Lung Histology

Acute respiratory distress syndrome (ARDS), a life-threatening condition with limited treatment options, requires better understanding through histopathological analysis. Scoring tissue damage based on specific features helps improve ARDS treatments, but this process is time-consuming and requires expert pathologists for consistency across species. In a study using an animal model, we aimed to automate this scoring system with Deep Convolutional Neural Networks (DCNNs). Pathologists defined key damage features, and we trained ML models to replicate their scoring, achieving a 94% correlation with gold standard scores. The following repositories are the projects have been done during PhD.

Histology [paper]

XAI

Histology Features