Automatic deep learning-based pleural effusion classification in lung ultrasound images for respiratory pathology diagnosis

Tsai, Cheng-Lung; Burgt, Jeroen van der; Vukovic, Damjan; Kaur, Nancy; Demi, Libertario; Canty, David; Wang, Andrew; Royse, Alistair; Royse, Colin; Haji, Kavi; Dowling, Jason; Chetty, Girija; Fontanarosa, Davide

doi:10.1016/j.ejmp.2021.02.023

Cited by 44 publications

(42 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…But these studies mainly focused on the intensity-related properties of the pleural line or the distribution of B-lines, and did not consider the pleural line-and B-line-related characteristics simultaneously. Some deep learning based techniques have shown promises in automatic detection of B lines and pleural effusion, binary classification and 4-level scoring system using artificially extracted quantitative parameters, frame-level and video-level ultrasound images or combination of images and clinical information as the input [33,34,36,46,47], but they need a large number of annotated samples and have poor interpretability. As a contrast, the quantitative method we propose aims at characterizing the image patterns of clinical findings and digging out more features from the pleural line and B-lines, which are further combined to provide more comprehensive information for the severity assessment.…”

Section: Discussionmentioning

confidence: 99%

Quantitative Analysis of Pleural Line and B-Lines in Lung Ultrasound Images for Severity Assessment of COVID-19 Pneumonia

Wang

Zhang

et al. 2022

IEEE Trans. Ultrason., Ferroelect., Freq. Contr.

View full text Add to dashboard Cite

Specific patterns of lung ultrasound images are utilized to assess the severity of COVID-19 pneumonia, while such assessment is mainly based on clinicians' qualitative and subjective observations. In this study, we quantitatively analyze the LUS images to assess the severity of COVID-19 pneumonia by characterizing the patterns related to the pleural line and B-lines. 27 patients with COVID-19 pneumonia, including 13 moderate cases, 7 severe cases, and 7 critical cases, are enrolled. Features related to the pleural line, including the thickness (TPL) and roughness of the pleural line (RPL), and the mean (MPLI) and standard deviation (SDPLI) of the pleural line intensities are extracted from the LUS images. Features related to the B-lines, including the number (NBL), accumulated width (AWBL), attenuation coefficient (ACBL), and accumulated intensity (AIBL) of B-lines are also extracted. The correlations of these features with the disease severity are evaluated. The performances of the binary severe / non-severe classification are assessed for each feature and support vector machine (SVM) classifiers with various combinations of features as input. Several features, including the RPL, NBL, AWBL, and AIBL show significant correlations with disease severity (all p < 0.05). The classification performance is optimal by employing the SVM classifier using all the features as input (area under the receiver operating characteristic curve = 0.96, sensitivity = 0.93, specificity = 1). These findings demonstrate that the proposed method may be a promising tool for automatic grading diagnosis and follow-up of patients with COVID-19 pneumonia.

show abstract

Section: Discussionmentioning

confidence: 99%

Quantitative Analysis of Pleural Line and B-Lines in Lung Ultrasound Images for Severity Assessment of COVID-19 Pneumonia

Wang

Zhang

et al. 2022

IEEE Trans. Ultrason., Ferroelect., Freq. Contr.

View full text Add to dashboard Cite

show abstract

“…All major non-contact techniques that can be effectively used for the detection and monitoring of COVID-19 symptoms are described in this paper. Moreover, Table 1 [ [81] , [82] , [83] , [84] , [85] , [86] , [87] , [88] , [89] , [90] , [91] , [92] , [93] ] summarizes studies in which different non-contact sensing technologies are effectively used to monitor an abnormal breathing rate, which is the primary sign of COVID-19. Fig.…”

Section: Intelligent Healthcare Technologymentioning

confidence: 99%

“… Technology Data Results Refs. Camera A total of 12 male and female volunteers were recorded at distinct resolutions 100% accuracy on HD 720 [ 81 ] Near-infrared camera A total of 28 near-infrared videos and 11 with subject were uncovered and partially covered 99.70% accuracy and 88.95% accuracy [ 82 ] Smartphone camera A total of 11 healthy subjects recorded at distinct breathing frequencies 1.43% average median error [ 83 ] Thermal & depth camera Physical activities were recorded by the home exercise bike 100% accuracy approximately [ 84 ] Thermal camera A total of 41 adults and 20 children undergoing elective polysomnography were recorded r=0.94 (correlation between thermal imaging and the contact method) [ 85 ] Ultrasound imaging A total of 1103 images (172 healthy, 277 pneumonia, and 654 COVID-19) 89% accuracy [ 86 ] Ultrasound imaging A total of 623 videos including 99,209 ultrasound images of 70 patients 92.4% and 91.1% accuracy [ 87 ] X-radiation (X-ray) imaging A total of 500 X-ray images in integration with generative adversarial networks 95.2%–97.6% accuracy [ 88 ] X-ray imaging A total of 6432 chest X-ray scan samples 97.97% accuracy [ 89 ] Computerized tomography (CT) scanning A total of 150 CT images containing 53 cases of COVID...…”

Section: Intelligent Healthcare Technologymentioning

confidence: 99%

Machine learning empowered COVID-19 patient monitoring using non-contact sensing: An extensive review

Saeed

Shah

Ahmad

et al. 2022

Journal of Pharmaceutical Analysis

View full text Add to dashboard Cite

The severe acute respiratory syndrome coronavirus 2 (SARS‐CoV‐2), which caused the COVID-19 pandemic, has affected more than 250 million people worldwide. With the recent rise of a new Delta variant, the efficacy of the vaccines has become an important question. The goal of various studies has been to limit the spread of the virus by utilizing wireless sensing technologies to prevent human-to-human interactions, particularly for healthcare workers. In this review paper, we discuss the current literature on invasive/contact and non-invasive/non-contact technologies (including Wi-Fi, RADAR, and software-defined radio) that have been effectively used to detect, diagnose, and monitor human activities and COVID-19 related symptoms, such as irregular respiration. In addition, we focused on cutting-edge machine learning algorithms (such as generative adversarial networks, random forest, multilayer perceptron, support vector machine, extremely randomized trees, and k-nearest neighbors) and their essential role in intelligent healthcare systems. Furthermore, this study highlights the limitations related to non-invasive techniques and prospective research directions.

show abstract

“…From data collected at the Royal Melbourne Hospital, 623 videos of LUS, containing 99,209 ultrasound images of 70 patients were used. In addition, a DL model using a Spatial Transformer Network (STN) for the automatic detection of pleural effusion focusing on COVID-19 was proposed in [ 123 ]. The model was trained using supervised and weakly supervised approaches.…”

Section: Related Workmentioning

confidence: 99%

Pulmonary COVID-19: Learning Spatiotemporal Features Combining CNN and LSTM Networks for Lung Ultrasound Video Classification

Barros

Lacerda

Albuquerque

et al. 2021

Sensors

View full text Add to dashboard Cite

Deep Learning is a very active and important area for building Computer-Aided Diagnosis (CAD) applications. This work aims to present a hybrid model to classify lung ultrasound (LUS) videos captured by convex transducers to diagnose COVID-19. A Convolutional Neural Network (CNN) performed the extraction of spatial features, and the temporal dependence was learned using a Long Short-Term Memory (LSTM). Different types of convolutional architectures were used for feature extraction. The hybrid model (CNN-LSTM) hyperparameters were optimized using the Optuna framework. The best hybrid model was composed of an Xception pre-trained on ImageNet and an LSTM containing 512 units, configured with a dropout rate of 0.4, two fully connected layers containing 1024 neurons each, and a sequence of 20 frames in the input layer (20×2018). The model presented an average accuracy of 93% and sensitivity of 97% for COVID-19, outperforming models based purely on spatial approaches. Furthermore, feature extraction using transfer learning with models pre-trained on ImageNet provided comparable results to models pre-trained on LUS images. The results corroborate with other studies showing that this model for LUS classification can be an important tool in the fight against COVID-19 and other lung diseases.

show abstract

Automatic deep learning-based pleural effusion classification in lung ultrasound images for respiratory pathology diagnosis

Cited by 44 publications

References 22 publications

Quantitative Analysis of Pleural Line and B-Lines in Lung Ultrasound Images for Severity Assessment of COVID-19 Pneumonia

Quantitative Analysis of Pleural Line and B-Lines in Lung Ultrasound Images for Severity Assessment of COVID-19 Pneumonia

Machine learning empowered COVID-19 patient monitoring using non-contact sensing: An extensive review

Pulmonary COVID-19: Learning Spatiotemporal Features Combining CNN and LSTM Networks for Lung Ultrasound Video Classification

Contact Info

Product

Resources

About