A pipeline for quantifying cell damage in whole-slide fluorescence images of skeletal muscle

Quantitative fluorescence imaging of formalin-fixed paraffin-embedded (FFPE) tissue is often limited by intensity heterogeneity, endogenous autofluorescence, and fixation-induced artifacts. Together, these factors reduce analytic accuracy and reproducibility. In murine skeletal muscle, dyes such as Procion Yellow (ProY) are used to identify membrane-compromised cells following injury; however, overlapping autofluorescence and uneven staining hinder reliable quantification. Existing segmentation workflows, including ImageJ-based approaches, are sensitive to these variations, and standard preprocessing methods often fail to adequately normalise fluorescence intensity across whole-slide images. Here, we present a workflow for quantitative analysis of ProY-stained FFPE skeletal muscle. The pipeline combines spectral characterisation of the dye and autofluorescence, optimised whole-slide fluorescence image acquisition, ratiometric intensity normalisation, automated segmentation using Cellpose, adaptive thresholding, and particle analysis. This approach improves segmentation robustness and consistency in highly autofluorescent FFPE tissue sections while reducing user-dependent variability. As proof-of-principle, validation in mechanically injured murine skeletal muscle demonstrated that the workflow could distinguish between different levels of tissue injury. This workflow provides a quantitative approach for fluorescence-based imaging in preclinical studies, with potential for future integration into more standardised clinical histopathology workflows. Although optimised for ProY-labelled skeletal muscle, the pipeline could be adapted to other dyes and tissue types affected by autofluorescence.