ABSTRACT
Detecting COVID-19 in computed tomography (CT) or radiography images has been proposed as a supplement to the RT-PCR test. We compare slice-based (2D) and volume-based (3D) approaches to this problem and propose a deep learning ensemble, called IST-CovNet, combining the best 2D and 3D systems with novel preprocessing and attention modules and the use of a bidirectional Long Short-Term Memory model for combining slice-level decisions. The proposed ensemble obtains 90.80% accuracy and 0.95 AUC score overall on the newly collected IST-C dataset in detecting COVID-19 among normal controls and other types of lung pathologies; and 93.69% accuracy and 0.99 AUC score on the publicly available MosMedData dataset that consists of COVID-19 scans and normal controls only. The system also obtains state-of-art results (90.16% accuracy and 0.94 AUC) on the COVID-CT-MD dataset which is only used for testing. The system is deployed at Istanbul University Cerrahpasa School of Medicine where it is used to automatically screen CT scans of patients, while waiting for RT-PCR tests or radiologist evaluation.
ABSTRACT
INTRODUCTION: We aimed to evaluate clinical and laboratory findings of hospitalized asthma and chronic obstructive pulmonary disease (COPD) patients with COVID-19 and demonstrate that they have different symptoms and/or laboratory results and outcomes than COVID-19 patients with comorbidity (CoV-com) and without comorbidity (CoV-alone). METHODOLOGY: The data of the demographic, clinical, laboratory findings of hospitalized CoV-alone, asthma, COPD patients with COVID-19 (CoV-asthma, CoV-COPD, respectively), and CoV-com were analyzed. RESULTS: Out of 1082 patients hospitalized for COVID-19, 585 (54.1%) had CoV-alone, 40 (3.7%) had CoV-asthma, 46 (4.3%) had CoV-COPD and 411 (38%) had CoV-com. Cough, shortness of breath, fever and weakness were the most common four symptoms seen in all COVID-19 patients. Shortness of breath, myalgia, headache symptoms were more common in CoV-asthma than the other groups (p < 0.001, p < 0.01, p < 0.05 respectively). Sputum was more common in CoV-COPD than other groups (p < 0.01). COPD group most frequently had increased values, different from the other groups with CRP>5ng/mL in 91.3%, D-dimer > 0.05mg/dL in 89.1%, troponin > 0.014micg/L in %63.9, INR>1.15 in 52.2%, CK-MB>25U/L in 48.5%, PT>14s in 40.9% of patients (p < 0.05, p < 0.001, p < 0.001, p < 0.001, p < 0.05, p < 0.001, respectively). NT-ProBNP was found to have the highest AUC value and the best differentiating parameter for CoV-asthma from CoV-alone. Typical CT findings were present in 44.4% of CoV-alone, 57.5% of CoV-asthma, 28.3% of CoV-COPD and 38.9% of CoV-com groups. CoV-COPD and CoV-com patients died more frequently than other groups (17.8%, 18.5%). CONCLUSIONS: CoV-asthma and CoV-COPD patients might have different symptoms and laboratory parameters than other COVID-19 patients which can guide the physicians.