Implicit Neural Image Field for Biological Microscopy Image Compression

TL;DR

This paper introduces an adaptive compression method using Implicit Neural Representation for biological microscopy images, achieving high compression ratios while preserving critical details for analysis.

Contribution

It presents a novel INR-based workflow that adapts to diverse bioimaging data, enabling efficient, application-specific compression of microscopy images.

Findings

Achieved up to 512x compression ratios.

Preserved detailed information for downstream analysis.

Applicable to a wide range of microscopy images.

Abstract

The rapid pace of innovation in biological microscopy imaging has led to large images, putting pressure on data storage and impeding efficient sharing, management, and visualization. This necessitates the development of efficient compression solutions. Traditional CODEC methods struggle to adapt to the diverse bioimaging data and often suffer from sub-optimal compression. In this study, we propose an adaptive compression workflow based on Implicit Neural Representation (INR). This approach permits application-specific compression objectives, capable of compressing images of any shape and arbitrary pixel-wise decompression. We demonstrated on a wide range of microscopy images from real applications that our workflow not only achieved high, controllable compression ratios (e.g., 512x) but also preserved detailed information critical for downstream analysis.