Next-generation efficient 3D image data platform for scalable storage, visualisation, and analysis.

DataFlight is a UK-based biotechnology compute SME led by a project team of Joel Jonsson (technical lead) and Bevan Cheeseman (project lead).

Organoids, 3D cell cultures mimicking human organs more accurately than animal models, are revolutionizing drug discovery and personalised medicine. Despite their potential, data processing challenges hinder widespread adoption in high-throughput experiments. A single microscope can generate 0.5-12 TB/day, leading to high storage, compute, and environmental costs.

High-throughput organoid imaging at 165 TB/year incurs storage costs of £132.79K and 2.01 tonnes CO2e, and compute costs of £81.19K and 122.34 tonnes CO2e, using 538.42 kWh of energy. These costs often leave experiments unanalyzed, hindering progress. Not adopting 3D organoid and fluorescence imaging technologies will delay breakthrough therapies and increase treatment costs.

DataFlight is developing a web-based, end-to-end data platform to address these bottlenecks, making the entire pipeline faster, cheaper, and more environmentally friendly.

Core to DataFlight's innovation is Adaptive Particle Representation (APR) technology, a smart adaptive image format that replaces pixel images, improving scalability and reducing data size without compromising quality. APR optimizes storage, visualization, and analysis, including 3D rendering and deep learning methods, enhancing speed and cost-efficiency.

Using APR instead of pixel images in all data operations, DataFlight's platform significantly reduces computational and environmental burdens across the entire pipeline, diminishes bottlenecks, and unlocks scale.

DataFlight's platform removes barriers to scale and accelerates analysis workflows for high-throughput 3D organoid imaging. For users generating 165 TB/year, it offers 10 times faster processing (3.94K processor-days/year), 83% reduced net cost (£178.7K annual savings), and a 90% smaller carbon footprint (112 tonnes CO2e/year). These improvements reduce infrastructure investments and barriers to entry, enhance experiment turnover, and enable additional R&D investments of increased scale, accelerating progress.