Spatial analysis of tumor images
Title: Spatial analysis of tumor images
DNr: Berzelius-2024-129
Project Type: LiU Berzelius
Principal Investigator: Anna Sandström Gerdtsson <>
Affiliation: Lunds universitet
Duration: 2024-04-01 – 2024-10-01
Classification: 10203


Spatial analysis of multiplex immunofluorescence images of tissue microarrays will be conducted using e.g. graph neural networks and/or other methods of spatial statistics. The objective is to generate prediction models of spatial components (e.g. cellular neighborhoods) in relation to molecular profiles established by high-plex spatial omics technologies and clincial outcomes such as survival. The long term goal is to provide spatio-molecular tumor microenvironment signatures for improved patient stratification and for guiding immuno-oncology treatments.