China
Africa
Explore chapters and articles related to this topic
Amiodarone pulmonary toxicity
Published in Philippe Camus, Edward C Rosenow, Drug-induced and Iatrogenic Respiratory Disease, 2010
Philippe Camus, Thomas V Colby, Edward C Rosenow
History-taking, the presence of dyspnoea when lying flat, the presence or absence of cardiomegaly or valvular heart disease, a history of definite heart failure, the diuresis test, cardiac ultrasound, review of previous imaging and pulmonary function and, in selected cases, cardiac catheterization with measurement of the pulmonary capillary wedge pressure should make it possible to weigh the respective likelihoods of hydrostatic pulmonary oedema or APT.42 If the pulmonary infiltrates clear suboptimally following a course of diuretics, continued consideration of amiodarone-induced lung disease is warranted.58 Other competing diagnoses that need to be particularly ruled out in this setting include interstitial lung disease of other cause, including those due to cardiovascular drugs, an infection, aspiration pneumonia, pulmonary embolism/infarction, idiopathic organizing-, non-specific- or eosinophilic interstitial pneumonia, exogenous lipoid pneumonia and, less commonly, bronchioloalveolar carcinoma or pulmonary lymphoma. With regard to cardiovascular drugs, patients may develop adverse pulmonary reactions to such drugs as aspirin, beta-blockers, ACEIs, oral or intravenous anticoagulants, platelet glycoprotein IIb/IIIa inhibitors (abciximab, clopidogrel), HMG-CoA reductase inhibitors (statins), hydrochlorothiazide, thiazolinediones, flecainide, procainamide and tocainide, in the form of interstitial lung disease, eosinophilic pneumonia, organizing pneumonia, alveolar haemorrhage or pulmonary oedema, with few features that might definitely separate these adverse reactions from APT.7
A Detailed Study on AI-Based Diagnosis of Novel Coronavirus from Radiograph Images
Published in Chhabi Rani Panigrahi, Bibudhendu Pati, Mamata Rath, Rajkumar Buyya, Computational Modeling and Data Analysis in COVID-19 Research, 2021
Malaya Kumar Nath, Aniruddha Kanhe
Current technology makes use of artificial intelligence (AI) to effectively handle healthcare issues and complications, such as breast cancer (Shen et al. 2019), skin cancer (Keerthana and Nath 2020), and brain tumor detection (Ismael et al. 2019; Mittal et al. 2019). Deep learning (DL) techniques have triggered much interest in their application to the medical imaging domain, as they reveal detailed image features that are not possible from original images. Convolutional neural networks (CNNs) are found to be favorable in a large group of research communities for classification and detection of various pathologies from radiograph images. DL techniques are popular due to the availability of deep CNN, which achieves good performance in some applications. This technique suffers due to the lack of a large quantity of training data. Transfer learning overcomes the issue by utilizing the knowledge acquired during the training process and retraining the deep CNN with a smaller amount of data. Chouhan et al. (2020) make use of this concept for pneumonia detection in chest X-ray images using different pre-trained models such as AlexNet, DenseNet121, InceptionV3, ResNet18, and GoogleNet on ImageNet dataset. Gu et al. (2018) have used deep 19-layers VGGNet and 22-layers GoogleNet for the diagnosis of bacterial or viral pneumonia from chest radiographs. Wang et al. (2017) have used AlexNet, GoogleNet, VGG, and ResNet-50 pre-trained models on ChestX-ray8 database to diagnose eight commonly occurring thoracic pathologies: atelectasis, cardiomegaly, effusion, infiltration, mass, nodule, pneumonia, and pneumathorax. Rajpukar et al. (2018) have developed CheXNeXt with a 121-layer DenseNet architecture to identify 14 types of thoracic diseases (atelectasis, cardiomegaly, consolidation, edema, effusion, emphysema, fibrosis, hernia, infiltration, mass, nodule, pleural thickening, pneumonia, and pneumothorax) from the largest publicly available ChestX-ray14 dataset. AlexNet and GoogleNet have been used by Lakhani and Sundaram (2017) to identify pulmonary TB or normal. Yadav and Jadhav (2019) use InceptionV3 and VGG16 models for training and support vector machine (SVM) for classification on pneumonia data.
Cardiovascular system
Published in David A Lisle, Imaging for Students, 2012
Abnormal pulmonary vascular patterns are outlined below:Pulmonary venous hypertensionBlood vessels in the upper lobes are larger than those in the lower lobes on erect CXRAssociated with cardiac failure and mitral valve disease, and often accompanied on CXR by cardiomegaly, pulmonary oedema and pleural effusionPulmonary arterial hypertension (Fig. 3.4)Bilateral hilar enlargement due to enlarged proximal pulmonary arteries with rapid decrease in the calibre of peripheral vessels (‘pruning’)Associated with long-standing pulmonary disease, including emphysema, multiple recurrent pulmonary emboli, left-to-right shunts (ventricular septal defect (VSD), atrial septal defect (ASD), patent ductus artery (PDA))Pulmonary plethoraIncreased size and number of pulmonary vesselsIncreased pulmonary blood flow caused by left-to-right cardiac shunts (VSD, ASD, PDA)Pulmonary oligaemiaGeneral lucency (blackness) of lungs with decreased size and number of pulmonary vessels and small main pulmonary arteriesReduced pulmonary blood flow associated with pulmonary stenosis/atresia, Fallot tetralogy, tricuspid atresia, Ebstein anomaly and severe emphysema.
R-Peak Identification in ECG Signals using Pattern-Adapted Wavelet Technique
Published in IETE Journal of Research, 2023
L. V. Rajani Kumari, Y. Padma Sai, N. Balaji
Electrocardiogram (ECG), the recording electrical activity of the heart, is prominent in identifying various heart diseases and heart conditions such as arrhythmias, cardiomegaly, hyperkalaemia, and myocardial infarction [1,2]. A huge amount of medical information can be determined by calculating amplitudes and the durations pertaining to the ECG signals [3,4]. R-peak identification in ECG signal plays a vital role in computing the heart rate and arrhythmia detection [5,6]. Accurate R-peak detection is a difficult task because of the non-stationary nature of ECG signal. Researchers have been applying different combinations of various signal processing techniques to the ECG in order to identify R-peaks accurately and enable further analyses.