Results Of The First Defocus Particle Tracking Challenge

This work presents the results of the First Defocus Particle Tracking (DPT) Challenge. The Challenge was structured with experimental and synthetic datasets. The experimental datasets consisted of a physical microfluidic device, the Chip-in-a-Box (ChiB), directly shipped to the interested participants. The synthetic datasets, referred to as the Multi-Flavour Datasets (MFDs), consisted of images generated with a ray-tracing approach capable of generating different types of defocusing images with a unified methodology. Seven international research teams submitted results in at least one case of the Challenge. A great variety of established defocus-based approaches (or flavors) was reflected by the contributing teams, spanning from three-pinhole techniques and shadowgraphy to astigmatism particle tracking. Five teams submitted results for the experimental test cases and applied a DPT method with their favorite flavor in one microfluidic experiment in the ChiB. By optimizing the experimental conditions, reliable particle tracking with a spatial resolution of up to 2.7 × 10⁻² velocity vectors per µm³ was achieved in two demanding benchmark experiments. Five teams submitted results for synthetic MFDs showing overall excellent results, with median errors down to 0.017 pixel for the in-plane coordinate and 0.17 pixel (0.07 % of the full depth size) for the depth coordinate up to a particle image density of 0.0016 particles-per-pixel (ppp). The results demonstrate that DPT evolved as a versatile and precise measurement technique, encompassing a broad spectrum of methodological flavors.