Welcome to my website!

I am a Mechanical Engineer turned Computational Scientist, fascinated by solving complex problems that arise when humans decide to quantify nature.

Applied AI/ML · Probabilistic Modeling · Data Science · Computer Simulations · High Performance Computing

During my Postdoc, I collaborated with biologists to count millions of tiny cells inside terabyte-sized whole-brain scans, to understand the progression of neurodegenerative brain diseases; partnered with neuroscientists to quantify the efficacy of novel drugs for Alzheimer's disease, based on the drug's dispersion in the brain; and joined highly inter-disciplinary efforts to measure genetically induced cell toxicity, by analyzing millions of images acquired from genome-wide genetic perturbation (CRISPR) screens. As a PhD student, I was captivated by the inherent flexibility of erythrocytes and studied how their shape is optimized by nature, for maximum delivery of oxygen in arterioles. I modeled the viscoelastic material properties of erythrocyte membranes and developed a hierarchical Bayesian model, able to predict their flow and reproduce experimental data and their uncertainty. I simulated the flow of blood with sub-cellular resolution using mesoscopic, particle-based methods involving energy potentials, and derived dimensionless quantities fully describing the dynamics of the system. I learned about fluid-structure interactions and modeled the interactions between erythrocytes, the surrounding fluid, and walls of microfluidics devices. To save computational resources, I trained ML-based surrogate models to emulate the computationally expensive high-fidelity blood flow simulations.

A few highlights are showcased below — you can explore the full set on the Projects page.

Get in touch Google Scholar

Selected Research Projects

Neuronal Cell Segmentation in Whole Mouse Brains
Under review — Cell

An international collaboration developing a protein-based delivery system to the CNS. 3D whole mouse-brain scans processed with light-sheet microscopy using a custom, in-house pipeline with shared-memory parallelism for fast processing of large 3D datasets.

Microglia Segmentation in Drug Efficacy Studies
In preparation

A novel image processing pipeline for quantifying drug efficacy at the sub-cellular level through object segmentation of stained neural cells in 3D whole mouse-brain scans.

Transferable Red Blood Cell Model with Bayesian Inference
Published

Introduced a transferable RBC model (t-RBC) with viscoelastic properties and stress-free state calibrated using Bayesian inference for the first time.

View all projects →

Open-Source Software

img3D — 3D Volumetric Image Processing
Python · C

Python/C package for image processing of large, 3D (volumetric) image data. Supports raw file I/O, NRRD metadata, and parallel computation with OpenMP.

AggreQuant — Automated High Content Screen Analysis
Python

Automated pipeline for segmentation and co-localization analysis of high content screening data, using deep learning (StarDist, Cellpose) for nuclei and cell segmentation.

View all tools →

Neuronal Cell Segmentation in Whole Mouse Brains Under review — Cell

Microglia Segmentation in Drug Efficacy Studies In preparation

Transferable Red Blood Cell Model with Bayesian Inference Published

img3D — 3D Volumetric Image Processing Python · C

AggreQuant — Automated High Content Screen Analysis Python

Athena Economides

Neuronal Cell Segmentation in Whole Mouse Brains
Under review — Cell

Microglia Segmentation in Drug Efficacy Studies
In preparation

Transferable Red Blood Cell Model with Bayesian Inference
Published

img3D — 3D Volumetric Image Processing
Python · C

AggreQuant — Automated High Content Screen Analysis
Python