BigDataProcessor2: a free and open-source Fiji plugin for inspection and processing of TB sized image data.

Nicolas Chiaruttini,Jinbo Xu,Christian Tischer,Rainer Pepperkok,Sabine Reither,Nils Norlin,Ashis Ravindran

doi:10.1093/bioinformatics/btab106

Abstract

SummaryModern bioimaging and related areas such as sensor technology have undergone tremendous development over the last few years. As a result, contemporary imaging techniques, particularly electron microscopy (EM) and light sheet microscopy, can frequently generate datasets attaining sizes of several terabytes (TB). As a consequence, even seemingly simple data operations such as cropping, chromatic- and drift-corrections and even visualisation, poses challenges when applied to thousands of time points or tiles. To address this we developed BigDataProcessor2—a Fiji plugin facilitating processing workflows for TB sized image datasets.Availability and implementationBigDataProcessor2 is available as a Fiji plugin via the BigDataProcessor update site. The application is implemented in Java and the code is publicly available on GitHub (https://github.com/bigdataprocessor/bigdataprocessor2).Supplementary informationSupplementary data are available at Bioinformatics online.

Highlights

Inspection and processing of TB sized image data as produced by state-of-the-art light-sheet and volume electron microscopy poses several practical challenges (Power and Huisken, 2017)
The challenges of big image data inspection can be addressed by lazy-loading schemes where only the portion of the data is loaded into Random Access Memory (RAM) that is needed to render the current view on the computer monitor
Due to write performance considerations, raw microscopy data is typically not saved in a format that is compatible with those requirements

Summary

Introduction

Inspection and processing of TB sized image data as produced by state-of-the-art light-sheet and volume electron microscopy poses several practical challenges (Power and Huisken, 2017). To this end, making use of ImageJ’s virtual stacks (Schneider et al, 2012) for lazy-loading of big image data we previously developed BigDataProcessor (https://github.com/bigdataprocessor/bigda taprocessor1), a Fiji plugin (Schindelin et al, 2012) facilitating interactive browsing, processing and re-saving of TB sized microscopy raw data.

Implementation and application

Discussion and conclusions