Respiratory infections are among the leading causes of consultation in primary care and emergency settings and account for substantial healthcare resource use. Their management relies heavily on chest imaging, particularly radiography and, more recently, lung ultrasound. Artificial intelligence (AI) applied to these modalities offers new opportunities to improve anomaly detection, reduce interobserver variability, and support clinical decision-making. This article summarizes recent evidence on the contribution of AI in thoracic imaging for the diagnosis of respiratory infections, particularly pneumonia and tuberculosis, while highlighting current limitations and challenges related to validation and implementation.
Pasquale et al. (Thu,) studied this question.