High-dimensional multinomial multiclass severity scoring of COVID-19 pneumonia using CT radiomics features and machine learning algorithms.

We aimed to construct a prediction model based on computed tomography (CT) radiomics features to classify COVID-19 patients into severe-, moderate-, mild-, and non-pneumonic. A total of 1110 patients were studied from a publicly available dataset with 4-class severity scoring performed by a radiologist (based on CT images and clinical features). The entire lungs were segmented and followed by resizing, bin discretization and radiomic features extraction. We utilized two feature selection algorithms, namely bagging random forest (BRF) and multivariate adaptive regression splines (MARS), each coupled to a classifier, namely multinomial logistic regression (MLR), to construct multiclass classification models. The dataset was divided into 50% (555 samples), 20% (223 samples), and 30% (332 samples) for training, validation, and untouched test datasets, respectively. Subsequently, nested cross-validation was performed on train/validation to select the features and tune the models. All predictive power indices were reported based on the testing set. The performance of multi-class models was assessed using precision, recall, F1-score, and accuracy based on the 4 × 4 confusion matrices. In addition, the areas under the receiver operating characteristic curves (AUCs) for multi-class classifications were calculated and compared for both models. Using BRF, 23 radiomic features were selected, 11 from first-order, 9 from GLCM, 1 GLRLM, 1 from GLDM, and 1 from shape. Ten features were selected using the MARS algorithm, namely 3 from first-order, 1 from GLDM, 1 from GLRLM, 1 from GLSZM, 1 from shape, and 3 from GLCM features. The mean absolute deviation, skewness, and variance from first-order and flatness from shape, and cluster prominence from GLCM features and Gray Level Non Uniformity Normalize from GLRLM were selected by both BRF and MARS algorithms. All selected features by BRF or MARS were significantly associated with four-class outcomes as assessed within MLR (All p values < 0.05). BRF + MLR and MARS + MLR resulted in pseudo-R2 prediction performances of 0.305 and 0.253, respectively. Meanwhile, there was a significant difference between the feature selection models when using a likelihood ratio test (p value = 0.046). Based on confusion matrices for BRF + MLR and MARS + MLR algorithms, the precision was 0.856 and 0.728, the recall was 0.852 and 0.722, whereas the accuracy was 0.921 and 0.861, respectively. AUCs (95% CI) for multi-class classification were 0.846 (0.805-0.887) and 0.807 (0.752-0.861) for BRF + MLR and MARS + MLR algorithms, respectively. Our models based on the utilization of radiomic features, coupled with machine learning were able to accurately classify patients according to the severity of pneumonia, thus highlighting the potential of this emerging paradigm in the prognostication and management of COVID-19 patients.

COVID-19 prognostic modeling using CT radiomic features and machine learning algorithms: Analysis of a multi-institutional dataset of 14,339 patients.

BACKGROUND: We aimed to analyze the prognostic power of CT-based radiomics models using data of 14,339 COVID-19 patients. METHODS: Whole lung segmentations were performed automatically using a deep learning-based model to extract 107 intensity and texture radiomics features. We used four feature selection algorithms and seven classifiers. We evaluated the models using ten different splitting and cross-validation strategies, including non-harmonized and ComBat-harmonized datasets. The sensitivity, specificity, and area under the receiver operating characteristic curve (AUC) were reported. RESULTS: In the test dataset (4,301) consisting of CT and/or RT-PCR positive cases, AUC, sensitivity, and specificity of 0.83 ± 0.01 (CI95%: 0.81-0.85), 0.81, and 0.72, respectively, were obtained by ANOVA feature selector + Random Forest (RF) classifier. Similar results were achieved in RT-PCR-only positive test sets (3,644). In ComBat harmonized dataset, Relief feature selector + RF classifier resulted in the highest performance of AUC, reaching 0.83 ± 0.01 (CI95%: 0.81-0.85), with a sensitivity and specificity of 0.77 and 0.74, respectively. ComBat harmonization did not depict statistically significant improvement compared to a non-harmonized dataset. In leave-one-center-out, the combination of ANOVA feature selector and RF classifier resulted in the highest performance. CONCLUSION: Lung CT radiomics features can be used for robust prognostic modeling of COVID-19. The predictive power of the proposed CT radiomics model is more reliable when using a large multicentric heterogeneous dataset, and may be used prospectively in clinical setting to manage COVID-19 patients.

Evaluation serum levels of Insulin Growth Factor-1 (IGF-1) and its association with clinical parameters in severe COVID-19.

BACKGROUND: Severe coronavirus disease-2019 (COVID-19) is associated with dysregulated immune response and extreme inflammatory injury. Considering the role of insulin growth factor-1 (IGF-1) in immune-mediated and inflammatory reactions, this study was conducted to investigate the IGF-1 contribution to the pathogenesis of severe form of COVID-19. MATERIAL AND METHODS: Sixty-two patients with severe COVID-19 and 52 healthy subjects were enrolled in this study. The serum levels of IGF-1 were measured using a solid-phase enzyme-linked chemiluminescent immunoassay on an Immulite 2000 system (Siemens Healthcare Diagnostics. RESULT: The serum levels of IGF-1 had no significant difference in COVID-19 patients compared to the healthy subjects (p = 0.359). There was a positive correlation between IGF-1 and age in the severe COVID-19 patients, while a negative correlation was observed for the serum levels of IGF-1 and age in the control group (r = 0.364, p = 0.036, r = - 0.536, p = 0.001, respectively). Moreover, IGF-1 was remarkably associated with hypertension, neurogenic disease, shock, and nausea in patients with the severe form of COVID-19 (p = 0.031, p = 0.044, p = 0.01, p = 0.03, respectively). CONCLUSION: Our results pointed to the complex role of IGF-1 in the severe form of COVID-19, and its association with clinical parameters, and some risk factors in the severe form of COVID-19.

Coronavirus disease 2019 in patients with Behcet's disease: a report of 59 cases in Iran.

OBJECTIVES: To present the clinical characteristics, disease course, management, and outcomes of COVID-19 infection in patients with Behcet's disease (BD). METHODS: In this retrospective cohort study, we retrieved BD patients with definite diagnosis of COVID-19 infection. Demographic data, comorbidities, features related both to BD and COVID-19 infection, treatments, and outcomes were collected. Comparisons between patients with or without hospitalization were performed. All statistical analyzes were performed using SPSS version 25. We considered p < 0.05 statistically significant. RESULTS: We identified 61 episodes of COVID-19 infection in 59 BD patients. The prevalence was 0.69%. The median age was 45 years (IQR = 20), and the median disease duration was 162 months (IQR = 195). BD features were similar except for higher rate of arterial involvement and positive pathergy test in infected patients. Thirty-five episodes (62.5%) happened in non-active patients; 39% had a comorbid disease. COVID manifestations were the same as the general population. Flu-like symptoms were the most common (85%), followed by fever (66%), ageusia/anosmia (56%), headache (51%), and pulmonary involvement (48%). There was no change in BD symptoms in 74%. Fifteen patients (25.4%) were hospitalized, and one patient (1.7%) died. Receiving glucocorticoids (p < 0.03) and cytotoxic drugs (p < 0.02) were associated with an increased rate of hospitalization. CONCLUSION: The incidence of COVID-19 infection in BD patients was not higher than general population in Iran. They showed milder form of disease with lower morbidity and mortality rate. Most were on immunosuppressive drugs, or had a comorbidity apart from BD. No significant effect on BD course was shown. Key Points • The incidence of COVID-19 infection in patients with Behcet's disease is not higher. • They showed milder form of infection with lower morbidity and mortality rate. • No significant effect on Behcet's disease course was shown with COVID19 infection. • BD patients can be managed according to the guidelines used for general population.

Tuberculosis under the Influence of COVID-19 Lockdowns: Lessons from Tehran, Iran.

This study investigates the short-term effects of the coronavirus disease 2019 (COVID-19) pandemic lockdown on tracing and detection of tuberculosis (TB) patients in Tehran, Iran. Results of this study have demonstrated that due to the significant decrease in the identification of patients with suspected TB during the COVID-19 outbreak in Tehran, it is imperative that patients with suspected TB be tracked and diagnosed more quickly to make up for some of the decline in TB diagnosis in recent months and to recover lost cases.

Clinical features and risk factors associated with acute respiratory distress syndrome in pregnant women diagnosed with COVID-19: a multi-center case-control study.

OBJECTIVES: The aim of this study was to evaluate differences in clinical features and laboratory parameters in critically ill pregnant women with acute respiratory distress syndrome (ARDS) compared to moderate and severe pregnant women with coronavirus disease-2019 (COVID-19) but without ARDS. METHODS: This was a retrospective multicenter study of all pregnant women with COVID-19 diagnosed with ARDS between February 15, and May 1, 2020 in nine level III maternity centers in Iran (ARDS group). The control COVID-19 pregnant women were selected from 3 of 9 level III maternity centers between March 15 and April 20, 2020. Univariate statistics were used to look at differences between groups. Cluster dendrograms were used to look at the correlations between clinical and laboratory findings in the groups. A value of p <.05 was considered statistically significant. RESULTS: Fifteen COVID-19 infected women with ARDS were compared to 29 COVID-19 positive and ARDS negative control (moderate: (n = 26) 89.7% and severe: (n = 3)10.3%). The mean maternal age (35.6 vs. 29.4 years; p = .002) and diagnosis of chronic hypertension (20.0% vs. 0%, p = .034) were significantly higher in the ARDS group. There was no significant difference between the two groups in their presenting symptoms. The ARDS group had a significantly higher prevalence of tachypnea (66.6% vs. 10.3%, p = .042) and blood oxygen saturation (SpO2) <93% (66.6% vs. 10.3%, p = .004) at presentation. Relative lymphopenia (lymphocyte ratio < 10.5%, 66.6% vs. 17.2%, p = .002), lymphocytes to leukocytes ratio (11.3% vs. 17.7%, p = .010), and neutrophils to lymphocytes ratio (NLR) >7.5 were significantly different between the two groups (all p < .05). CONCLUSION: Our data demonstrate that symptom-based strategies for identifying the critically ill pregnant women with SARS-CoV-2 are insufficient; however, vital signs and laboratory data might be helpful to predict ARDS in critically ill COVID-19 pregnant patients.

Intravenous methylprednisolone pulse as a treatment for hospitalised severe COVID-19 patients: results from a randomised controlled clinical trial.

INTRODUCTION: There are no determined treatment agents for severe COVID-19. It is suggested that methylprednisolone, as an immunosuppressive treatment, can reduce the inflammation of the respiratory system in COVID-19 patients. METHODS: We conducted a single-blind, randomised controlled clinical trial involving severe hospitalised patients with confirmed COVID-19 at the early pulmonary phase of the illness in Iran. The patients were randomly allocated in a 1:1 ratio by the block randomisation method to receive standard care with methylprednisolone pulse (intravenous injection, 250 mg·day-1 for 3 days) or standard care alone. The study end-point was the time of clinical improvement or death, whichever came first. Primary and safety analysis was done in the intention-to-treat (ITT) population. RESULTS: 68 eligible patients underwent randomisation (34 patients in each group) from April 20, 2020 to June 20, 2020. In the standard care group, six patients received corticosteroids by the attending physician before the treatment and were excluded from the overall analysis. The percentage of improved patients was higher in the methylprednisolone group than in the standard care group (94.1% versus 57.1%) and the mortality rate was significantly lower in the methylprednisolone group (5.9% versus 42.9%; p<0.001). We demonstrated that patients in the methylprednisolone group had a significantly increased survival time compared with patients in the standard care group (log-rank test: p<0.001; hazard ratio 0.293, 95% CI 0.154-0.556). Two patients (5.8%) in the methylprednisolone group and two patients (7.1%) in the standard care group showed severe adverse events between initiation of treatment and the end of the study. CONCLUSIONS: Our results suggest that methylprednisolone pulse could be an efficient therapeutic agent for hospitalised severe COVID-19 patients at the pulmonary phase.

Primary Symptoms, Comorbidities, and Outcomes of 431 Hospitalized Patients with Confirmative RT-PCR Results for COVID-19.

This study aimed to evaluate the primary symptoms, comorbidities, and outcomes of inpatients with confirmed reverse transcription-PCR (RT-PCR) for SARS-CoV-2 infection among 2077 suspected/diagnosed cases of COVID-19. Based on the results of Least Absolute Shrinkage and Selection Operator (LASSO) logistic regression, age, and suggestive chest X-ray (CXR) findings for SARS-CoV-2 infection, cardiovascular diseases, diabetes mellitus, chronic lung diseases, and intensive care units admission had significant associations with positive RT-PCR results for COVID-19 infection. Also, the highest area under the curve (AUC) was related to cough (AUC = 0.53, 95% CI: 0.51-0.56), dyspnea (AUC = 0.52, 95% CI: 0.50-0.54), and abnormal CXR (AUC = 0.52, 95% CI: 0.50-0.54), as significant predictors. This study showed that some symptoms including cough and dyspnea, as well as abnormal CXR, could be proper predictors of positive RT-PCR result for SARS-CoV-2 infection. It seems that patients with underlying disease(s), such as cardiovascular diseases, diabetes mellitus, and chronic lung diseases, had a higher probability to have positive RT-PCR for SARS-CoV-2 infection than those with no underlying disease(s).