Algorithms for morphometric measurements on cancer cells in electron microscopy. Pilot tests

Area of Science:

• Quantitative pathology
• Cell biology
• Biostatistics

Background:

• Accurate quantitative data is crucial for understanding neoplastic cell morphology.
• Electron microscopy provides high-resolution imaging for detailed cellular analysis.

Purpose of the Study:

• To develop automated procedures for quantitative morphometric data analysis of neoplastic cells.
• To assess the variability of stereological parameters in breast carcinoma cells.
• To evaluate the utility of correlation analysis for stereological parameters.

Main Methods:

• Development of automated procedures for handling quantitative data from electron micrographs.
• Stereological analysis of neoplastic cells from 3 breast carcinoma samples.
• Coding of each tumor cell using 30 stereological parameters describing subcellular organization.

Main Results:

• Significant variability observed in some stereological parameters, with standard deviations up to 90% of the mean.
• Correlation analysis showed limited value, with significant correlations primarily found between parameters of the same subcellular structures.

Conclusions:

• Automated morphometric analysis can handle complex quantitative data from neoplastic cells.
• High variability in stereological parameters necessitates careful interpretation.
• Correlation analysis is most effective when applied to related subcellular structural parameters.