Research Projects

Most of my work starts with a deceptively simple question - why, how many, how much, or where exactly? Answering it usually requires building something new: a pipeline, a model, a framework. Here are some of the problems I've enjoyed solving.

Neuroscience & Brain Imaging

Neuronal Cell Segmentation in Whole Mouse Brains
Under review — Cell

An international collaboration aiming to develop a new protein-based delivery system to the central nervous system. 3D images of whole mouse-brain scans were used to detect the bio-distribution of the delivery system, using light-sheet microscopy. Image processing was performed using a custom, in-house developed pipeline employing shared-memory parallelism, targeted for fast processing of large 3D image datasets.

Light-sheet microscopy 3D segmentation HPC

Plaque and Microglia Quantification in Prime-Edited Alzheimer's Brains
Published — Nature Communications 2025

Quantification of plaques and microglia in whole murine brains for a study on prime editing of the β1 adrenoceptor to restore physiological REM sleep in an Alzheimer's disease mouse model. Machine learning-based instance segmentation (ilastik) combined with deformable registration to the Allen Brain Atlas for region-specific analysis of pathological markers.

Alzheimer's disease Prime editing ilastik Allen Brain Atlas

Brainwide Spatial Distribution of Microglia (Netoglitazone)
Published — PLoS One 2025

Quantification of brainwide, region-specific effects of the antidiabetic drug netoglitazone in Alzheimer's mouse models, using quantitative whole-brain 3D histochemistry (Q3D) and light-sheet microscopy. Whole-brain 3D stacks processed with a custom pipeline for automated microglia segmentation, registration to the Allen Brain Atlas, and statistical mapping.

Neurodegeneration Drug efficacy 3D image analysis

Computational Biophysics

Stress-Free State & Viscoelastic Properties of Red Blood Cells
Published — Biophysical Journal 2023

Introduced a transferable RBC model (t-RBC), and calibrated its viscoelastic properties as well as the stress-free state for the first time, using Bayesian inference. The t-RBC model reproduces a wide range of experimental data across multiple flow configurations without further calibration of its parameters.

Bayesian inference Red blood cells Uncertainty quantification

Hierarchical Bayesian UQ for Red Blood Cell Models
Published — Phys. Rev. Applied 2021

Developed a hierarchical Bayesian framework for uncertainty quantification of a red blood cell membrane model, enabling systematic calibration and prediction of cell dynamics under different flow conditions.

Bayesian UQ Hierarchical models Membrane mechanics

HPC & Frameworks

Korali — Large-Scale Bayesian UQ Framework
Published — CMAME 2022

An open-source framework for large-scale Bayesian uncertainty quantification and stochastic optimization. The framework relies on non-intrusive sampling of complex multiphysics models with a distributed sampling engine featuring novel fault tolerance and load balancing for massively-parallel architectures.

Open-source HPC Bayesian optimization

Load Balancing in Large Scale Bayesian Inference
Published — PASC 2020

Developed load balancing strategies for large-scale Bayesian inference on massively-parallel architectures, demonstrated across molecular dynamics, blood microfluidics, and pipe flow optimization applications.

Load balancing Distributed computing MPI

Neuronal Cell Segmentation in Whole Mouse Brains Under review — Cell

Plaque and Microglia Quantification in Prime-Edited Alzheimer's Brains Published — Nature Communications 2025

Brainwide Spatial Distribution of Microglia (Netoglitazone) Published — PLoS One 2025

Stress-Free State & Viscoelastic Properties of Red Blood Cells Published — Biophysical Journal 2023

Hierarchical Bayesian UQ for Red Blood Cell Models Published — Phys. Rev. Applied 2021

Korali — Large-Scale Bayesian UQ Framework Published — CMAME 2022

Load Balancing in Large Scale Bayesian Inference Published — PASC 2020