Pathadin

Pathadin performs quantitative analysis of whole slide histology images to separate glandular and stromal compartments and quantify markers such as Masson's trichrome and Ki67 for prostate carcinoma research.

Key Features:

• Comprehensive Format Support: Supports a wide range of modern whole slide image (WSI) formats for histology analysis.
• Integration of Machine Learning: Incorporates basic and advanced analytical algorithms, including machine learning methods.
• AI-Enhanced Quantification: Applies a U-net model to separate glandular and stromal tissues in prostate carcinoma samples, enabling independent component analysis.
• Prognostic Marker Analysis: Quantifies Masson's trichrome and Ki67 staining and integrates HistomicsTK for Ki67 evaluation.
• Epithelial and Stromal Quantification: Measures epithelial and stromal changes and collagen expression for comparison with ISUP grade.

Scientific Applications:

• Prostate carcinoma histopathology: Quantifies epithelial and stromal compartment changes and correlates those metrics with ISUP grade.
• Stromal biomarker evaluation: Detects statistically significant increases in collagen expression between tumor grades, supporting stromal components as potential diagnostic markers.
• Proliferative activity assessment: Measures Ki67 across sample groups to detect gradual increases and distinguish high- and low-grade carcinomas.

Methodology:

Uses a U-net model for glandular and stromal separation; training included 5 radical prostatectomy samples and subsequent analyses used 83 additional samples (including 11 autopsies); leverages HistomicsTK for Ki67 evaluation.