedures.

Contribution

It introduces a novel mounting and calibration approach to minimize cover glass deformation effects in multi-camera imaging systems.

Findings

Over 90% of surface variation explained by depth and tilt modes.

Customized mounts significantly reduce plate-to-plate tilt and depth variation.

Calibration procedures effectively compensate for cover glass curvature.

Abstract

The 96-eyes instrument is capable of computational extended depth of focus (eDOF) of up to +/- 30 micrometer in the phase channel, and conventional depth of field (DOF) of +/- 5 micrometer in the fluorescence channel. However, it requires minimal plate-to-plate cover glass depth variation to function. Plate depths are measured using a third-party plate scanner (Opera Phenix) grouped by plate types (Greiner UV-Star, Cell-Star, and Eppendorf meniscus-free). The two-dimensional (2D) depth dataset is aggregated through principal component analysis to obtain the top eight dominating 2D surface deformation modes. More than 90% of the variation can be explained by the plate's absolute depth and tilt (Pitch, Gradient-Y, and Gradient-X), followed by (~= 2%) the cover glass's curvature (Curve-Y and Curve-XY). Plate-to-plate average depth and tilt variations are suppressed by a customized kinematic mount anchoring the plate's cover glass at the instrument's imaging plane. The plate's average curvature is compensated by manually aligning all 96-eyes microscope objective lenses to track the plate's surface; an one-off calibration procedure aided by the backlash-free piezo-flexure z-stage. Design validation is conducted in silico, with the proof of concept experiment conducted on the 96-eyes with new mounting bracket retrofits.