Research on Classification of COVID-19 Chest X-Ray Image Modal Feature Fusion Based on Deep Learning

Most detection methods of coronavirus disease 2019 (COVID-19) use classic image classification models, which have problems of low recognition accuracy and inaccurate capture of modal features when detecting chest X-rays of COVID-19. This study proposes a COVID-19 detection method based on image modal feature fusion. This method first performs small-sample enhancement processing on chest X-rays, such as rotation, translation, and random transformation. Five classic pretraining models are used when extracting modal features. A global average pooling layer reduces training parameters and prevents overfitting. The model is trained and fine-tuned, the machine learning evaluation standard is used to evaluate the model, and the receiver operating characteristic (ROC) curve is drawn. Experiments show that compared with the classic model, the classification method in this study can more effectively detect COVID-19 image modal information, and it achieves the expected effect of accurately detecting cases.

Download Full-text

COVID-19 detection and disease progression visualization: Deep learning on chest X-rays for classification and coarse localization

10.21203/rs.3.rs-34534/v1 ◽

2020 ◽

Cited By ~ 1

Author(s):

Tahmina Zebin ◽

Shahadate Rezvy ◽

Wei Pang

Keyword(s):

Disease Progression ◽

Detection Accuracy ◽

X Rays ◽

Automated Classification ◽

X Ray ◽

Feature Extractor ◽

Chest X Ray ◽

Activation Mapping ◽

Activation Map

Abstract Chest X-rays are playing an important role in the testing and diagnosis of COVID-19 disease in the recent pandemic. However, due to the limited amount of labelled medical images, automated classification of these images for positive and negative cases remains the biggest challenge in their reliable use in diagnosis and disease progression. We applied and implemented a transfer learning pipeline for classifying COVID-19 chest X-ray images from two publicly available chest X-ray datasets {https://github.com/ieee8023/covid-chestxray-dataset},{https://www.kaggle.com/paultimothymooney/chest-xray-pneumonia}}. The classifier effectively distinguishes inflammation in lungs due to COVID-19 and pneumonia (viral and bacterial) from the ones with no infection (normal). We have used multiple pre-trained convolutional backbones as the feature extractor and achieved an overall detection accuracy of 91.2% , 95.3%, 96.7% for the VGG16, ResNet50 and EfficientNetB0 backbones respectively. Additionally, we trained a generative adversarial framework (a cycleGAN) to generate and augment the minority COVID-19 class in our approach. For visual explanations and interpretation purposes, we visualized the regions of input that are important for predictions and a gradient class activation mapping (Grad-CAM) technique is used in the pipeline to produce a coarse localization map of the highlighted regions in the image. This activation map can be used to monitor affected lung regions during disease progression and severity stages.

Download Full-text

COVID-19 detection and disease progression visualization: Deep learning on chest X-rays for classification and coarse localization

Applied Intelligence ◽

10.1007/s10489-020-01867-1 ◽

2020 ◽

Author(s):

Tahmina Zebin ◽

Shahadate Rezvy

Keyword(s):

Disease Progression ◽

Input Image ◽

Mapping Technique ◽

Detection Accuracy ◽

X Rays ◽

Automated Classification ◽

X Ray ◽

Chest X Ray ◽

Activation Mapping

Abstract Chest X-rays are playing an important role in the testing and diagnosis of COVID-19 disease in the recent pandemic. However, due to the limited amount of labelled medical images, automated classification of these images for positive and negative cases remains the biggest challenge in their reliable use in diagnosis and disease progression. We implemented a transfer learning pipeline for classifying COVID-19 chest X-ray images from two publicly available chest X-ray datasets1,2. The classifier effectively distinguishes inflammation in lungs due to COVID-19 and Pneumonia from the ones with no infection (normal). We have used multiple pre-trained convolutional backbones as the feature extractor and achieved an overall detection accuracy of 90%, 94.3%, and 96.8% for the VGG16, ResNet50, and EfficientNetB0 backbones respectively. Additionally, we trained a generative adversarial framework (a CycleGAN) to generate and augment the minority COVID-19 class in our approach. For visual explanations and interpretation purposes, we implemented a gradient class activation mapping technique to highlight the regions of the input image that are important for predictions. Additionally, these visualizations can be used to monitor the affected lung regions during disease progression and severity stages.

Download Full-text

Design of Accurate Classification of COVID-19 Disease in X-Ray Images Using Deep Learning Approach

Journal of ISMAC - June 2019 ◽

10.36548/jismac.2021.2.006 ◽

2021 ◽

Vol 2 (2) ◽

pp. 132-148

Author(s):

Joy Iong-Zong Chen

Keyword(s):

Transfer Learning ◽

Research Work ◽

Well Being ◽

Classification Model ◽

Classification Method ◽

World Health ◽

Detection Methods ◽

X Rays ◽

X Ray

COVID-19 appears to be having a devastating influence on world health and well-being. Moreover, the COVID-19 confirmed cases have recently increased to over 10 million worldwide. As the number of verified cases increase, it is more important to monitor and classify healthy and infected people in a timely and accurate manner. Many existing detection methods have failed to detect viral patterns. Henceforth, by using COVID-19 thoracic x-rays and the histogram-oriented gradients (HOG) feature extraction methodology; this research work has created an accurate classification method for performing a reliable detection of COVID-19 viral patterns. Further, the proposed classification model provides good results by leveraging accurate classification of COVID-19 disease based on the medical images. Besides, the performance of our proposed CNN classification method for medical imaging has been assessed based on different edge-based neural networks. Whenever there is an increasing number of a class in the training network, the accuracy of tertiary classification with CNN will be decreasing. Moreover, the analysis of 10 fold cross-validation with confusion metrics can also take place in our research work to detect various diseases caused due to lung infection such as Pneumonia corona virus-positive or negative. The proposed CNN model has been trained and tested with a public X-ray dataset, which is recently published for tertiary and normal classification purposes. For the instance transfer learning, the proposed model has achieved 85% accuracy of tertiary classification that includes normal, COVID-19 positive and Pneumonia. The proposed algorithm obtains good classification accuracy during binary classification procedure integrated with the transfer learning method.

Download Full-text

Effective Survey on Detection and Classification of COVID-19 Suspected Individual Using CT scan Images

International Journal of Scientific Research in Computer Science Engineering and Information Technology ◽

10.32628/cseit217339 ◽

2021 ◽

pp. 294-299

Author(s):

Snehal R. Sambhe ◽

Dr. Kamlesh A. Waghmare

Keyword(s):

Ct Scan ◽

X Rays ◽

Rt Pcr ◽

X Ray ◽

Feature Extractor ◽

Chest X Ray ◽

Coronavirus Infection ◽

Aided Diagnosis ◽

Excessive Risk

As insufficient testing kits are available, the development of new testing kits for detecting COVID remains an open vicinity of research. It’s impossible to test each and every patient suffering from coronavirus symptoms using the traditional method i.e. RT-PCR. This test requires more time to produce results and have less sensitivity. Detecting feasible coronavirus infection using chest X-Ray may also assist quarantine excessive risk sufferers while testing results are disclosed. A learning model can be built based on CT scan images or Chest X-rays of individuals with higher accuracy. This paper represents a computer-aided diagnosis of COVID 19 infection bases on a feature extractor by using CNN models.

Download Full-text

Leadless pacemakers: where is the device?

EP Europace ◽

10.1093/europace/euab116.375 ◽

2021 ◽

Vol 23 (Supplement_3) ◽

Author(s):

M Elrefai ◽

C Menexi ◽

P Roberts

Keyword(s):

Heart Diseases ◽

Sinus Node ◽

X Rays ◽

Funding Sources ◽

Safety And Reliability ◽

Chest X Ray ◽

Implantation Sites ◽

The Right ◽

Funding Acknowledgements

Abstract Funding Acknowledgements Type of funding sources: None. Introduction Leadless pacemakers (LPs) were designed to avoid lead-related complications associated with transvenous pacing. To minimise the risk of complications, there is preference towards implanting LPs into the septal aspect of the right ventricle rather than the apex or free wall. The Transcatheter Pacing Study (TPS) and the international post-approval registry demonstrated the safety and reliability of the LP systems in real-world settings. The registry demonstrated that more than half of the LPs were implanted into the septum and most required <2 attempts at deployment. We report a radiological method of defining LP position. Methods We reviewed the first 100 LPs implanted at our centre. Two independent observers who didn’t implant LPs reviewed the patients’ post-implant fluoroscopy images and/or post-implant CXRs when available. The reviewers assessed the devices’ positions in postero-anterior (PA) and/or right anterior oblique (RAO) views based on conventional fluoroscopic criteria for lead position. We used the proposed criteria interchangeably on fluoroscopic images and post implant CXRs (Figure). Differences in classification of device position were resolved by consensus. Results Three experienced operators implanted 100 LPs at our centre. Patients (61% male) 56.6 ± 22.2 years had normal hearts (74%), ischaemic cardiomyopathies (12%), congenital heart diseases (6%), valvular pathologies (5%) and dilated cardiomyopathies (3%). Indications for pacing were symptomatic sinus node dysfunction (36%), followed by high grade atrio-ventricular block (33%), bradyarrhythmia associated with atrial tachyarrhythmias (27%) and other indications for pacing (4%). We had a 100% successful implant rate, 88% required ≤2 attempts and 70% required one attempt. There were no major complications. We were able to classify the site of the LPs implants in a total of 90 patients who had fluoroscopic projections or chest x-rays that would allow us to classify the implant sites. A total of 32 implants were in the apex (35.6%). 28 were in mid-septum (31.1 %), 15 in the apical septum (16.7%), 14 on the septal aspect of the right ventricular inflow (15.5%) and 1 implant (1.1%) in the septum of the RV outflow tract. Conclusion Our proposed method of defining LP position demonstrated that the rate of implants into the true apex at our centre was highly comparable to that of the international registry. It also showed that we had lower rates of implants into the mid-septum in favour of apical septum. There were no pericardial effusions or cardiac perforations resulting from our implant procedures regardless of the site of the implant. We utilised widely used fluoroscopic and chest x-ray criteria for categorisation of the LPs implantation sites. However, a recognised limitation to our analysis is that our findings were not validated using other imaging modalities such as echocardiogram or cardiac computerised tomography (CT). Abstract Figure. Criteria to classify device position

Download Full-text

Ensemble Deep Learning for the Detection of COVID-19 in Unbalanced Chest X-ray Dataset

Applied Sciences ◽

10.3390/app112210528 ◽

2021 ◽

Vol 11 (22) ◽

pp. 10528

Author(s):

Khin Yadanar Win ◽

Noppadol Maneerat ◽

Syna Sreng ◽

Kazuhiko Hamamoto

Keyword(s):

Deep Learning ◽

Area Under Curve ◽

Ensemble Classifier ◽

Saliency Map ◽

X Rays ◽

Automated Classification ◽

Deep Convolutional Neural Networks ◽

X Ray ◽

Chest X Ray

The ongoing COVID-19 pandemic has caused devastating effects on humanity worldwide. With practical advantages and wide accessibility, chest X-rays (CXRs) play vital roles in the diagnosis of COVID-19 and the evaluation of the extent of lung damages incurred by the virus. This study aimed to leverage deep-learning-based methods toward the automated classification of COVID-19 from normal and viral pneumonia on CXRs, and the identification of indicative regions of COVID-19 biomarkers. Initially, we preprocessed and segmented the lung regions usingDeepLabV3+ method, and subsequently cropped the lung regions. The cropped lung regions were used as inputs to several deep convolutional neural networks (CNNs) for the prediction of COVID-19. The dataset was highly unbalanced; the vast majority were normal images, with a small number of COVID-19 and pneumonia images. To remedy the unbalanced distribution and to avoid biased classification results, we applied five different approaches: (i) balancing the class using weighted loss; (ii) image augmentation to add more images to minority cases; (iii) the undersampling of majority classes; (iv) the oversampling of minority classes; and (v) a hybrid resampling approach of oversampling and undersampling. The best-performing methods from each approach were combined as the ensemble classifier using two voting strategies. Finally, we used the saliency map of CNNs to identify the indicative regions of COVID-19 biomarkers which are deemed useful for interpretability. The algorithms were evaluated using the largest publicly available COVID-19 dataset. An ensemble of the top five CNNs with image augmentation achieved the highest accuracy of 99.23% and area under curve (AUC) of 99.97%, surpassing the results of previous studies.

Download Full-text

COVID-19 Classification of X-ray Images Using Deep Neural Networks

10.1101/2020.10.01.20204073 ◽

2020 ◽

Author(s):

Elisha Goldstein ◽

Daphna Keidar ◽

Daniel Yaron ◽

Yair Shachar ◽

Ayelet Blass ◽

...

Keyword(s):

Machine Learning ◽

Neural Networks ◽

Receiver Operating Characteristic ◽

Deep Neural Networks ◽

Operating Characteristic ◽

Learning Model ◽

X Ray ◽

Machine Learning Model ◽

Chest X Ray

AbstractBackgroundIn the midst of the coronavirus disease 2019 (COVID-19) outbreak, chest X-ray (CXR) imaging is playing an important role in the diagnosis and monitoring of patients with COVID-19. Machine learning solutions have been shown to be useful for X-ray analysis and classification in a range of medical contexts.PurposeThe purpose of this study is to create and evaluate a machine learning model for diagnosis of COVID-19, and to provide a tool for searching for similar patients according to their X-ray scans.Materials and MethodsIn this retrospective study, a classifier was built using a pre-trained deep learning model (ReNet50) and enhanced by data augmentation and lung segmentation to detect COVID-19 in frontal CXR images collected between January 2018 and July 2020 in four hospitals in Israel. A nearest-neighbors algorithm was implemented based on the network results that identifies the images most similar to a given image. The model was evaluated using accuracy, sensitivity, area under the curve (AUC) of receiver operating characteristic (ROC) curve and of the precision-recall (P-R) curve.ResultsThe dataset sourced for this study includes 2362 CXRs, balanced for positive and negative COVID-19, from 1384 patients (63 +/- 18 years, 552 men). Our model achieved 89.7% (314/350) accuracy and 87.1% (156/179) sensitivity in classification of COVID-19 on a test dataset comprising 15% (350 of 2326) of the original data, with AUC of ROC 0.95 and AUC of the P-R curve 0.94. For each image we retrieve images with the most similar DNN-based image embeddings; these can be used to compare with previous cases.ConclusionDeep Neural Networks can be used to reliably classify CXR images as COVID-19 positive or negative. Moreover, the image embeddings learned by the network can be used to retrieve images with similar lung findings.SummaryDeep Neural Networks and can be used to reliably predict chest X-ray images as positive for coronavirus disease 2019 (COVID-19) or as negative for COVID-19.Key ResultsA machine learning model was able to detect chest X-ray (CXR) images of patients tested positive for coronavirus disease 2019 with accuracy of 89.7%, sensitivity of 87.1% and area under receiver operating characteristic curve of 0.95.A tool was created for finding existing CXR images with imaging characteristics most similar to a given CXR, according to the model’s image embeddings.

Download Full-text

Chest X-ray classification using Deep learning for automated COVID-19 screening

10.1101/2020.06.21.20136598 ◽

2020 ◽

Cited By ~ 1

Author(s):

Ankita Shelke ◽

Madhura Inamdar ◽

Vruddhi Shah ◽

Amanshu Tiwari ◽

Aafiya Hussain ◽

...

Keyword(s):

Deep Learning ◽

Early Stage ◽

Ground Glass Opacity ◽

Classification Model ◽

X Rays ◽

X Ray ◽

The People ◽

History Of ◽

Chest X Ray

AbstractIn today’s world, we find ourselves struggling to fight one of the worst pandemics in the history of humanity known as COVID-2019 caused by a coronavirus. If we detect the virus at an early stage (before it enters the lower respiratory tract), the patient can be treated quickly. Once the virus reaches the lungs, we observe ground-glass opacity in the chest X-ray due to fibrosis in the lungs. Due to the significant differences between X-ray images of an infected and non-infected person, artificial intelligence techniques can be used to identify the presence and severity of the infection. We propose a classification model that can analyze the chest X-rays and help in the accurate diagnosis of COVID-19. Our methodology classifies the chest X-rays into 4 classes viz. normal, pneumonia, tuberculosis (TB), and COVID-19. Further, the X-rays indicating COVID-19 are classified on severity-basis into mild, medium, and severe. The deep learning model used for the classification of pneumonia, TB, and normal is VGG16 with an accuracy of 95.9 %. For the segregation of normal pneumonia and COVID-19, the DenseNet-161 was used with an accuracy of 98.9 %. ResNet-18 worked best for severity classification achieving accuracy up to 76 %. Our approach allows mass screening of the people using X-rays as a primary validation for COVID-19.

Download Full-text

A Method to Discriminate Pulmonary Contusion Severity Through Analysis of Hounsfield Unit Frequency

ASME 2007 Summer Bioengineering Conference ◽

10.1115/sbc2007-176906 ◽

2007 ◽

Author(s):

F. Scott Gayzik ◽

Melissa Daly ◽

Joel Stitzel

Keyword(s):

Three Dimensional ◽

Hounsfield Unit ◽

Pulmonary Contusion ◽

Thoracic Injury ◽

Strong Positive Correlation ◽

X Rays ◽

Novel Approach ◽

Chest X Ray ◽

Predicting Outcome

This study presents a novel approach for the quantification and classification of pulmonary contusion (PC). PC is a common thoracic injury, affecting up to 25% of patients sustaining blunt chest trauma. [1] Contusion volume at the time of hospitalization has been shown to be an independent predictor for the development of Acute Respiratory Distress Syndrome (ARDS), with the risk of ARDS increasing sharply with PC in excess of 20% by volume. [1] Despite the frequency of the injury and strong positive correlation between contusion volume and outcome, there are relatively few contusion quantification methods in the current literature. One such study utilized chest x-ray film to score PC according the amount of lung appearing to be damaged. [2] The study concluded that despite the limitations in using chest x-rays, a PC scoring system may be of value in determining the need for ventilator assistance and predicting outcome. A potentially more accurate approach to quantifying the severity of PC is through the use of computed tomography (CT) chest scans. CT is the preferred modality for obtaining volumetric pulmonary contusion data since the complete three-dimensional lung anatomy is captured. In this work a semi-automated approach is used to analyze PC in an isolated model of lung contusion in the rat. [3, 4] The CT-based approach enables the PC to be precisely quantified as the lesion progresses in time. The technique distinguishes the severity of the contusion by analyzing the composition of bands in the Hounsfield Unit (HU) range of lung image masks.

Download Full-text

Simple method for detecting idiopathic interstitial pneumonias by measuring vertical lung length on chest X-ray

Scientific Reports ◽

10.1038/s41598-021-87452-z ◽

2021 ◽

Vol 11 (1) ◽

Author(s):

Masato Karayama ◽

Yoichiro Aoshima ◽

Hideki Yasui ◽

Hironao Hozumi ◽

Yuzo Suzuki ◽

...

Keyword(s):

Operating Characteristic ◽

Characteristic Curve ◽

Area Under The Curve ◽

X Rays ◽

Simple Method ◽

X Ray ◽

Idiopathic Interstitial Pneumonias ◽

Chest X Ray ◽

Costophrenic Angle ◽

Tracheal Carina

AbstractDetection of idiopathic interstitial pneumonias (IIPs) on chest X-ray is difficult for non-specialist physicians, especially in patients with mild IIPs. The current study aimed to evaluate the usefulness of a simple method for detecting IIPs by measuring vertical lung length (VLL) in chest X-rays to quantify decreased lung volume. A total of 280 consecutive patients with IIPs were randomly allocated to exploratory and validation cohorts, and 140 controls were selected for each cohort by propensity score-matching. Upper (uVLL; from apex to tracheal carina), lower (lVLL; from carina to costophrenic angle), and total VLL (tVLL; from apex to costophrenic angle), and the l/uVLL ratio were measured on chest X-rays. Patients in the exploratory cohort had significantly decreased uVLL, lVLL, tVLL, and l/uVLL ratio compared with controls (all p < 0.001). Receiver operating characteristic curve analyses demonstrated that lVLL (area under the curve [AUC] 0.86, sensitivity 0.65, specificity 0.92), tVLL (AUC 0.83, sensitivity 0.75, specificity 0.80), and l/uVLL ratio (AUC 0.80, sensitivity 0.72, specificity 0.79) had high diagnostic accuracies for IIPs. These results were reproduced in the validation cohort. IIP patients thus have decreased VLLs, and measurements of VLL may thus aid the accurate detection of IIPs.

Download Full-text