Sonja Aits

This paper applies explainable AI techniques to improve transparency, data quality, and performance in lung histopathology image classification.

The paper compiles a comprehensive database of cell death regulators by curating existing resources and UniProt/Gene Ontology entries, providing a manually curated resource with associated code.

This study evaluates the reproducibility of human histology scoring for ARDS in a porcine model and demonstrates that CNNs can automate the process with acceptable performance.