CBIAS_2021_Poster_17.pdf

  

Method and software for model-based artifact correction in time-lapse fluorescence microscopy images

_{(Presented in CBIAS 2021 Conference in Francis Crick Institute, 22-23 November 2021)}

Time-lapse microscopy (TLM) is a widely used method for studying dynamic biochemical and morphological responses in biological systems. 

Unfortunately, fluorescence images often suffer from hardware-dependent intensity artifacts that degrade the quality of the image data.

Starting from a general fluorescence image formation model, we have designed several artifact correction methods addressing the most common practical cases. 

The proposed techniques may correct images for XY-dependent artifacts, including excitation non-uniformity, optical vignetting, and stray light, as well as time-dependent artifacts, e.g., arising from excitation light source power variations during a time course and inherent photobleaching of cell culture medium. 

We have implemented these methods in a convenient open-source, cross-platform MATLAB software tool with GUI, aiming to make TLM image data more amenable for further quantitative analysis.