Related Experiment Video
Updated: Dec 2, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
3.2K
Introducing Biomedisa as an open-source online platform for biomedical image segmentation
Philipp D Lösel1,2, Thomas van de Kamp3,4, Alejandra Jayme5,6
1Engineering Mathematics and Computing Lab (EMCL), Interdisciplinary Center for Scientific Computing (IWR), Heidelberg University, Im Neuenheimer Feld 205, 69120, Heidelberg, Germany. philipp.loesel@uni-heidelberg.de.
Nature Communications
|November 5, 2020
Summary
Biomedisa is a new platform for semi-automatic segmentation of large volumetric images. It significantly reduces time and effort for scientists, even with limited computational expertise.
Area of Science:
- Biomedical imaging
- Computational biology
- Medical image analysis
Background:
- Accurate segmentation of large volumetric images is crucial for various biomedical applications.
- Manual segmentation is time-consuming and requires significant expertise.
- Existing semi-automatic tools often require complex configurations or lack efficiency.
Purpose of the Study:
- To introduce Biomedisa, a user-friendly, open-source platform for semi-automatic segmentation of large volumetric images.
- To provide an efficient tool for scientists with limited computational expertise.
- To reduce the time and human effort required for image segmentation.
Main Methods:
- Developed an open-source online platform, Biomedisa.
- Implemented a smart interpolation technique for sparsely pre-segmented slices.
- Leveraged complete underlying image data for enhanced segmentation accuracy.
- Ensured web-browser accessibility with no complex software configuration.
Main Results:
- Biomedisa drastically reduces time and human effort for segmenting large images.
- Demonstrated significant improvement over conventional morphological interpolation methods.
- Outperformed other segmentation tools that consider underlying image data.
- Validated its effectiveness across different 3D imaging modalities and biomedical applications.
Conclusions:
- Biomedisa offers an efficient and accessible solution for semi-automatic image segmentation.
- The platform is particularly valuable for tasks requiring dense annotation, such as training deep neural networks.
- Biomedisa empowers scientists without extensive computational backgrounds to perform complex segmentation tasks effectively.
