A novel abnormality annotation database for COVID-19 affected frontal lung X-rays.

Consistent clinical observations of characteristic findings of COVID-19 pneumonia on chest X-rays have attracted the research community to strive to provide a fast and reliable method for screening suspected patients. Several machine learning algorithms have been proposed to find the abnormalities in the lungs using chest X-rays specific to COVID-19 pneumonia and distinguish them from other etiologies of pneumonia. However, despite the enormous magnitude of the pandemic, there are very few instances of public databases of COVID-19 pneumonia, and to the best of our knowledge, there is no database with annotation of abnormalities on the chest X-rays of COVID-19 affected patients. Annotated databases of X-rays can be of significant value in the design and development of algorithms for disease prediction. Further, explainability analysis for the performance of existing or new deep learning algorithms will be enhanced significantly with access to ground-truth abnormality annotations. The proposed COVID Abnormality Annotation for X-Rays (CAAXR) database is built upon the BIMCV-COVID19+ database which is a large-scale dataset containing COVID-19+ chest X-rays. The primary contribution of this study is the annotation of the abnormalities in over 1700 frontal chest X-rays. Further, we define protocols for semantic segmentation as well as classification for robust evaluation of algorithms. We provide benchmark results on the defined protocols using popular deep learning models such as DenseNet, ResNet, MobileNet, and VGG for classification, and UNet, SegNet, and Mask-RCNN for semantic segmentation. The classwise accuracy, sensitivity, and AUC-ROC scores are reported for the classification models, and the IoU and DICE scores are reported for the segmentation models.

Comparison of Elastography Point Quantification with Transient Elastography in Patients with Chronic Viral Hepatitis and Nonalcoholic Fatty Liver Disease: A Pilot Study.

AIMS: The objective of this study was to compare diagnostic accuracy of elastography point quantification (ElastPQ) with transient elastography (TE) and liver histology for measuring liver stiffness in patients with chronic viral hepatitis (CVH) and nonalcoholic fatty liver disease (NAFLD). METHODS: Thirty-two patients with chronic liver disease (CVH and NAFLD) were evaluated by ElastPQ and TE within 7 days of liver biopsy. Within the CVH group, subgroup analysis was carried out in patients with end-stage renal disease (ESRD) and without ESRD. Area under the receiver operating characteristic (AUROC) curves were calculated for ElastPQ and TE. RESULTS: There were 15 patients with CVH and 17 patients with NAFLD. In the CVH group, there were 8 patients with ESRD and 7 patients without ESRD. Taking liver histopathology as the gold standard, liver stiffness measurement by ElastPQ (ρ = 0.826;P < 0.0001) and TE (ρ = 0.649; P < 0.0001) correlated significantly with the stage of fibrosis. AUROCs of ElastPQ and TE for the diagnosis of any fibrosis (F ≥ 1), significant fibrosis (F ≥ 2), and advanced fibrosis (F ≥ 3) were 0.907, 0.959, 0.926 and 0.870, 0.770, 0.881, respectively, in both CVH and NAFLD groups. However, the accuracy of both these techniques was poor in patients with CVH and ESRD (AUROCs for ElastPQ and TE of 0.667 and 0.167 for the diagnosis of significant fibrosis, respectively, and 0.429 and 0.143 for the diagnosis of advanced fibrosis, respectively). The diagnostic accuracy of both ElastPQ and TE for detecting significant fibrosis was excellent in patients with NAFLD (AUROC of 1.000 and 0.936, respectively). ElastPQ was superior to TE in the diagnosis of significant fibrosis in the combined analysis (P = 0.0149) and in the CVH group (P = 0.0391), while both modalities were comparable in patients of the NAFLD group (P = 0.2539). CONCLUSION: ElastPQ may be equally accurate as Fibroscan, and large prospective studies are required to validate the same.

Chest MRI Using Multivane-XD, a Novel T2-Weighted Free Breathing MR Sequence.

OBJECTIVE: To compare image quality of free-breathing T2-weighted MultiVane-XD (MVXD) sequence (non-Cartesian k-space filling using radial rectangular blades) with conventional MR sequences (short tau inversion recovery [STIR],balanced true field echo [BTFE], T1 in phase fast field echo [T1 FFE], and T1-fat saturated postgadolinium [T1PG]) in MR imaging of chest. MATERIALS AND METHODS: Twenty-one patients (10 men and 11 women) underwent chest MRI including T2W MVXD, STIR, BTFE (18/21), T1 FFE, T1PG (10/21) sequences at 1.5 T. Two reviewers (A.S.B and M.J. with 20 and 10 years of experience in pulmonary imaging, respectively) evaluated each sequence with respect to overall image quality, image sharpness, definition of mediastinal vessels including the aorta, pulmonary arteries, superior vena cava, intrapulmonary vessels; trachea, main bronchi, intrapulmonary airways; lung-mediastinal interface, pulmonary lesion detection, and artefacts in the upper, middle, and lower third of chest using 5-point scales. No sedation was given. Pairwise comparisons between T2W MVXD and the 4 conventional sequences were made using unpaired student's t test. RESULTS: Mean age of patients was 30.67 years (range: 6-60 years). T2 MVXD showed significantly better overall image quality and sharpness than STIR, T1 FFE, and T1PG (P < 0.01) while it was comparable to BTFE. Mediastinal vessels were significantly better visualized on T2 MVXD as compared to STIR and T1 (P < 0.003). However, BTFE and T1PG were superior to T2 MVXD for visualization of great vessels, SVC, and intrapulmonary vessels (P < 0.01). Visualization of trachea, major bronchi, intrapulmonary airways as well as intrapulmonary lesion detection was significantly better on T2 MVXD images in comparison to any of the other 4 sequences (P < 0.03). Intrapulmonary artifacts were significantly lesser in BTFE images as compared to T2 MVXD (P < 0.01). No significant difference was found between the severity of intrapulmonary artifacts in other MR sequences as compared to T2 MVXD. CONCLUSIONS: By virtue of its better overall image quality, sharpness, superior visualization of mediastinal airways, and lesion detection, T2 MultiVane-XD promises to be a robust addition in the armamentarium of thoracic radiologists.

Pediatric Chest MRI: A Review.

Chest radiographs and CT scans have been the cornerstone of pulmonary imaging given their advantages of being rapid and easily available techniques. However, a significant concern with their use in the pediatric population is the associated ionisation radiation. The use of magnetic resonance imaging (MRI) in pulmonary imaging has lagged behind its adoption in other organ systems. Previously, the lung parenchyma was considered difficult to evaluate by magnetic resonance due to low proton density in the pulmonary tissue, susceptibility artefacts within the lungs, and respiratory motion artefacts. However, in recent years, there have been a multitude of technical advancements to overcome these limitations. MRI can be an excellent radiation-free alternative in patients who require protracted follow-up like in cases such as cystic fibrosis, complicated pneumonias, tuberculosis and mediastinal neoplasms. An added advantage of MRI is that it can provide functional information in addition to the structural information provided by traditional imaging techniques. One of the major reasons of limited use of MRI despite its established utility is the lack of clarity regarding its indications, and a paucity of data on tailored MRI protocols customised to clinical needs. This article aims to review the basic MRI techniques, indications and terminologies used in chest imaging, with special emphasis on imaging findings of common pathologies in the pediatric population.