Development and Validation of a Deep Learning-Based Model Using Computed Tomography Imaging for Predicting Disease Severity of Coronavirus Disease 2019.

OBJECTIVES: Coronavirus disease 2019 (COVID-19) is sweeping the globe and has resulted in infections in millions of people. Patients with COVID-19 face a high fatality risk once symptoms worsen; therefore, early identification of severely ill patients can enable early intervention, prevent disease progression, and help reduce mortality. This study aims to develop an artificial intelligence-assisted tool using computed tomography (CT) imaging to predict disease severity and further estimate the risk of developing severe disease in patients suffering from COVID-19. MATERIALS AND METHODS: Initial CT images of 408 confirmed COVID-19 patients were retrospectively collected between January 1, 2020 and March 18, 2020 from hospitals in Honghu and Nanchang. The data of 303 patients in the People's Hospital of Honghu were assigned as the training data, and those of 105 patients in The First Affiliated Hospital of Nanchang University were assigned as the test dataset. A deep learning based-model using multiple instance learning and residual convolutional neural network (ResNet34) was developed and validated. The discrimination ability and prediction accuracy of the model were evaluated using the receiver operating characteristic curve and confusion matrix, respectively. RESULTS: The deep learning-based model had an area under the curve (AUC) of 0.987 (95% confidence interval [CI]: 0.968-1.00) and an accuracy of 97.4% in the training set, whereas it had an AUC of 0.892 (0.828-0.955) and an accuracy of 81.9% in the test set. In the subgroup analysis of patients who had non-severe COVID-19 on admission, the model achieved AUCs of 0.955 (0.884-1.00) and 0.923 (0.864-0.983) and accuracies of 97.0 and 81.6% in the Honghu and Nanchang subgroups, respectively. CONCLUSION: Our deep learning-based model can accurately predict disease severity as well as disease progression in COVID-19 patients using CT imaging, offering promise for guiding clinical treatment.

Altered intrinsic regional brain spontaneous activity in patients with comitant strabismus: a resting-state functional MRI study.

OBJECTIVE: To investigate the underlying regional homogeneity (ReHo) of brain-activity abnormalities in patients with comitant strabismus (CS) and their relationship with behavioral performance. METHODS: Twenty patients with CS (ten men and ten women) and 20 (ten men and ten women) age-, sex-, and education-matched healthy controls (HCs) underwent resting-state functional magnetic resonance imaging scans. The ReHo method was used to assess local features of spontaneous brain activities. Patients with CS were distinguished from HCs by receiver operating characteristic curve. Correlation analysis was performed to explore the relationship between the observed mean ReHo values of the different brain areas and behavioral performance. RESULTS: Compared to HCs, the patients with CS showed significantly increased ReHo values in the right inferior temporal cortex/fusiform gyrus/cerebellum anterior lobe, right lingual gyrus, and bilateral cingulate gyrus. We did not find any relationship between the observed mean ReHo values of the different brain areas and behavioral performance. CONCLUSION: CS causes dysfunction in many brain regions, which may explain the fusion compensation in CS.

Altered intrinsic regional spontaneous brain activity in patients with optic neuritis: a resting-state functional magnetic resonance imaging study.

OBJECTIVE: To investigate the underlying regional homogeneity (ReHo) in brain-activity deficit in patients with optic neuritis (ON) and its relationship with behavioral performance. MATERIALS AND METHODS: In total, twelve patients with ON (four males and eight females) and twelve (four males and eight females) age-, sex-, and education-matched healthy controls underwent resting-state functional magnetic resonance imaging scans. The ReHo method was used to assess the local features of spontaneous brain activity. Correlation analysis was used to explore the relationship between the observed mean ReHo values of the different brain areas and the visual evoked potential (VEP) in patients with ON. RESULTS: Compared with the healthy controls, patients with ON showed lower ReHo in the left cerebellum, posterior lobe, left middle temporal gyrus, right insula, right superior temporal gyrus, left middle frontal gyrus, bilateral anterior cingulate cortex, left superior frontal gyrus, right superior frontal gyrus, and right precentral gyrus, and higher ReHo in the cluster of the left fusiform gyrus and right inferior parietal lobule. Meanwhile, we found that the VEP amplitude of the right eye in patients with ON showed a positive correlation with the ReHo signal value of the left cerebellum posterior lobe (r=0.701, P=0.011), the right superior frontal gyrus (r=0.731, P=0.007), and the left fusiform gyrus (r=0.644, P=0.024). We also found that the VEP latency of the right eye in ON showed a positive correlation with the ReHo signal value of the right insula (r=0.595, P=0.041). CONCLUSION: ON may involve dysfunction in the default-mode network, which may reflect the underlying pathologic mechanism.

Disturbed spontaneous brain-activity pattern in patients with optic neuritis using amplitude of low-frequency fluctuation: a functional magnetic resonance imaging study.

OBJECTIVE: To use the amplitude of low-frequency fluctuation (ALFF) technique to investigate the local features of spontaneous brain activity in optic neuritis (ON) and their relationship with behavioral performance. MATERIALS AND METHODS: Twelve patients with ON (four male, eight female) and twelve age-, sex-, and education status-matched healthy controls (HCs) (four male, eight female) underwent resting-state functional magnetic resonance imaging (rs-fMRI) scans. The ALFF technique was used to assess local features of spontaneous brain activity. Correlation analysis was used to explore the relationship between the observed mean ALFF values of the different areas and visual evoked potentials (VEPs) in patients with ON. RESULTS: Compared with HCs, patients with ON had significantly decreased ALFF values in the posterior and anterior lobes of the right cerebellum, right putamen, right inferior frontal gyrus, right insula, right supramarginal gyrus, right inferior parietal lobule, left medial frontal gyrus, left superior temporal gyrus, bilateral anterior cingulate/medial frontal gyrus, and bilateral precuneus, and significantly increased ALFF values in the posterior lobes of the left and right cerebellum, right inferior temporal gyrus, right inferior temporal/fusiform gyrus, left parahippocampal gyrus, left fusiform gyrus, left calcarine fissure, left inferior parietal lobule, and left cuneus. We found negative correlations between the mean ALFF signal value of the left parahippocampal gyrus and the VEP amplitude of the right eye in ON (r=-0.584, P=0.046), and a positive correlation between the mean ALFF signal value of the bilateral precuneus and the best-corrected visual acuity of the left eye (r=0.579, P=0.048) in patients with ON. CONCLUSION: ON mainly seems to involve dysfunction in the default-mode network, cerebellum, and limbic system, which may reflect the underlying pathologic mechanism of ON.