Improving CBCT diagnosis with pathological descriptors in radiology training.

Area of Science:

• Oral and Maxillofacial Radiology
• Medical Imaging Education

Background:

• Cone-beam computed tomography (CBCT) is crucial for diagnosing oral pathologies.
• Standardized interpretation and diagnostic accuracy remain areas for improvement in radiology training.

Purpose of the Study:

• To evaluate the impact of pathological descriptor selection on CBCT interpretation accuracy in Oral Radiology residents.
• To identify which descriptor categories are most predictive of accurate differential diagnoses.

Main Methods:

• Three Oral Radiology residents interpreted 200 CBCT cases with known pathologies.
• Residents selected descriptors from 101 options and provided differential diagnoses, compared against a gold standard.
• Statistical analyses (Spearman's Rho, Kruskal-Wallis) assessed correlations (P < .05).

Main Results:

• Residents demonstrated significant improvement in descriptor selection and diagnostic accuracy (P < .05).
• Descriptor categories like location, size, and dentoalveolar involvement strongly predicted diagnostic accuracy (Rho > 0.6).
• Shape and border descriptors showed weaker correlations with diagnostic outcomes.

Conclusions:

• Training focused on pathological descriptors effectively enhances CBCT diagnostic accuracy in radiology residents.
• Targeted training based on descriptor performance can refine diagnostic skills.
• This methodology provides a scalable framework for radiology education and clinical practice.