A Dual Image Sensor Approach for Automated, High Resolution, Region-of-Interest Imaging in a 96-Well Plate

Au, Aaron 1 ;  Maximiliano Giuliani 2 ;  Yip, Christopher 1, 2, 3

1. Chemical engineering and Applied Chemistry; 2. Institute of Biomaterials and Biomedical Engineering; 3. Biochemistry

Phenotypic screens are performed on model organisms to understand how chemicals compounds affect biological systems. Recently we developed a low cost high throughput imaging solution was developed to capture the effects of chemicals on the egg-laying behavior of C. elegans [1]. We present a second generation prototype capable of high-resolution (0.2 µm/px), automatic imaging of 96-well plates while efficiently collecting only the areas where worms are present.

In order to efficiently obtain automatic high-resolution imaging we developed a dual imaging arm approach: a low resolution arm provides overall tracking of regions of interests (ROIs); while a second high resolution arm provides a high-resolution of the respective ROIs. In the low resolution arm a CCD camera (Thorlabs) collect an image of the entire well.  The image is fed to a python algorithm that segments the worms and the skeleton of the segmented worms are used as a path for the motorized stage to capture image of individual worms with a line-scanning system.  The scan time of a 96-well plate will be proportional to the number of ROIs, but will be significantly less than imaging an entire plate.

 

The high resolution images are fed into an image processing pipeline that will be able to classify the effects of chemicals on the internal structure of the nematodes.  With continuous tracking from the low resolution camera, automatic adjustments with the motorized stage can be made to compensate for the movement of non-sedated worms.  This feature maybe useful in tracking worm movements and screen for deficiencies in worm mobility. This low cost solution provides users with an automated approach to obtain high-resolution images and dynamic information of C. elegans.

 

[1] Au A., Giuliani M., Harrington S., Roy P., & Yip C. (2016, Feb). A low cost, high-resolution, high-throughput imager for phenotypic screening on a 96-well plate. Poster presented at the Biophysical Society Annual Meeting in Los Angeles, California