Related Experiment Video
Updated: Jul 10, 2026
09:48
In vitro Quantitative Imaging Assay for Phagocytosis of Dead Neuroblastoma Cells by iPSC-Macrophages
Published on: February 14, 2021
Image analysis for neuroblastoma classification: segmentation of cell nuclei
Metin N Gurcan1, Tony Pan, Hiro Shimada
1Biomed. Informatics Dept., Ohio State Univ., Columbus, OH 43210, USA. gurcan.1@osu.edu
Summary
An automated method accurately segments neuroblastoma cell nuclei in images, aiding childhood cancer diagnosis. This technique achieves 90.24% accuracy, improving prognostic classification for this nervous system cancer.
Area of Science:
- Computational pathology
- Medical image analysis
- Pediatric oncology
Background:
- Neuroblastoma, a childhood nervous system cancer, requires accurate prognostic classification.
- Current classification relies partly on morphological analysis of H&E-stained cell images.
- Manual segmentation of cell nuclei is time-consuming and prone to variability.
Purpose of the Study:
- To develop an automated method for segmenting cell nuclei in neuroblastoma images.
- To improve the accuracy and efficiency of cell nuclei detection for prognostic purposes.
- To provide a quantitative tool for analyzing cellular morphology in neuroblastoma.
Main Methods:
- Development of an automated cell nuclei segmentation algorithm.
- Utilized morphological top-hat by reconstruction and hysteresis thresholding.
- Algorithm performance evaluated by comparison against manual segmentation.
Main Results:
- The automated segmentation method successfully detected and segmented cell nuclei.
- Achieved an average segmentation accuracy of 90.24% +/- 5.14%.
- Demonstrated high concordance with manual segmentation benchmarks.
Conclusions:
- The developed automated method provides accurate and reliable cell nuclei segmentation for neuroblastoma.
- This technique can enhance the objectivity and efficiency of prognostic classification.
- Potential to improve diagnostic workflows in pediatric oncology.
