Related Experiment Video
Updated: Dec 2, 2025
07:17
Author Spotlight: Studying Spatial Protein Expression Using Agarose Embedded Lung Tissue Sections
Published on: October 6, 2023
7.1K
Deep-LUMEN assay - human lung epithelial spheroid classification from brightfield images using deep learning
Lyan Abdul1, Shravanthi Rajasekar2, Dawn S Y Lin2
1School of Biomedical Engineering, McMaster University, 1280 Main Street West, Hamilton, ON, L8S 4L8, Canada.
Lab on a Chip
|November 5, 2020
Summary
A new deep learning assay, Deep-LUMEN, non-invasively measures changes in 3D epithelial spheroid morphology. This method accurately assesses drug toxicity by considering tissue architecture, improving pre-clinical drug studies.
Area of Science:
- Biomedical Engineering
- Cell Biology
- Drug Discovery
Background:
- Three-dimensional (3D) tissue models like spheroids and organoids are crucial for pre-clinical drug studies.
- Characterizing 3D tissue models non-invasively from brightfield images presents a significant challenge compared to 2D cultures.
Purpose of the Study:
- To develop a novel deep learning algorithm for automated, non-invasive characterization of epithelial spheroid morphology.
- To enable the tracking of luminal structure changes and differentiation between polarized and non-polarized lung epithelial spheroids.
Main Methods:
- Development of Deep-LUMEN (Deep Learning Uncovered Measurement of Epithelial Networks), an object detection algorithm fine-tuned for spheroid analysis.
- Utilizing brightfield microscopy to capture images of 3D tissue models.
- Validation through screening morphological changes in response to extracellular matrices and drug treatments.
Main Results:
- Deep-LUMEN successfully identified subtle differences in epithelial spheroid morphology.
- The assay distinguished between polarized and non-polarized lung epithelial spheroids.
- Cyclosporin's dose-dependent toxicity was found to be underestimated without considering morphological changes, highlighting the assay's utility.
Conclusions:
- Deep-LUMEN provides a non-invasive method for assessing drug effects on 3D spheroid models.
- The assay captures critical morphological changes, enhancing the accuracy of pre-clinical drug evaluation.
- This technology can improve the reliability of drug screening using advanced 3D tissue models.
