Background COVID-19 has overwhelmed health systems worldwide. It is important to identify severe cases as early as possible, such that resources can be mobilized and treatment can be escalated. Objective This study aims to develop a machine learning approach for automated severity assessment of COVID-19 based on clinical and imaging data. Methods Clinical data—including demographics, signs, symptoms, comorbidities, and blood test results—and chest computed tomography scans of 346 patients from 2 hospitals in the Hubei Province, China, were used to develop machine learning models for automated severity assessment in diagnosed COVID-19 cases. We compared the predictive power of the clinical and imaging data from multiple machine learning models and further explored the use of four oversampling methods to address the imbalanced classification issue. Features with the highest predictive power were identified using the Shapley Additive Explanations framework. Results Imaging features had the strongest impact on the model output, while a combination of clinical and imaging features yielded the best performance overall. The identified predictive features were consistent with those reported previously. Although oversampling yielded mixed results, it achieved the best model performance in our study. Logistic regression models differentiating between mild and severe cases achieved the best performance for clinical features (area under the curve [AUC] 0.848; sensitivity 0.455; specificity 0.906), imaging features (AUC 0.926; sensitivity 0.818; specificity 0.901), and a combination of clinical and imaging features (AUC 0.950; sensitivity 0.764; specificity 0.919). The synthetic minority oversampling method further improved the performance of the model using combined features (AUC 0.960; sensitivity 0.845; specificity 0.929). Conclusions Clinical and imaging features can be used for automated severity assessment of COVID-19 and can potentially help triage patients with COVID-19 and prioritize care delivery to those at a higher risk of severe disease.

Objective: Patients with diabetes mellitus, diabetic nephropathy (DN) and healthy donor were analyzed to test whether the early DN patients can be detected using both blood oxygenation level dependent (BOLD) and diffusion tensor imaging. Methods: This study was approved by the Ethics Committee of our hospital. MR images were acquired on a 3.0-Tesla MR system (Discovery MR750, General Electric, Milwaukee, WI). 30 diabetic patients were divided into NAU (normal to mildly increased albuminuria, N = 15) and MAU (moderately increased albuminuria, N = 15) group based on the absence or presence of microalbuminuria. 15 controls with sex- and age-matched were enrolled in the study. Prior to MRI scan, all participants were instructed to collect their fresh morning urine samples for quantitative measurement of urinary microalbumin and urinary creatinine. Then, the estimations of serum creatinine, serum uric acid, HbAlc and fasting plasma glucose as well as fundus examinations were performed in all subjects. Then, the values of albumin–creatinine ratio (ACR) and estimated glomerular filtration rate were also calculated. All subjects underwent renal diffusion tensor imaging (DTI) and BOLD acquisition after fasting for 4 h. Regions of interest were placed in renal medulla and cortex for evaluating apparent diffusion coefficient (ADC), fractional anisotropy (FA) and R2* values by two experienced radiologists. The consistency between the two observations was estimated using intragroup correlation coefficients. To test differences in ADC, FA and R2* values across the three groups, the data were analyzed using separate one-way ANOVAs. Post-hoc pair wise comparisons were then performed using t-test. To investigate the clinical relevance of imaging parameters in both regions across the three groups, the correlations of values of the ACR/estimated glomerular filtration rate and of the ADC/FA/R2* were calculated. Results: There was a high level of consistency of those ADC, FA and R2* values across the three groups on both renal cortex and medulla measured by the two doctors. The FA value of medulla in MAU group was lower than that in control (p < 0.01). The R2* value of medulla in the NAU group was higher than that in the control (p < 0.01), and the R2* value of medulla in the MAU group was lower than that in the control (p = 0.009) . Moreover, the current study revealed a decreasing trend in FA values of the renal medulla from the control group to NAU and MAU groups. Finally, a weak negatively correlation between medullary R2* and ACR was found in current study. Conclusion: Medullary R2* value might be a new more sensitive predictor of early DN. Meanwhile, BOLD imaging detected the medullary hypoxia at the simply diabetic stage, while DTI didn’t identify the medullary directional diffusion changes at this stage. Based on our assumption mentioned above, it’s presumable that BOLD imaging may be more sensitive for assessment of the early renal function changes than DTI. These imaging techniques are more accurate and practical than conventional tests. Advances in knowledge: Non-invasive MRI was used to detect renal function changes at early DN stage.

IVIM DWI might be helpful in noninvasively identifying early-stage DN. The IVIM parametric values are more sensitive than the ACR in detecting early-stage kidney changes.

Automatic severity assessment and progression prediction can facilitate admission, triage, and referral of COVID-19 patients. This study aims to explore the potential use of lung lesion features in the management of COVID-19, based on the assumption that lesion features may carry important diagnostic and prognostic information for quantifying infection severity and forecasting disease progression. A novel LesionEncoder framework is proposed to detect lesions in chest CT scans and to encode lesion features for automatic severity assessment and progression prediction. The LesionEncoder framework consists of a U-Net module for detecting lesions and extracting features from individual CT slices, and a recurrent neural network (RNN) module for learning the relationship between feature vectors and collectively classifying the sequence of feature vectors. Chest CT scans of two cohorts of COVID-19 patients from two hospitals in China were used for training and testing the proposed framework. When applied to assessing severity, this framework outperformed baseline methods achieving a sensitivity of 0.818, specificity of 0.952, accuracy of 0.940, and AUC of 0.903. It also outperformed the other tested methods in disease progression prediction with a sensitivity of 0.667, specificity of 0.838, accuracy of 0.829, and AUC of 0.736. The LesionEncoder framework demonstrates a strong potential for clinical application in current COVID-19 management, particularly in automatic severity assessment of COVID-19 patients. This framework also has a potential for other lesion-focused medical image analyses. NOTE: This manuscript has been submitted to the Medical Image Analysis Special Issue on Intelligent Analysis of COVID-19 Imaging Data on the 20th of May, 2020. This version is submitted version, last updated on the 20th of May. Its current status with Medical Image Analysis is still "under review" (last accessed on 2nd of August, 2020).

No abstract

Automatic severity assessment and progression prediction can facilitate admission, triage, and referral of COVID-19 patients. This study aims to explore the potential use of lung lesion features in the management of COVID-19, based on the assumption that lesion features may carry important diagnostic and prognostic information for quantifying infection severity and forecasting disease progression. A novel LesionEncoder framework is proposed to detect lesions in chest CT scans and to encode lesion features for automatic severity assessment and progression prediction. The LesionEncoder framework consists of a U-Net module for detecting lesions and extracting features from individual CT slices, and a recurrent neural network (RNN) module for learning the relationship between feature vectors and collectively classifying the sequence of feature vectors. Chest CT scans of two cohorts of COVID-19 patients from two hospitals in China were used for training and testing the proposed framework. When applied to assessing severity, this framework outperformed baseline methods achieving a sensitivity of 0.818, specificity of 0.952, accuracy of 0.940, and AUC of 0.903. It also outperformed the other tested methods in disease progression prediction with a sensitivity of 0.667, specificity of 0.838, accuracy of 0.829, and AUC of 0.736. The LesionEncoder framework demonstrates a strong potential for clinical application in current COVID-19 management, particularly in automatic severity assessment of COVID-19 patients. This framework also has a potential for other lesion-focused medical image analyses.

Objectives This study aims to develop a machine learning approach for automated severity assessment of COVID-19 patients based on clinical and imaging data. Materials and Methods Clinical data, including demographics, signs, symptoms, comorbidities and blood test results and chest CT scans of 346 patients from two hospitals in the Hubei province, China, were used to develop machine learning models for automated severity assessment of diagnosed COVID-19 cases. We compared the predictive power of clinical and imaging data by testing multiple machine learning models, and further explored the use of four oversampling methods to address the imbalance distribution issue. Features with the highest predictive power were identified using the SHAP framework. Results Targeting differentiation between mild and severe cases, logistic regression models achieved the best performance on clinical features (AUC:0.848, sensitivity:0.455, specificity:0.906), imaging features (AUC:0.926, sensitivity:0.818, specificity:0.901) and the combined features (AUC:0.950, sensitivity:0.764, specificity:0.919). The SMOTE oversampling method further improved the performance of the combined features to AUC of 0.960 (sensitivity:0.845, specificity:0.929). Discussion Imaging features had the strongest impact on the model output, while a combination of clinical and imaging features yielded the best performance overall. The identified predictive features were consistent with findings from previous studies. Oversampling yielded mixed results, although it achieved the best performance in our study. Conclusions This study indicates that clinical and imaging features can be used for automated severity assessment of COVID-19 patients and have the potential to assist with triaging COVID-19 patients and prioritizing care for patients at higher risk of severe cases. [Manuscript last updated on 31 July, 2020]

No abstract