What question did this study set out to answer?

To develop a methodology for estimating pleural sliding velocity and assess its diagnostic potential in differentiating lung pathologies.

April 27, 2026Open Access

Novel velocity model for the quantitative characterization of pleural sliding, in vivo multicenter clinical study on RF lung ultrasound data

Key Points

To develop a methodology for estimating pleural sliding velocity and assess its diagnostic potential in differentiating lung pathologies.
Evaluated the novel methodology on two in-vivo datasets with multifrequency RF lung ultrasound data from 79 patients.
Analyzed a total of 536 LUS videos involving patients with cardiogenic pulmonary edema, pneumonia, COPD, and fibrosis.
Utilized binary classifiers to assess the potential of the sliding velocity feature in distinguishing between different lung pathologies.
Pleural sliding velocities ranged from approximately 0.49 mm/s to 4.39 mm/s, remaining within the physiological range.
COPD was associated with the minimum sliding velocity while pneumonia exhibited the maximum average velocity.
Binary classifiers achieved accuracies of up to 94.06% in distinguishing between COPD and pneumonia.

Abstract

Quantitative lung ultrasound (LUS) has emerged as a way to improve current LUS data analysis, based on the visual interpretation of imaging patterns. However, most studies on quantitative LUS have neglected dynamic imaging patterns. Among these, pleural sliding is well-known for its clinical relevance, but few studies have focused on its quantitative characterization. Therefore, in this study, we introduce a novel methodology to estimate the velocity of pleural sliding inspired by functional ultrasound (fUS). Additionally, we assessed, through binary classifiers, the potential of this feature in distinguishing between lung pathologies. We applied the methodology to two in-vivo datasets, comprising multifrequency RF data acquired from patients affected by cardiogenic pulmonary edema (CPE), pneumonia, COPD, fibrosis, and exacerbated. Specifically, we analyzed a total of 536 multifrequency LUS videos from 79 patients. The resulting sliding velocity values were within the physiological range ( ≈ 0.49 mm/s - ≈ 4.39 mm/s). Among all pathological populations, COPD was characterized by the minimum velocity and pneumonia by the maximum velocity, on average. The statistical analysis highlighted some statistically significant differences, especially when comparing COPD with the other pathologies. These differences were confirmed also with binary classifiers that showed accuracies up to 94.06% (when differentiating between COPD and pneumonia). These results demonstrate the potential of this novel methodology to extract clinically valuable information and highlight its applicability in future diagnostic processes. • We propose a novel methodology to estimate the velocity of pleural sliding, inspired by functional ultrasound (fUS). • The proposed methodology was evaluated on two datasets, comprising five different pathological populations. • The estimated sliding velocities were within the physiological range. • Statistical analysis and binary classification highlighted the diagnostic potential of this feature.

Novel velocity model for the quantitative characterization of pleural sliding, in vivo multicenter clinical study on RF lung ultrasound data

Key Points

Abstract

Cite This Study