Background Chest CT is used for diagnosis of 2019 novel coronavirus disease (COVID-19), as an important complement to the reverse-transcription polymerase chain reaction (RT-PCR) tests. Purpose To ...
Objectives The purpose of this study was to observe the imaging characteristics of the novel coronavirus pneumonia. Methods Sixty-three confirmed patients were enrolled from December 30, 2019 to January 31, 2020. High-resolution CT (HRCT) of the chest was performed. The number of affected lobes, ground glass nodules (GGO), patchy/punctate ground glass opacities, patchy consolidation, fibrous stripes and irregular solid nodules in each patient's chest CT image were recorded. Additionally, we performed imaging follow-up of these patients. Results CT images of 63 confirmed patients were collected. M/F ratio: 33/30. The mean age was 44.9 ± 15.2 years. The mean number of affected lobes was 3.3 ± 1.8. Nineteen (30.2%) patients had one affected lobe, five (7.9%) patients had two affected lobes, four (6.3%) patients had three affected lobes, seven (11.1%) patients had four affected lobes while 28 (44.4%) patients had 5 affected lobes. Fifty-four (85.7%) patients had patchy/punctate ground glass opacities, 14 (22.2%) patients had GGO, 12 (19.0%) patients had patchy consolidation, 11 (17.5%) patients had fibrous stripes and 8 (12.7%) patients had irregular solid nodules. Fifty-four (85.7%) patients progressed, including single GGO increased, enlarged and consolidated; fibrous stripe enlarged, while solid nodules increased and enlarged. Conclusions Imaging changes in novel viral pneumonia are rapid. The manifestations of the novel coronavirus pneumonia are diverse. Imaging changes of typical viral pneumonia and some specific imaging features were observed. Therefore, we need to strengthen the recognition of image changes to help clinicians to diagnose quickly and accurately. Key Points • High-resolution CT (HRCT) of the chest is critical for early detection, evaluation of disease severity and follow-up of patients with the novel coronavirus pneumonia. • The manifestations of the novel coronavirus pneumonia are diverse and change rapidly. • Radiologists should be aware of the various features of the disease and temporal changes.
Purpose To quantitatively evaluate lung burden changes in patients with COVID-19 using serial CT scan by an automated deep learning method. Materials and Methods Patients with COVID-19 who underwen...
The recovery process of COVID-19 patients is unclear. Some recovered patients complain of continued shortness of breath. Vasculopathy has been reported in COVID-19, stressing the importance of probing microstructure and function of lungs at the alveolar-capillary interface. While CT detects structural abnormalities, little is known about the impact of disease on lung function. 129Xe MRI is a technique uniquely capable of assessing ventilation, microstructure and gas exchange. Using 129Xe MRI, we found COVID-19 patients have higher ventilation defects percentage (5.9% vs 3.7%), unchanged microstructure, longer gas-blood exchange time (43.5 ms vs 32.5 ms), and reduced RBC/TP (0.279 vs 0.330) compared with healthy subjects. These findings suggest regional ventilation and alveolar airspace dimensions are relatively normal around the time of discharge, while gas-blood exchange function is diminished. This study establishes the feasibility of localized lung function measurement in COVID-19 patients. Such readouts could be useful as a supplement to structural imaging.
Non-invasive prediction of isocitrate dehydrogenase (IDH) genotype plays an important role in tumor glioma diagnosis and prognosis. Recently, research has shown that radiology images can be a potential tool for genotype prediction, and fusion of multi-modality data by deep learning methods can further provide complementary information to enhance prediction accuracy. However, it still does not have an effective deep learning architecture to predict IDH genotype with three-dimensional (3D) multimodal medical images. In this paper, we proposed a novel multimodal 3D DenseNet (M3D-DenseNet) model to predict IDH genotypes with multimodal magnetic resonance imaging (MRI) data. To evaluate its performance, we conducted experiments on the BRATS-2017 and The Cancer Genome Atlas breast invasive carcinoma (TCGA-BRCA) dataset to get image data as input and gene mutation information as the target, respectively. We achieved 84.6% accuracy (area under the curve (AUC) = 85.7%) on the validation dataset. To evaluate its generalizability, we applied transfer learning techniques to predict World Health Organization (WHO) grade status, which also achieved a high accuracy of 91.4% (AUC = 94.8%) on validation dataset. With the properties of automatic feature extraction, and effective and high generalizability, M3D-DenseNet can serve as a useful method for other multimodal radiogenomics problems and has the potential to be applied in clinical decision making.
Background: Computed tomography–derived fractional flow reserve (CT-FFR) using on-site machine learning enables identification of both the presence of coronary artery disease and vessel-specific ischemia. However, it is unclear whether on-site CT-FFR improves clinical or economic outcomes when compared with the standard of care in patients with stable coronary artery disease. Methods: In total, 1216 patients with stable coronary artery disease and an intermediate stenosis of 30% to 90% on coronary computed tomographic angiography were randomized to an on-site CT-FFR care pathway using machine learning or to standard care in 6 Chinese medical centers. The primary end point was the proportion of patients undergoing invasive coronary angiography without obstructive coronary artery disease or with obstructive disease who did not undergo intervention within 90 days. Secondary end points included major adverse cardiovascular events, quality of life, symptoms of angina, and medical expenditure at 1 year. Results: Baseline characteristics were similar in both groups, with 72.4% (881/1216) having either typical or atypical anginal symptoms. A total of 421 of 608 patients (69.2%) in the CT-FFR care group and 483 of 608 patients (79.4%) in the standard care group underwent invasive coronary angiography. Compared with standard care, the proportion of patients undergoing invasive coronary angiography without obstructive coronary artery disease or with obstructive disease not undergoing intervention was significantly reduced in the CT-FFR care group (28.3% [119/421] versus 46.2% [223/483]; P <0.001). Overall, more patients underwent revascularization in the CT-FFR care group than in the standard care group (49.7% [302/608] versus 42.8% [260/608]; P =0.02), but major adverse cardiovascular events at 1 year did not differ (hazard ratio, 0.88 [95% CI, 0.59–1.30]). Quality of life and symptoms improved similarly during follow-up in both groups, and there was a trend towards lower costs in the CT-FFR care group (difference, –¥4233 [95% CI, –¥8165 to ¥973]; P =0.07). Conclusions: On-site CT-FFR using machine learning reduced the proportion of patients with stable coronary artery disease undergoing invasive coronary angiography without obstructive disease or requiring intervention within 90 days, but increased revascularization overall without improving symptoms or quality of life, or reducing major adverse cardiovascular events. Registration: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT03901326.