Ultrasound Estimation of Pleural Effusion in the Sitting Patients

The aim of this study is the feasibility assessment of a simple and affordable model for the quantification of Pleural Effusion through thoracic Ultra Sounds images. Two US scans will be performed to measure: the height of Pleural Effusion column (hPEUS) and the area of the effusion in correspondence of the midline of hPEUS (aPEUS). The proposed model will estimate the Pleural effusion volume (PEVUS) by multiplying hPEUS and aPEUS. PEVUS will be compared with volumes estimated by CT scans (PEVCT), obtained within 24h from the US examination.

The pleural fluid is the thin film of serous fluid between the visceral and parietal pleurae, whose physiological value is around 20 mL in healthy adults. It plays crucial role in the respiratory mechanics, as allows the pleurae to slide effortlessly against each other during ventilation, and its surface tension leads to close apposition of the lung surfaces with the chest wall. The abnormal collection of fluid in the pleural cavity is defined pleural effusion (PE). The most common causes of PE in adults are congestive heart failure, liver cirrhosis, pneumonia, malignant pleural disease, pulmonary embolism, and gastrointestinal disease. PE represents the 10% of admissions in pulmonary units, and affects about 10 million people each year in industrialized countries. The high incidence demands methods for PE accurate estimation, in order to guide the clinician in the choice of the adequate therapy and its follow up . Strategies for the estimation of PE can be divided into: i) qualitative methods, which are non-invasive; and ii) quantitative methods, which are invasive. Qualitative methods provide coarse estimation of PE (minimal, small, moderate and massive PE), and usually ultrasonography (US) is devoted for this aim; quantitative approaches give accurate information about PE volume, at the expenses of the invasiveness, because of the need of X-ray, CT imaging and thoracentesis. The interest in the use of US for the evaluation of chest diseases, especially for the study of bedridden, critically ill patients increased. In fact, US-based methods present various advantages: i) absence of ionizing radiation; ii) noninvasiveness; iii) they can be performed at the bedside; iv) being inexpensive, can be repeated if necessary; v) short examination time when compared with CT-based methods. Moreover, US methods are particularly sensitive in imaging the chest wall, pleura, and pleural space thanks to their superficial locations, and are often used to detect PE and guide thoracentesis and drainage, especially in minimal effusions. Some authors proposed approaches for the estimation of PE volume by means of US images. The PE is identified like an anechoic area on the US image. The basic idea is to measure characteristic lengths or surfaces, and to correlate them to the volume. Such approaches can be distinguished into linear and planar. In the linear ones, usually one length (e.g., the high of the PE column, or intrapleural distance) measured on US image is correlated to the PE volume. These methods are assessed by comparing their output with the values obtained by the gold standard (CT, or thoracentesis). The pro of these techniques related to the quick evaluation is restrained by the low accuracy of the estimation. In the planar approaches, volumes are directly calculated by multiplying a specific length (e.g., the height of the PE column) with the effusion areas measured in correspondence of defined anatomical landmarks (e.g., the half height of PE column). For instance, the model proposed by Remérand et al. requires three measurements: 2 for the detection of the PE column extremities, and 1 for the area. The US exam is performed on supine patients, at the end-expiration. Such approach is more time-consuming than the linear one, but allows achieving more accurate results. The present study implements a simple and affordable planar model, based on the product between the height of the PE column and the area of the surface at half PE column. Such landmarks can be easily detected by the clinician during the exam. The model requires only two US measurements, performed on seated patients. Volumes estimated by US images are compared with volumes calculated from the CT scans of the patient (gold standard), acquired within the 24 hours. Materials and Methods: Images collection In this retrospective study, US images and CT scans will be collected from patients, hospitalized at the Campus Bio Medico Teaching Hospital in Rome. Thoracic and total body CT scans will be performed for clinical reasons related to patients care. The consent to participate a clinical study will be obtained. The CT scans will be collected within 24 hours from the US exam, in order to avoid significant differences due to spontaneous resorption of PE. Model development Bedside pleural US imaging will be performed using ECM-EXAGYNE with an abdominal convex 3.5 MHz probe. US exam will be performed with the seating patients and the US measurements will be collected at the end-expiration apnea. The pleural cavity will be explored on frontal plane to detect the maximal height of PE column (hPEUS). In correspondence of the midline of hPEUS, the probe will be rotated of 90° on transverse plane, and the PE area (aPEUS) will be detected. hPEUS and aPEUS will be manually collected. Thoracic CT scan CT scan will be performed using Somatom Sensation 64 (Siemens) with tube voltage of 120 kVp, tube current of 460mA and B30f as convolution kernel. The 3.0 mm-thick contiguous sections of the whole lung will be acquired during a prolonged expiratory pause. CT data will be stored on computerized disks and subsequently analyzed using OsiriX software (v7.0.2). PE volume will be quantified through a manual delineation of cross-sectional PE area. PEVCT will be computed by the software with the sum of the total number of pixels present in all PE cross-sectional area delineated. The following analysis will be performed: 1. Assessment of the intra-observer reproducibility on PEVCT 2. Correlation between PE volumes estimated with US and CT