Related Experiment Video
Updated: Jun 26, 2025
05:49
Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
Published on: February 23, 2024
830
Oropharyngeal Cancer Staging Health Record Extraction Using Artificial Intelligence.
Elif Baran1, Melissa Lee1, Steven Aviv2
1Department of Otolaryngology-Head and Neck Surgery, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, Ontario, Canada.
JAMA Otolaryngology-- Head & Neck Surgery
|May 16, 2024
Summary
Artificial intelligence shows promise for improving oropharyngeal cancer staging accuracy. The AI engine achieved fair to excellent results for tumor, nodal, and metastatic staging, aiding treatment decisions.
Area of Science:
- Oncology
- Medical Informatics
- Artificial Intelligence
Background:
- Accurate oropharyngeal cancer staging is vital for prognosis and treatment but often suffers from documentation inaccuracies.
- Artificial intelligence (AI) offers potential for more efficient and accurate cancer staging through automated data abstraction.
Purpose of the Study:
- To evaluate an AI algorithm for extracting essential information from medical records to assign tumor, nodal, and metastatic stages for oropharyngeal cancer.
- To assess the AI's accuracy based on the American Joint Committee on Cancer (AJCC) eighth edition guidelines.
Main Methods:
- A retrospective study of 806 patients with oropharyngeal squamous cell carcinoma.
- Development of a ground truth dataset and staging rules.
- Training of four distinct AI models for tumor, nodal, metastasis, and p16 status.
- Comparison of AI-derived stages against expert-established ground truth.
Main Results:
- The AI engine achieved accuracies of 55.9% for tumor, 56.0% for nodal, and 87.6% for metastatic staging.
- p16 status was accurately identified with 92.1% accuracy.
- Differentiation between localized and advanced cancers reached 80.7% accuracy.
Conclusions:
- The AI algorithm demonstrated fair to excellent accuracy for oropharyngeal cancer staging, particularly for metastatic disease and p16 status.
- The findings suggest clinical relevance for optimizing treatment and reducing toxicity.
- Further refinement and external validation are recommended for improved accuracy and clinical applicability.
