Skip to main content Skip to navigation
BioImage Informatics Conference 2017

Digital image processing for drug screening; hierarchical approaches for sparse image data of cells, model organisms and tissue

Carolina Wählby, Dept. of Information Technology, Uppsala University

Microscopy imaging is one of the most powerful tools to investigate complex biological processes, and automated analysis methods are capable of measuring a large number of parameters from a broad range of samples in parallel. Rapidly developing high-throughput techniques are capable of generating data at an unprecedented rate placing the biological and biomedical sciences on the verge of a digital explosion. Transformative approaches to analysis of massive spatial and temporal image data are urgently needed, or there is a real risk that the promised, rapid advancement in knowledge will not materialize. To meet this challenge, we make two fundamental observations: (1) Not all data contain valuable information. With datasets outgrowing resources we cannot afford to store and analyze data that lack scientifically relevant information or suffers from quality problems. (2) Given limited resources, or if real-time decisions are needed, we have to be smart about which subsets of the data we use for a detailed, costly analysis.

I will present ongoing studies where data is approached in a hierarchical way. In the first study we characterize cancer stem cells with the aim is to identify disease subgroups and elucidate drug-specific mechanisms of action. In the second study we use zebrafish, a well-established and effective model organism, and develop digital image processing and analysis approaches using both traditional and deep learning to score phenotypic responses. Finally, I will present our most recent work on analysis of gene expression in intact tissue samples.  

O. Ishaq, S.K. Sadanandan, and C. Wählby. Deep Fish: Deep Learning-Based Classification of Zebrafish Deformation for High-Throughput Screening. Journal of Biomolecular Screening, Sept 2016.

C. Pardo-Martin, A. Allalou, J. Medina, P.M. Eimon, C. Wählby, and M.F. Yanik. High-throughput hyperdimensional vertebrate phenotyping. Nature Communications, 2013 Feb 12; 4:1467

C. Wählby, L. Kamentsky L, Z.H. Liu, T. Riklin-Raviv, A.L. Conery, E.J. O’Rourke, K.L. Sokolnicki, O. Visvikis, V. Ljosa, J.E. Irazoqui, P. Golland, G. Ruvkun G, F.M. Ausubel, and A. E. Carpenter. An image analysis toolbox for high-throughput C. elegans assays. Nature Methods, 2012 Apr 22; 9(7): 714-716.

R. Ke, M. Mignardi, A. Pacureanu, J. Svedlund, J. Botling, C. Wählby, and M. Nilsson.In situ sequencing for RNA analysis in preserved tissue and cells. Nature Methods, 2013 (10), 857-860.

Carolina Wählby is Professor at the Dept. of Information Technology, Uppsala University, chairman of the board of the Centre for Image Analysis (cb.uu.se), and Director of the BioImage Informatics facility of the SciLifeLab (www.scilifelab.se). She received a MSc in Molecular Biotechnology in 1998, a PhD in Digital Image Analysis in 2003, and carried out PostDoctoral research at the Dept. of Genetics and Pathology, all at Uppsala University. She joined the Broad Institute of Harvard and MIT in 2009 as PI at the Imaging Platform developing algorithm for high-throughput screening using model organisms. She stayed at the Broad until 2015 but returned part-time to Sweden as a SciLifeLab strategic recruitment in 2011 and became full professor in 2014. She was elected ISAC scholar 2014, 

received the SBI2 President’s innovation award in 2014, and the Thuréus prize from The Royal Society of Sciences in 2015. She received an ERC consolidator grant in 2015 for TissueMaps; a project aimed at image-based sequencing of RNA in situ. In 2017 she received a Big Data framework grant from the Swedish Foundation for Strategic Research, further expanding her research in the interface between biomedicine, microscopy, and computer science in collaboration with AstraZeneca and Vironova.