Diagnostic Performance of General Practitioners in Carotid Plaque Detection Using AI-Enhanced Point-of-Care...

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Point-of-care ultrasound (POCUS) is increasingly used by general practitioners (GPs).
• Limited data exists on Chinese GPs' use of POCUS for carotid plaque detection.
• Artificial intelligence (AI) is being integrated into diagnostic imaging.

Purpose of the Study:

• To evaluate the diagnostic performance of Chinese GPs in detecting carotid plaque using AI-enhanced POCUS.
• To assess the effectiveness of systematic training for GPs in utilizing this technology.
• To explore the application of AI-assisted POCUS in a clinical setting for high-risk populations.

Main Methods:

• A diagnostic accuracy trial involving 7 GPs who received systematic training.
• Recruitment of 169 high-risk patients for carotid plaque screening.
• Evaluation of diagnostic performance using per-patient and per-vessel analyses, including sensitivity, specificity, PPV, NPV, and kappa statistic.

Main Results:

• High diagnostic accuracy was achieved at the per-patient level: sensitivity 0.87, specificity 0.91, PPV 0.89, NPV 0.89.
• Excellent agreement with the gold standard was observed (kappa = 0.78).
• Per-vessel diagnostic performance was comparable to per-patient results.

Conclusions:

• Systematically trained GPs can effectively perform carotid plaque screening using AI-enhanced POCUS.
• AI-enhanced POCUS is a viable tool for clinical workflow integration in high-risk elderly populations.
• This study supports the expanded use of AI-assisted POCUS by primary care physicians.