Ultrasound estimation of pleural effusion in geriatric patients

摘要

The pleural fluid is a viscous substance between the two pleural layers of the lungs, important in the respiratory mechanics. A pleural effusion (PE) is an excess of pleural fluid in the pleural cavity, caused by several pathologies (e.g., pulmonary embolism, cancer, and infections). Because of its incidence, PE identification and quantification are important to lead the clinician in the selection of the optimal therapeutic strategy. PE volume can be estimated by invasive (quantitative) and non-invasive (qualitative) approaches. Invasive methods provide accurate measurement of PE volume and utilize Computed Tomography (CT) imaging and thoracentesis. Non-invasive methods are based on Ultrasound (US) imaging and are able to identify the PE, but do not quantify its volume. Aiming to spare the patients from risks due to X-ray exposure and invasiveness of thoracentesis, a non-invasive approach for the estimation of PE is coveted. The aim of this study is the feasibility assessment of a simple and affordable model for the quantification of PE through US images. Two US scans are performed to measure: the height of PE column (hPEUS) and the area of the effusion (aPEUS) in correspondence of half hPEUS. The proposed model estimates the PE volume (PEVUS) by multiplying hPEUS and aPEUS, based on cylindrical approximation. PEVUS are compared with PE volumes estimated by CT scans (PEVct), achieved within 24 h from the US exam. The model was successfully tested on 7 patients for PE volumes ranging from 70 mL to 550 mL. The good agreement between PEVus and PEVct volumes is witnessed by the slope of the best fitting line (0.988) and the high correlation coefficient (R=0.99). The two measurements are also compared by Bland-Altman plot, which provides a mean of difference of 1.7 mL and limits of agreement of -32.5 mL and +36.0 mL. The findings of this work show that the proposed model can be a valid tool for the estimation of PE volume by means of US exam. An important advantage of this model is the possibility to quantify PE by the measurements of only two parameters. This approach is not time consuming, hence it can be easily implemented in clinical routine.