M-DDC: MRI based demyelinative diseases classification with U-Net segmentation and convolutional network.

Childhood demyelinative diseases classification (DDC) with brain magnetic resonance imaging (MRI) is crucial to clinical diagnosis. But few attentions have been paid to DDC in the past. How to accurately differentiate pediatric-onset neuromyelitis optica spectrum disorder (NMOSD) from acute disseminated encephalomyelitis (ADEM) based on MRI is challenging in DDC. In this paper, a novel architecture M-DDC based on joint U-Net segmentation network and deep convolutional network is developed. The U-Net segmentation can provide pixel-level structure information, that helps the lesion areas location and size estimation. The classification branch in DDC can detect the regions of interest inside MRIs, including the white matter regions where lesions appear. The performance of the proposed method is evaluated on MRIs of 201 subjects recorded from the Children's Hospital of Zhejiang University School of Medicine. The comparisons show that the proposed DDC achieves the highest accuracy of 99.19% and dice of 71.1% for ADEM and NMOSD classification and segmentation, respectively.

Differentiation of MOGAD in ADEM-like presentation children based on FLAIR MRI features.

OBJECTIVE: The differences in magnetic resonance imaging (MRI) between children with classic acute disseminated encephalomyelitis (ADEM) and myelinal oligodendrocyte glycoprotein antibody associated disease (MOGAD) with ADEM-like presentation are controversial. The purpose of this study was to investigate whether the radiological characteristics of the MRI-FLAIR sequence can predict MOGAD in children with ADEM-like presentation and to further explore its imaging differences. METHODS: We extracted 1041 radiomics features from MRI-FLAIR lesions. Then we used the redundancy analysis (Spearman correlation coefficient), significance test (student test or Mann-Whitney U test), least absolute contraction and selection operator (LASSO) to select potential predictors from the feature groups. The selected potential predictors and MOG antibody test results were used to fit the machine learning model for classification. Combined with feature selection and machine learning classifiers, the optimal model for each subgroup was derived. The resulting models have been evaluated using the receiver operator characteristic curve (ROC) at the lesion level and the model performance was evaluated at the case level using decision curve analysis. RESULTS: We retrospectively reviewed and re-diagnosed 70 ADEM-like presentation cases in our center from April 2015 to January 2020. Including 49 cases with classic ADEM and 21 cases with MOGAD. 30(43%) were female, with a median age of 5.3 years. On the four subgroups by age and gender, the area under the curve (AUC) of the optimal models were 89%, 90%, 98%, and 99%, and the MOGAD detection rates (Specificity) were 83%, 83%, 92%, and 75%, respectively. CONCLUSIONS: The machine learning model trained on radiomics features of MR-FLAIR images can effectively predict patients' MOGAD. This study provides a fast, objective, and quantifiable method for MOGAD diagnosis.

Determination of significant parameters in remote ischemic postconditioning for ischemic stroke in experimental models: A systematic review and meta-analysis study.

OBJECTIVES: To systematically review studies using remote ischemia postconditioning (RIPostC) for ischemic stroke in experimental models and obtain factors that significantly influence treatment outcomes. MATERIALS AND METHODS: Peer-reviewed studies were identified and selected based on the eligibility criteria, followed by extraction of data on potentially influential factors related to model preparation, postconditioning, and measure time based on outcome measures including infarct size, neurological scales, and cell tests with autophagy, apoptosis, normal-neuron, and damaged-neuron counting. Then, all data were preprocessed, grouped, and meta-analyzed with the indicator of the standardized mean difference. RESULTS: Fifty-seven studies with 224 experiments (91 for infarct size, 92 for neurological scales, and 41 for cell-level tests) were included. There was little statistical difference between different model preparations, treated body parts, number of treatments, and sides. And treatment effect was generally a positive correlation with the duration of conditioning time to stroke onset with exceptions at some time points. Based on infarct size, the number of cycles per treatment, duration of occlusion, and release per cycle showed significant differences. Combined with the effect sizes by other measures, the occlusion/release duration of 8-10 min per cycle is better than 5 min, and three cycles per treatment were most frequently used with good effects. Effect also varied when measuring at different times, showing statistical differences in infarct size and most neurological scales. RIPostC is confirmed as an effective therapeutic intervention for ischemic stroke, while the RIPostC-mediated autophagy level being activated or inhibited remained conflicting. CONCLUSIONS: Conditioning time, number of cycles per treatment, duration of occlusion, and release per cycle were found to influence the treatment effects of RIPostC significantly. More studies on the relevant influential factors and autophagy mechanisms are warranted.