ReproHack23 Bio-image analysis

A small and diverse group of enthusiastic image data analysts, an imaging data steward, and a bioinformatics trainer supported by technical HPC experts joined forces to assessing the expected variability in reproducible data analysis for some biological imaging data sets.   

Despite numerous efforts towards FAIR data and an imminent reproducibility crisis, reproducibility in the context of biological imaging remains an elusive concept. Intuitively there is an expectation for biological variation; however, for the image analysis, there is the expectation of digital duplicability. However, in the light of different implementations, hardware, and image input, some yet unexplored amount of variability is inevitable. During the ReproHack, we explore the effect of the amount of analysis metadata on result variability in order to define expectations with respect to the reproducibility of biological image analysis. Next, we quantify the differences stemming from reproducing the results using the same analysis tools as well as using similar analysis tools for reproducing the results of earlier analysis. In addition, we aim to unravel the contribution of acquisition to the variability. 

Some impressions of the days can be found on https://twitter.com/hashtag/reprohack23

Here is the list of participants:
Alexander Botzki (1), Benjamin Pavie (2,3,4), Nicolas Peredo (2,3,4), Tatiana Woller (2,3,4,5), Till Korten (6), Robert Haase (6), Tim Van De Looverbosch (7,8,9), Marlies Verschuuren (7,8,9), Winnok H. De Vos (7,8,9), Chris Cawthorne (10),  Ingrid Barcena Roig (11),  Jan Ooghe (11), Sebastian Munck (2,3,4)

1 VIB Technology Training
2 VIB BioImaging Core
3 VIB - KU Leuven Center for Brain and Disease Research Center
4 KU Leuven Department for Neuroscience
5 VIB Data Core
6 DFG Cluster of Excellence 'Physics of Life,' TU Dresden, Germany.
7 Department of Veterinary Sciences, Laboratory of Cell Biology and Histology and Antwerp
8 Centre for Advanced Microscopy (ACAM), University of Antwerp, 2610 Antwerp, Belgium.
9 μNEURO Research Centre of Excellence, University of Antwerp, 2610 Antwerp, Belgium.
10 Nuclear Medicine and Molecular Imaging, Department of Imaging and Pathology, KU Leuven, 3000 Leuven, Belgium
11 Flemish supercomputing center (Vlaams Supercomputer Centrum, VSC) and KU Leuven’s central IT infrastructure

supported by