(A) new approach to lung ultrasound diagnosis

Abstract

The attempt to diagnose the lung, an organ directly associated with life, is underway using ultrasound, which has the advantage of real-time imaging. However, the air in the lung poses a challenge as ultrasound cannot penetrate the tissue and is strongly reflected by the pleural. Lung diagnosis using ultrasound relies on indirectly utilizing the reverberation artifacts created by reflections occurring at the pleural. In this paper, we aim to automatically detect vertical reverberation artifacts present in lungs affected by lesions, thereby automating the lung scores employed in clinical tests. Additionally, we propose a quantitative lung ultrasound by utilizing the ultrasound RF signals that give rise to reverberation artifacts, quantifying lung damage based on air volume using air porosity. The proposed system demonstrates the potential of using ultrasound to diagnose the lungs quantitatively, presenting clinical feasibility.