No abstract

Edaravone, a novel free radical scavenger, demonstrates neuroprotective effects by inhibiting vascular endothelial cell injury and ameliorating neuronal damage in ischemic brain models. The present study was undertaken to verify its therapeutic efficacy following acute ischemic stroke. We performed a multicenter, randomized, placebo-controlled, double-blind study on acute ischemic stroke patients commencing within 72 h of onset. Edaravone was infused at a dose of 30 mg, twice a day, for 14 days. At discharge within 3 months or at 3 months after onset, the functional outcome was evaluated using the modified Rankin Scale. Two hundred and fifty-two patients were initially enrolled. Of these, 125 were allocated to the edaravone group and 125 to the placebo group for analysis. Two patients were excluded because of subarachnoid hemorrhage and disseminated intravascular coagulation. A significant improvement in functional outcome was observed in the edaravone group as evaluated by the modified Rankin Scale (p = 0.0382). Edaravone represents a neuroprotective agent which is potentially useful for treating acute ischemic stroke, since it can exert significant effects on functional outcome as compared with placebo.

A new subtype classification of ischemic stroke was developed to reflect recent therapeutic strategies as well as evolving concepts of stroke definitions and mechanisms. In 200 consecutive patients with acute ischemic stroke, the inter-rater reliability and proportion of subtypes of the new classification system were assessed and compared with those of the Trial of ORG 10172 in the Acute Stroke Treatment (TOAST) classification. The most frequent subtype of the new classification was atherothrombosis (n = 80, 40%), followed by stroke of undetermined etiology (n = 54, 27%), small artery disease (n = 33, 16.5%), cardioembolism (n = 26, 13%), and stroke of other determined etiology (n = 7, 3.5%). Three raters agreed to the stroke subtype diagnosis in 165 out of 200 cases and the overall ĸ value was excellent (ĸ = 0.82). The new classification system for brain infarction was easy to use and had high inter-rater reliability.

Machine learning and artificial intelligence have achieved a human-level performance in many application domains, including image classification, speech recognition and machine translation. However, in the financial domain expert-based credit risk models have still been dominating. Establishing meaningful benchmark and comparisons on machine-learning approaches and human expert-based models is a prerequisite in further introducing novel methods. Therefore, our main goal in this study is to establish a new benchmark using real consumer data and to provide machine-learning approaches that can serve as a baseline on this benchmark. We performed an extensive comparison between the machine-learning approaches and a human expert-based model—FICO credit scoring system—by using a Survey of Consumer Finances (SCF) data. As the SCF data is non-synthetic and consists of a large number of real variables, we applied two variable-selection methods: the first method used hypothesis tests, correlation and random forest-based feature importance measures and the second method was only a random forest-based new approach (NAP), to select the best representative features for effective modelling and to compare them. We then built regression models based on various machine-learning algorithms ranging from logistic regression and support vector machines to an ensemble of gradient boosted trees and deep neural networks. Our results demonstrated that if lending institutions in the 2001s had used their own credit scoring model constructed by machine-learning methods explored in this study, their expected credit losses would have been lower, and they would be more sustainable. In addition, the deep neural networks and XGBoost algorithms trained on the subset selected by NAP achieve the highest area under the curve (AUC) and accuracy, respectively.

No abstract