Bioimaging data analysis using High Performance Computing (HPC)
Bioimaging data analysis using High Performance Computing (HPC)
TBA
General context
Asa datasets grow in size and complexity, traditional desktop tools often fall short. This introductory training shows you how to leverage the Flemish Supercomputer Center (VSC) infrastructure for efficient and reproducible image analysis. It is intended for scientists who work with microscopy or bioimage data and want to scale up their analysis using high-performance computing (HPC)
You will learn to access HPC resources via the Open OnDemand portal, run GUI-based tools (like napari or Fiji), and adapt existing pipelines (e.g., from ZeroCostDL4Mic or the BioImage Model Zoo) in a Jupyter notebook environment.
- Transfer and organize microscopy data on the HPC system to prepare scalable analysis workflows.
- Execute pre-installed bioimage analysis tools (e.g., Ilastik, Fiji, Napari) on the VSC for tasks such as segmentation and denoising.
- Run image analysis pipelines using a Jupyter notebook.
- Adapt existing image analysis pipelines (e.g., from ZeroCostDL4Mic or BioImage Model Zoo) to run efficiently on the HPC infrastructure.
- Evaluate the reproducibility of bioimage analysis workflows and propose improvements using HPC-based solutions.
This course combines short lectures with guided hands-on exercises on the HPC.
This training is recommended for people with no prior experience in HPC who plan to utilize HPC for bioimage analysis tasks.
Notions of command line and scripting, being comfortable with a remote desktop, and general knowledge of image analysis.
This course is part of the learning path Bioimaging.
AI methods used
The course covers machine learning and deep learning methods for large-scale bioimage analysis, including Random Forest for classification, neural networks for segmentation with StarDist, and panoptic segmentation with Empanada.
