Voxel-level forecast system for lesion development in patients with COVID-19 (preprint)

The global spread of COVID-19 seriously endangers human health and even lives. By predicting patients' individualized disease development and further performing intervention in time, we may rationalize scarce medical resources and reduce mortality. Based on 1337 multi- stage ([≥]3) high-resolution chest computed tomography (CT) images of 417 infected patients from three centers in the epidemic area, we proposed a random forest + cellular automata (RF+CA) model to forecast voxel-level lesion development of patients with COVID-19. The model showed a promising prediction performance (Dice similarity coefficient [DSC] = 71.1%, Kappa coefficient = 0.612, Figure of Merit [FoM] = 0.257, positional accuracy [PA] = 3.63) on the multicenter dataset. Using this model, multiple driving factors for the development of lesions were determined, such as distance to various interstitials in the lung, distance to the pleura, etc. The driving processes of these driving factors were further dissected and explained in depth from the perspective of pathophysiology, to explore the mechanism of individualized development of COVID-19 disease. The complete codes of the forecast system are available at https://github.com/keyunj/VVForecast_covid19.

GASNet: Weakly-supervised Framework for COVID-19 Lesion Segmentation (preprint)

Segmentation of infected areas in chest CT volumes is of great significance for further diagnosis and treatment of COVID-19 patients. Due to the complex shapes and varied appearances of lesions, a large number of voxel-level labeled samples are generally required to train a lesion segmentation network, which is a main bottleneck for developing deep learning based medical image segmentation algorithms. In this paper, we propose a weakly-supervised lesion segmentation framework by embedding the Generative Adversarial training process into the Segmentation Network, which is called GASNet. GASNet is optimized to segment the lesion areas of a COVID-19 CT by the segmenter, and to replace the abnormal appearance with a generated normal appearance by the generator, so that the restored CT volumes are indistinguishable from healthy CT volumes by the discriminator. GASNet is supervised by chest CT volumes of many healthy and COVID-19 subjects without voxel-level annotations. Experiments on three public databases show that when using as few as one voxel-level labeled sample, the performance of GASNet is comparable to fully-supervised segmentation algorithms trained on dozens of voxel-level labeled samples.

Epidemiological,clinical and radiological findings in medical staff with COVID-19 in Wuhan, China: a single-centered, retrospective cohort study (preprint)

Backgrounds In December 2019, a pneumonia associated with the severe acute respiratory syndrome coronavirus 2 (SARS-Cov-2) emerged in Wuhan city, China. As of 20 Feb 2020, a total of 2,055 medical staff infected with SARS-Cov-2 in China had been reported. The predominant cause of the infection and the failure of protection among medical staff remains unclear. We sought to explore the epidemiological, clinical characteristics and prognosis of novel coronavirus-infected medical staff.Methods Medical staff who infected with SARS-Cov-2 and admitted to Union Hospital, Wuhan between 16 Jan, 2020 to 25 Feb, 2020 were included retrospectively. Epidemiological, clinical and radiological data were compared by occupation and analyzed with the Kaplan-Meier and Cox regression methods.Results A total of 101 medical staff (32 males and 69 females; median age: 33 years old) were included in this study and 74% were nurses. None had an exposure to Huanan seafood wholesale market or wildlife. A small proportion of the cohort had contact with specimens (3%) as well as patients infected with SARS-Cov-2 in fever clinics (15%) and isolation wards (3%). 80% of medical staff showed abnormal IL-6 levels and 33% had lymphocytopenia. Chest CT mainly manifested as bilateral (62%), septal/subpleural (77%) and ground­glass opacities (48%). The major differences between doctors and nurses manifested in laboratory indicators. As of the last observed date, no patient was transferred to intensive care unit or died, and 98 (97%) had been discharged. Fever (HR=0.57; 95% CI 0.36-0.90) and IL-6 levels greater than >2.9 pg/ml (HR=0.50; 95% CI 0.30-0.86) on admission were unfavorable factors for discharge.Conclusions Our findings suggested that the infection of medical staff mainly occurred at the early stages of SARS-CoV-2 epidemic in Wuhan, and only a small proportion of infection had an exact mode. Meanwhile, medical staff infected with COVID-19 have relatively milder symptoms and favorable clinical course than other ordinary patients, which may be partly due to their medical expertise, younger age and less underlying diseases. The potential risk factors of presence of fever and IL-6 levels greater than >2.9 pg/ml could help to identify medical staff with poor prognosis at an early stage.

Prognostic Value of Baseline Clinical and HRCT Findings in 101 Patients with Severe COVID-19 in Wuhan, China (preprint)

Background It has been reported that patients with severe Coronavirus disease 2019 (COVID-19) had high mortality rate, but the information about prognostic factors for these patients is unknown. The aim of this study was to assess the prognostic value of baseline clinical and high resolution CT (HRCT) findings in patients with severe COVID-19.Methods In this retrospective, two-center study, we included two groups of inpatients with severe COVID-19 who had been discharged or had died in Jin Yin-tan hospital and Wuhan union hospital between January 5, 2020, and February 22, 2020. Cases were confirmed by real-time polymerase chain reaction. Demographic, clinical, laboratory data, and HRCT data were collected and compared between discharged and deceased patients. Univariable and multivariable logistic regression models were used to assess predictors of mortality risk in these patients.Results 101 patients were included in this study, of whom 66 were discharged and 35 died in the hospital. The mean age was 56.6 ± 15.1 years and 67 (66.3%)were men. Of the 101 patients, hypertension (38, 37.6%), cardiovascular disease (21,20.8%), diabetes (18,17.8%), and chronic pulmonary disease (16,15.8%) were the most common coexisting conditions. The multivariable regression analysis showed older age (OR:1.142, 95%CI:1.059–1.231, p༜0.001), acute respiratory distress syndrome (ARDS) (OR:10.142, 95%CI:1.611–63.853, p = 0.014), reduced lymphocyte count (OR:0.004, 95%CI: 0.001–0.306, p = 0.013), and elevated HRCT score (OR:1.276, 95%CI:1.002–1.625, p = 0.049) to be independent predictors of mortality risk on admission in severe COVID-19 patients.Conclusions These initial data indicate that older age, ARDS, lymphocytopenia and elevated HRCT score on admission were strong predictors of mortality risk in severe COVID-19 patients.

iCTCF: an integrative resource of chest computed tomography images and clinical features of patients with COVID-19 pneumonia (preprint)

The outbreak of coronavirus disease 2019 (COVID-19) caused by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was initially reported in Wuhan, China since December, 2019. Here, we reported a timely and comprehensive resource named iCTCF to archive 256,356 chest computed tomography (CT) images, 127 types of clinical features (CFs), and laboratory-confirmed SARS-CoV-2 clinical status from 1170 patients, reaching a data volume of 38.2 GB. To facilitate COVID-19 diagnosis, we integrated the heterogeneous CT and CF datasets, and developed a novel framework of Hybrid-learning for UnbiaSed predicTion of COVID-19 patients (HUST-19) to predict negative cases, mild/regular and severe/critically ill patients, respectively. Although both CT images and CFs are informative in predicting patients with or without COVID-19 pneumonia, the integration of CT and CF datasets achieved a striking accuracy with an area under the curve (AUC) value of 0.978, much higher than that when exclusively using either CT (0.919) or CF data (0.882). Together with HUST- 19, iCTCF can serve as a fundamental resource for improving the diagnosis and management of COVID-19 patients.Authors Wanshan Ning, Shijun Lei, Jingjing Yang, and Yukun Cao contributed equally to this work.

Development and Evaluation of an AI System for COVID-19 (preprint)

Early detection of COVID-19 based on chest CT will enable timely treatment of patients and help control the spread of the disease. With rapid spreading of COVID-19 in many countries, however, CT volumes of suspicious patients are increasing at a speed much faster than the availability of human experts. We proposed an artificial intelligence (AI) system for fast COVID-19 detection and performed extensive statistical analysis of CTs of COVID-19 based on the AI system. We developed and evaluated our system on a large dataset with more than 10 thousand CT volumes from COVID-19, influenza-A/B, non-viral community acquired pneumonia (CAP) and non-pneumonia subjects. In such a difficult multi-class diagnosis task, our deep convolutional neural network-based system is able to achieve an area under the receiver operating characteristic curve (AUC) of 97.17%, a sensitivity of 90.19%, and a specificity of 95.76% for COVID-19 on internal test cohort of 3,203 scans and AUC of 97.77% on the publicly available CC-CCII database with 1,943 test samples. In a reader study involving five radiologists, the AI system outperforms all of radiologists in more challenging tasks at a speed of two orders of magnitude above them. Diagnosis performance of chest x-ray (CXR) is compared. Detailed interpretation of deep network is also performed to relate AI results with CT findings. The code is available at https://github.com/ChenWWWeixiang/diagnosis_covid19.