Related Experiment Video
Updated: May 19, 2026
08:40
Quantitation of Protein Expression and Co-localization Using Multiplexed Immuno-histochemical Staining and Multispectral Imaging
Published on: April 8, 2016
Machine learning and automation methods for the segmentation, classification and quantification of testicular tissue
Adam J R Gadd1, Iris Sanou2,3, Eleanor Brain4
1Centre for Reproductive Health, Institute for Regeneration and Repair, The University of Edinburgh, Edinburgh, UK.
Reproduction & Fertility
|May 18, 2026
Summary
Automating immunofluorescent image analysis of testicular tissues using machine learning significantly speeds up cell quantification and improves data accuracy. This novel method enhances the analysis of cellular organization and mitotic index in both human and mouse samples.
Area of Science:
- Reproductive biology
- Computational pathology
- Bioinformatics
Background:
- Immunofluorescent image analysis is crucial for studying testicular tissue.
- Manual analysis is time-consuming, subjective, and can lead to inaccurate results.
- Developing automated methods is essential for efficient and reliable research.
Purpose of the Study:
- To develop and validate a standardized automated method for analyzing immunofluorescent images of human and mouse testicular tissues.
- To compare the efficiency and accuracy of automated analysis with manual methods.
- To enable detailed analysis of cellular parameters like mitotic index and organization.
Main Methods:
- Utilized QuPath software to train an artificial neural network with multilayer perception (ANN-MLP).
- Employed StarDist and Watershed transformations for automated cell segmentation in regions of interest.
- Applied QuPath's object classification system for cell marker analysis (SOX9+, MAGE-A+).
Main Results:
- Automated segmentation and classification maintained correlations between tubular area and cell counts (SOX9+: r2=0.56, p=0.03; MAGE-A+: r2=0.93, p=0.002), unlike manual methods.
- The automated method was significantly faster (3 seconds vs. 6,247 seconds per image).
- Exported data to R/Rstudio for advanced analysis of mitotic index and cellular organization.
Conclusions:
- The developed automated method provides a standardized, efficient, and accurate approach for analyzing testicular immunofluorescent images.
- This technique overcomes limitations of manual analysis, improving data reliability and research throughput.
- The method facilitates in-depth quantification and analysis of cellular characteristics in reproductive tissues.
