Project: High-Throughput Computational and Data Pipeline for iSCAT Microscopy
Description
Interferometric scattering (iSCAT) microscopy is a cutting-edge optical technique that detects nanoscale objects (like individual proteins or live viruses) by capturing the interference between scattered light and a reference reflection. While the physics allows for unprecedented sensitivity, the primary bottleneck is now computational: these instruments generate massive, high-speed video streams that require a robust, end-to-end data architecture to be useful for biological research.
This project focuses on the intersection of Computer Vision and Large-Scale Data Engineering. You will build a fully automated system that begins with raw video ingestion and ends with a high-performance queryable database. On the vision side, you will implement algorithms for sub-pixel 2D localization of interference ring patterns and implement multi-object tracking to reconstruct particle trajectories. These patterns encode 3D spatial data and physical properties which must be extracted with high precision.
A major pillar of this project is a Data Engineering challenge. You will design and implement a scalable storage schema and query interface capable of managing terabytes of experimental data. This involves architecting an efficient way to store raw video alongside derived time-series trajectories and physical metadata, ensuring that downstream analysis remains fast and interactive for end-users. Working closely with physicists, you will ensure the entire pipeline (including the GPU-accelerated processing and the database backend) is tailored to real-world experimental workflows.
This project is joint with the Molecular Biosensing research group of TU/e.
Details
- Supervisor
-
Robert Brijder
- Secondary supervisor
-
AKAnna Kashkanova
- Interested?
- Get in contact