Ultrasound in point-of-care lung assessment is becoming increasingly relevant. This is further reinforced in the context of the COVID-19 pandemic, where rapid decisions on the lung state must be made for staging and monitoring purposes. The lung structural changes due to severe COVID-19 modify the way ultrasound propagates in the parenchyma. This is reflected by changes in the appearance of the lung ultrasound images. In abnormal lungs, vertical artifacts known as B-lines appear and can evolve into white lung patterns in the more severe cases. Currently, these artifacts are assessed by trained physicians, and the diagnosis is qualitative and operator dependent. In this article, an automatic segmentation method using a convolutional neural network is proposed to automatically stage the progression of the disease. 1863 B-mode images from 203 videos obtained from 14 asymptomatic individual,14 confirmed COVID-19 cases, and 4 suspected COVID-19 cases were used. Signs of lung damage, such as the presence and extent of B-lines and white lung areas, are manually segmented and scored from zero to three (most severe). These manually scored images are considered as ground truth. Different test-training strategies are evaluated in this study. The results shed light on the efficient approaches and common challenges associated with automatic segmentation methods.
Although X-Ray Computed Tomography (CT) is widely used for detecting pulmonary nodules inside the parenchyma, it cannot be used during video-assisted surgical procedures. Real-time, non-ionizing, ultrasound-based techniques are an attractive alternative for nodule localization to ensure safe resection margins during surgery. Conventional ultrasound B-mode imaging of the lung is challenging due to multiple scattering. However, the multiple scattering contribution can be exploited to detect regions inside the lung containing no scatterers. Pulmonary nodules are homogeneous regions in contrast to the highly scattering parenchyma containing millions of air-filled alveoli. We developed a method relying on mapping the multiple scattering contribution inside the highly scattering lung to detect and localize pulmonary nodules. Impulse response matrices were acquired in ex-vivo pig and dog lungs using a linear array transducer to semi-locally investigate the backscattered field. Extracting the multiple-scattering contribution using singular-value decomposition and combining it with a depression detection algorithm allowed us to detect and localize regions with less multiple scattering, associated with the nodules. The feasibility of this method was demonstrated in five ex-vivo lungs containing a total of 20 artificial nodules. Ninety-five percent of the nodules were detected. Nodule depth and diameter significantly correlated with their ex-vivo CT-estimated counterparts (R = 0.960, 0.563, respectively).
Quantitative ultrasound methods based on the backscatter coefficient (BSC) and envelope statistics have been used to quantify disease in a wide variety of tissues, such as prostate, lymph nodes, breast, and thyroid. However, to date, these methods have not been investigated in the lung. In this study, lung properties were quantified by BSC and envelope statistical parameters in normal, fibrotic, and edematous rat lungs in vivo. The average and standard deviation of each parameter were calculated for each lung as well as the evolution of each parameter with acoustic propagation time within the lung. The transport mean free path and backscattered frequency shift, two parameters that have been successfully used to assess pulmonary fibrosis and edema in prior work, were evaluated in combination with the BSC and envelope statistical parameters. Multiple BSC and envelope statistical parameters were found to provide contrast between control and diseased lungs. BSC and envelope statistical parameters were also significantly correlated with fibrosis severity using the modified Ashcroft fibrosis score as the histological gold standard. These results demonstrate the potential for BSC and envelope statistical parameters to improve the diagnosis of pulmonary fibrosis and edema as well as monitor pulmonary fibrosis.
Although CT is widely used for detecting pulmonary nodules inside the parenchyma, the accurate real time lesion detection during video-assisted surgical procedures using ultrasound-based techniques is very attractive to improve to resection margins. Imaging the parenchyma using conventional B-mode is impossible due to multiple scattering in lung. However, the multiple scattering contribution of ultrasonic waves can be exploited, to detect the regions inside the lung with no scatterers. Nodules are homogeneous regions with relatively uniform properties compared to the healthy heterogeneous parenchyma containing millions of alveoli. We took advantage of this and developed an algorithm to extract multiple scattering contributions inside the highly scattering lung to localize pulmonary nodules. Inter-element response matrices were acquired using translated sections of a linear array transducer to semi-locally investigate the backscattered field. Extracting the multiple-scattering contribution using singular value decomposition and combining it with a depression detection algorithm allowed to detect regions with less multiple scattering, associated with the nodules. We validated this method in lung phantoms and demonstrated their feasibility in ex vivo pig and dog lungs containing artificial Vaseline nodules. We evaluated 7 lung blocks (4 pigs, 3 dogs) with multiple nodules and have localized all nodules in the last 4 lungs.
Using ultrasound for point-of-care lung assessment is becoming more and more relevant. Today, the outbreak of the coronavirus disease 2019 (COVID-19) has spread over the world at a very high rate. The most severe cases of COVID-19 are associated with lung damage such as ground-glass opacities and areas of lung consolidation, leading to acute respiratory distress. During the COVID-19 pandemic the need to detect and monitor the lung state is critical. Changes in the COVID-19 lung structure modify the way ultrasound propagates in the lung and are reflected by changes in the appearance of lung ultrasound images. Vertical artifacts known as B-lines appear and can evolve into white lung patterns in the more severe cases. Currently, these artifacts are assessed by trained physicians and sonographers, and the diagnosis is qualitative and operator dependent. We propose an automatic segmentation method using a convolutional neural network, to automatically stage the progression of the disease and predict the severity of the lung damage. By classifying the images based on illness severity we can define different scores—from healthy lung to most severe case—and produce a reliable tool to establish severity of COVID-19.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.