Dr Christoffer Nellåker talk

Disease and healthy phenotyping from image data

Dr. Nellaker has been working in the translational field of computational phenotyping since 2011.  Specifically, this has encompassed machine learning approaches to analyse imaging data to extract biologically meaningful phenotype descriptors. Using facial images to aid diagnosis of rare diseases is one such domain where visually detectable phenotypes can be associated to distinct genetic causes. Recent work into histology analyses shows the potential for cellular phenotype information to be extracted from placental tissue samples stained with H&E. From this data large scale quantification of tissue phenotypes can be distilled and mapped onto cell biology traits. These methods hold promise to speed and scale high throughput analyses of imaging data for novel and accurate quantification of normative and pathological biological processes.