Scientific Image Processing and Analysis
Please note, that the registration has to occur
via the iGRAD-Office!
|Audience||Doctoral Researchers and Postdocs, no previous software knowledge required|
|Competence Area||This course is assigned to Competence Area 3:|
allocation not defined
|Registration||Registration is open for all doctoral iGRAD members and doctoral researchers of the associated programs. Postdocs can participate upon request. |
For a binding registration please send an email to:
The workshop – “Scientific Image Processing and Analysis” aims to teach natural scientists from different areas of life science how to handle and process digital images starting from microscopic image acquisition on until the incorporation into the final publication figure. This includes important theory about high quality digital images in general as well as a broad spectrum of methods for correct image processing and specific analytical purposes according to high scientific standards. You will learn how to extract most information from your images and how to quantify regions of interest. Additionally, the workshop includes a lot of hands-on sessions and explains how to save time during repetitive image processing steps or while building your publication figures in a way that preserves image quality and stores processing data. You will be able to revisit the learned material using the provided exercises and script also later on. The workshop content is generally of importance for scientists working with digital images and their analysis. Optional topics will be adjusted to meet the participants needs as good as possible if there is time left.
Furthermore, specific participant question regarding image processing or solutions for analysis issues can be personally discussed if communicated beforehand (e-mail with question and example images).
Specific Topics (among others)
- Basics about correct image acquisition
- How to achieve good image resolution (sampling)
- Image formats - which formats serve scientific images and which should be avoided
- Metadata - information saved beyond the visible image
- Information content of images - learn about bit-depth, color spaces and different image types (or: how much information can be saved in and retrieved from an image).
- Correct image adjustments avoiding alterations - contrast and brightness, image rotation, size changes, background subtraction methods.
- Use of different image filters to improve extractability and preparation for further analysis
- Image segmentation - How to extract specific objects of interest (e.g. cells positive for a certain marker stain)
- Automated object counting and tracking of moving objects (optional)
- Basic 3D reconstructions (e.g. microscopic z-stacks, or medical MRI-Data)
- Image Quantifications (selected topics depending on participants field of interest):
- Measurements of areas, length, …
- 3D object analysis (volume, surface, distances, angles…)
- How to correctly measure intensities in images (e.g. fluorescence)
- Dimension scaling and calibration of images
- Labeling of images and time series/movies (text, numbers, scale bars, calibration bars,…)
- Ethics in image handling, processing and publication - where are the limits?!
- How to efficiently prepare publication figures.
The workshop should give scientists a better understanding about the Do's and Don'ts during digital image processing and insight in the methodology of extracting a multiplicity of information from their images. The participants will gain extensive knowledge about the possibilities they have to analyze their imaging data.
PhD Students and PostDocs working or planning to work with digital images. The workshop has a focus on microscopic images but all the content is applicable for digital images of different origin in general. No previous software knowledge required.
During the practical parts of the workshop we will use the professional software Fiji (customized ImageJ bundle). All necessary software will be provided (open source!).
Dr. rer. nat. Jan Brocher, Bioimaging Analyst, biovoxxel