Life Beyond the Pixels:
Deep Learning Methods for Single Cell Analysis
Prof Dr Peter Horvarth
Institute Director and Group Leader, Institute of Biochemistry, Biological Research Centre, Szeged and Research Fellow at the Institute for Molecular Medicine Finland (FIMM), Helsinki
Read BioPeter Horvath is the director of and a group leader in the Biological Research Center in Szeged and holds a Finland Distinguished Professor (FiDiPro) Fellow position in the Institute for Molecular Medicine Finland. He received his PhD from INRIA and the University of Nice Sophia Antipois in satellite image analysis. He co-founded the European Cell-based Assays Interest Group and is a Society of Biomolecular Imaging and Informatics councillor.
CloseIn this webinar, you will discover:
- the computational steps in the analysis of single cell-based large-scale microscopy experiments;
- a novel microscopic image correction method designed to eliminate illumination and uneven background effects;
- new single-cell image segmentation methods;
- the Advanced Cell Classifier, a machine learning software tool for identifying cellular phenotypes;
- how the above machine learning models were used to select cells and perform a range of measurements.
This webinar provides an overview of the computational steps in the analysis of single cell-based large-scale microscopy experiments. First, you’ll learn about a novel microscopic image correction method designed to eliminate illumination and uneven background effects, which, left uncorrected, corrupt intensity-based measurements. New single-cell image segmentation methods will be presented using differential geometry, energy minimization, and deep learning methods (www.nucleAIzer.org).
Second, we’ll introduce the Advanced Cell Classifier (ACC) (www.cellclassifier.org), a machine learning software tool capable of identifying cellular phenotypes based on features extracted from the image. It provides an interface for users to efficiently train machine learning methods to predict various phenotypes. For cases where discrete cell-based decisions are not suitable, we suggest using multi-parametric regression to analyze continuous biological phenomena. To improve learning speed and accuracy, we propose an active learning scheme that selects the most informative cell samples.
Finally, you will see how the above machine learning models were used to develop single-cell isolation methods based on laser-microcapturing and patch clamping. You will also discover how DNA and RNA sequencing, proteomics, lipidomics, and targeted electrophysiology measurements were successfully performed on isolated cells.