Related Experiment Video
Updated: Sep 3, 2025
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
215
Evaluation of a Probability-Based Predictive Tool on Pathologist Agreement Using Urinary Bladder as a Pilot Tissue.
Emily Jones1, Solomon Woldeyohannes1, Fernanda Castillo-Alcala2
1School of Veterinary Science, The University of Queensland, Gatton, QLD 4343, Australia.
Veterinary Sciences
|July 25, 2022
Summary
Veterinary pathologists showed significant variation in diagnosing bladder conditions. A new predictive tool improved agreement, highlighting the need for standardized data and high-quality whole-slide imaging in pathology.
Area of Science:
- Veterinary Pathology
- Digital Pathology
- Computational Pathology
Background:
- Inter-pathologist variation is a known issue in veterinary pathology, often worsened by incomplete case information.
- Discrepancies in interpreting inflammatory cell thresholds contribute to reduced agreement among pathologists.
Purpose of the Study:
- To evaluate the impact of a predictive modeling tool on inter-pathologist agreement for bladder histology assessments.
- To compare agreement using digital whole-slide images (WSI) versus glass slides, with and without clinical information.
Main Methods:
- Four pathologists assessed bladder histology slides under three conditions: no information, with signalment/history, and with a predictive tool.
- Assessments were repeated using both digital WSI and traditional glass slides.
- Kappa agreement statistics were calculated to quantify inter-pathologist variation.
Main Results:
- Initial agreement without information was low (kappa 7-37%).
- Agreement improved with signalment/history (kappa 23-37%) and further with the predictive tool (kappa 31-42%).
- Overall concurrence with the reference diagnosis was 60%.
Conclusions:
- Predictive modeling shows potential for standardizing pathologist agreement in veterinary pathology.
- High-quality WSI is crucial to minimize digitization effects on diagnostic agreement.
- Standardization of tissue assessment and data quality is essential for improving veterinary diagnostic accuracy.
