[DNA oriented automated malignancy diagnosis and malignancy grading in prostatic cancer].

Area of Science:

• Urology
• Oncology
• Biomedical Engineering

Context:

• Prostate cancer diagnosis and grading are critical for effective treatment planning.
• Accurate automated methods are needed to improve diagnostic efficiency and consistency.

Purpose:

• To evaluate the efficacy of DNA single-cell fluorescence photometry for the automated diagnosis and grading of prostate cancer.
• To assess the statistical accuracy of DNA histogram analysis in distinguishing malignant from benign prostate tissues and in grading cancer severity.

Summary:

• 85 prostate cancer patients and 20 benign prostatic hypertrophy patients underwent DNA single-cell fluorescence photometry.
• Statistical analysis of DNA histograms using univariate description, stepwise linear discriminant analysis, and regression analysis was performed.
• The method achieved 85.1% accuracy for malignancy diagnosis but only 39.5% accuracy for malignancy grading, with notable false-negative and false-positive rates.

Impact:

• This technique shows potential for improving the accuracy of prostate cancer malignancy diagnosis.
• The poor performance in grading necessitates further research and development for automated cancer grading applications.