Computer-assisted breast cancer grading

Area of Science:

• Biomedical Engineering
• Computational Pathology
• Artificial Intelligence in Medicine

Background:

• Histological grading of breast cancer is crucial for prognosis but subjective, leading to inconsistencies.
• Conventional methods for assessing morphological tumor differentiation lack structure and reproducibility.
• Computer-aided analysis offers potential for objective and consistent histological assessment.

Purpose of the Study:

• To develop a structured, computer-aided system for breast cancer histological grading.
• To enhance the accuracy and reproducibility of quantitative morphometric data appraisal.
• To integrate histopathologist expertise with computational methods for improved tumor differentiation analysis.

Main Methods:

• Development of a microcomputer system for collecting and appraising morphometric data.
• Formalization of histological grading methods for computer-aided analysis.
• Application of mathematical signal processing and artificial intelligence techniques.

Main Results:

• A structured framework for histological analysis was developed.
• The system facilitates more accurate and reproducible quantitative assessment of tumor differentiation.
• Potential for improved consistency in breast cancer grading was demonstrated.

Conclusions:

• The developed system addresses limitations of subjective histological grading.
• Computer-aided analysis, utilizing artificial intelligence, enhances accuracy and reproducibility in breast cancer assessment.
• The system has applications in teaching and routine histological analysis, improving prognostic reliability.