Automated Image Processing for On-Board Biochemical Analysis in Space: Improvement of the image processing routine for Immunolab/IBA/IBICA

Abstract: Samples for biochemical analysis taken today on board the International Space Station need to be sent to Earth for analysis, doubling the transportation costs and making tests with short life-time impossible. In order to allow the analysis of samples in-situ, a state of the art immuno-analyser is needed, which minimises the required volume of sample and auxiliary fluids. Astrium proposes a solution using a microscope and an imaging system to measure analyte concentration using fluorescent bead-based immunoassays. The inherent limitations of microscopy required a novel image processing algorithm designed for this particular purpose.

During this study, different techniques were tested until a suitable method was developed and implemented in software. The designed algorithm included pre-processing, segmentation and classification, as well as the determination of an optical value representing the concentration of the analyte. Tests showed that the new method could identify up to three times the amount of beads in comparison to previous methods based solely on thresholding, because of the added capability of evaluating lines and clusters of beads. The statistical support was then increased and therefore the lower limit of detection lowered, enabling the measurement of much smaller concentrations.