An Explainable Two Stage Deep Learning Framework for Pericoronitis Assessment in Panoramic Radiographs Using YOLOv8 and ResNet-50

AI can help dentists detect gum disease around wisdom teeth more reliably: The system achieved 86% accuracy in identifying pericoronitis (inflammation around impacted teeth) on dental X-rays, with the ability to detect normal cases at 88% accuracy, suggesting it could serve as a useful second opinion tool for radiologists.

Explainability is crucial for doctors to trust AI recommendations: When the system highlighted which parts of X-rays it used to make decisions (using a technique called Grad-CAM), radiologists agreed with these highlighted regions 84% of the time, demonstrating that showing why an AI makes a diagnosis—not just what it decides—builds clinical confidence.

Two-stage AI systems can break down complex diagnostic tasks: Rather than trying to diagnose pericoronitis directly, the system first identified wisdom teeth and their positions (92% accuracy), then analyzed those specific regions for disease signs, showing that structured AI approaches may be more reliable than single-step analysis.

AI-assisted diagnosis doesn't replace radiologists but supports them: The paper frames this as an "AI-assisted assessment system" rather than an autonomous diagnostic tool, suggesting the responsible approach is to use AI to enhance human expertise rather than substitute for it in clinical decision-making.