Related Experiment Video
Updated: Jun 24, 2026
09:58
DNA-barcode-based Multiplex Immunofluorescence Imaging to Analyze FFPE Specimens from Genetically Reprogrammed Murine Melanoma
Published on: June 6, 2025
A comparative study of individual and ensemble majority vote cDNA microarray image segmentation schemes, originating
Antonis Daskalakis1, Dimitris Glotsos, Spiros Kostopoulos
1Department of Medical Physics, Medical Image Processing and Analysis Laboratory, School of Medicine, University of Patras, Rio, Patras, Greece. daskalakis@med.upatras.gr
Computer Methods and Programs in Biomedicine
|March 13, 2009
Summary
This study evaluated segmentation algorithms for gene expression analysis in cDNA microarray images. An ensemble method achieved the highest accuracy, while the Histogram Concavity algorithm offered optimal performance for low-quality images.
Area of Science:
- Bioinformatics
- Computational Biology
- Genomics
Background:
- Accurate gene expression level extraction from cDNA microarray images is crucial for genomic studies.
- Existing segmentation methods may not be optimal for diverse image qualities and noise levels.
- Novel segmentation approaches are needed to improve the reliability of microarray data analysis.
Purpose of the Study:
- To comparatively evaluate the performance of various segmentation algorithms for gene expression intensity extraction in cDNA microarray images.
- To assess the impact of noise reduction techniques on segmentation accuracy.
- To identify optimal algorithms or ensemble structures for both high-quality and low-quality microarray images.
Main Methods:
- Employed segmentation algorithms based on histogram and unsupervised classification methods, some novel to microarray analysis.
- Utilized individual algorithms and ensemble majority vote structures for spot-image segmentation.
- Evaluated algorithm performance using simulated and real cDNA microarray images, assessing validity and reproducibility.
Main Results:
- An ensemble structure (Histogram Concavity, Gaussian Kernelized Fuzzy-C-Means, Seeded Region Growing) achieved the highest segmentation accuracy on high-quality simulated images.
- The Histogram Concavity algorithm demonstrated optimal performance regarding processing time and segmentation precision for low-quality simulated and real cDNA microarray images.
- Noise reduction steps were incorporated to enhance the segmentation process.
Conclusions:
- Ensemble segmentation strategies can significantly improve accuracy in gene expression analysis from cDNA microarrays.
- The Histogram Concavity algorithm provides a robust and efficient solution for segmenting challenging, low-quality microarray images.
- These findings contribute to more reliable and reproducible gene expression data extraction.
