Related Experiment Video
Updated: Aug 14, 2026
05:33
Introduction of an Integrated Pathology Image Management, Artificial Intelligence, and Reporting System
Published on: July 11, 2025
Artificial Intelligence-Based Histopathological Analysis to Assist Pathologists in Diagnosing Ewing Sarcoma and
Francisco Giner1,2, Álvaro Pastor-Naranjo3, Pablo Meseguer3
1Pathology Department, University of Valencia, 46010 Valencia, Spain.
International Journal of Molecular Sciences
|August 13, 2026
Summary
Artificial intelligence (AI) aids in diagnosing Ewing sarcoma (ES) and similar tumors by analyzing histopathology images. This AI tool accurately classifies tumor types, supporting pathologists and improving diagnostic accuracy for challenging cases.
Area of Science:
- Oncology
- Pathology
- Artificial Intelligence
Background:
- Ewing sarcoma diagnosis is challenging due to overlapping histological features with other round-cell tumors.
- Accurate diagnosis is critical, especially with limited biopsy material, requiring preservation for molecular testing.
- Artificial intelligence (AI) offers a promising approach for histopathological image analysis in oncology.
Purpose of the Study:
- To evaluate AI-based histopathological analysis for classifying Ewing sarcoma and its mimics.
- To assess the AI model's ability to distinguish between diagnostically challenging tumor entities based on morphology.
- To determine the AI's potential as a diagnostic adjunct for pathologists.
Main Methods:
- Analysis of 1926 digitized histological cores from 729 patients using tissue microarrays (TMAs).
- Dataset included Ewing sarcoma (ES), rhabdomyosarcoma (RMS), chondrosarcoma (CHS), gastrointestinal stromal tumor (GIST), and synovial sarcoma (SS).
- Development of a weakly supervised multiple-instance learning (MIL) framework with transformer-based aggregation.
Main Results:
- The AI model achieved high classification accuracies: 97.1% for ES, 80.0% for RMS, 85.7% for GIST, 80.0% for CHS, and 76.0% for SS.
- Overall classification accuracy was 91.6%, with no misclassifications between Ewing sarcoma and rhabdomyosarcoma.
- The AI demonstrated robustness in distinguishing challenging tumor entities based on histopathological features.
Conclusions:
- AI-based histopathological analysis shows strong potential as a diagnostic adjunct for classifying Ewing sarcoma and related tumors.
- The AI model supports clinical decision-making by recognizing morphologic patterns, complementing molecular diagnostics.
- Integration of AI with existing techniques can improve diagnostic accuracy and patient management.
