Texture-based analysis of droplet microfluidics for detecting bacterial growth and heteroresistance

Texture feature engineering for bacterial growth detection in droplet microfluidics.

This project focused on developing robust texture estimation parameters from droplet microfluidics images to detect bacterial growth and heteroresistance patterns. Texture features extracted from each droplet were used to identify the presence of bacterial growth by quantifying changes in spatial intensity patterns and local image heterogeneity. The pipeline combines image preprocessing, feature extraction, and quantitative analysis to characterize droplets and support rapid antimicrobial resistance screening. The work was carried out in collaboration with the BioImage Informatics Unit and delivered as reproducible code in Python.