Related Experiment Videos
Image segmentation of cribriform gland tissue
D Thompson1, P H Bartels, H G Bartels
1Optical Sciences Center, University of Arizona, Tucson 85721, USA.
Analytical and Quantitative Cytology and Histology
|October 1, 1995
Summary
This study presents an automated method for segmenting cribriform prostatic glands, achieving 70-80% accuracy. This technique enables detailed histometric analysis of prostate lesions.
Area of Science:
- Histopathology
- Computational Pathology
- Medical Image Analysis
Background:
- Cribriform glands in prostate tissue are challenging to segment accurately.
- Accurate segmentation is crucial for quantitative histopathology and cancer diagnosis.
Purpose of the Study:
- To develop automated procedures for segmenting cribriform prostatic glands.
- To enable precise histometric characterization of prostate lesions.
Main Methods:
- A knowledge-guided, model-based reasoning approach was employed.
- An expert systems framework for machine vision in histometry was utilized.
- A knowledge file with 78 entities was developed for automated segmentation.
Main Results:
- Achieved 70-80% fully automated, correct segmentation of cribriform glands.
- Segmentation accuracy agreed with visual assessment of histologic components.
- Enabled measurement of gland size, shape, lumen area, epithelial thickness, and cribriformity.
Conclusions:
- The automated procedure facilitates histometric characterization of premalignant and malignant prostate lesions.
- Future work incorporating spectral information may improve segmentation rates.