State Estimation for Population Balance Models

Multicellular systems play a key role in biomanufacturing and biomedical engineering. Isogenic cell populations commonly show variability with respect to phenotypic properties, like size and biochemical composition. As this variability may result in process instabilities, and reduced product yields, close monitoring and control of the cell population heterogeneity is important. Experimental data of heterogeneous cell populations is available through high-throughput single cell measurements (e.g. flow cytometry) in the form of population snapshot data. The samples can be represented by density distributions with respect to the measured cellular properties. Technical and financial restrictions may prevent the direct measurement of all intracellular states. To reconstruct the non-measurable quantities, model-based online state estimation methods are required. Here, available data is continuously combined with mathematical model predictions. Current methods are computationally not efficient for high-dimensional models, which often arise from cell population models. This work aims at developing novel online state estimation methods that are computationally efficient for high-dimensional cell population models.