No abstract

Objective To investigate the effects of the combination of extracorporeal cardiopulmonary resuscitation (ECPR) and thrombolytic therapy on the recovery of vital organ function after prolonged cardiac arrest. Design Laboratory investigation Setting University Laboratory Subjects Pigs Interventions Animals underwent 30-minute untreated ventricular fibrillation cardiac arrest followed by extracorporeal cardiopulmonary resuscitation (ECPR) for 6 hours. Animals were allocated into two experimental groups: t-ECPR, which received Streptokinase 1 MU and c-ECPR which did not receive Streptokinase. In both groups the resuscitation protocol included the following physiologic targets: mean arterial pressure (MAP) > 70 mmHg, Cerebral perfusion pressure (CerPP) > 50 mmHg, PaO2 150 ± 50 mmHg, PaCO2 40 ± 5 mmHg and core temperature 33 ± 1 °C. Defibrillation was attempted after 30 minutes of ECPR. Measurements and Main Results A cardiac resuscitability score was assessed on the basis of: success of defibrillation; return of spontaneous heart beat; weanability form ECPR; and left ventricular systolic function after weaning. The addition of thrombolytic to ECPR significantly improved cardiac resuscitability (3.7 ± 1.6 in t-ECPR vs 1.0 ± 1.5 in c-ECPR). Arterial lactate clearance was higher in t-ECPR than in c-ECPR (40 ± 15% VS 18 ± 21 %). At the end of the experiment, the intracranial pressure was significantly higher in c-ECPR than in t-ECPR. Recovery of brain electrical activity, as assessed by quantitative analysis of EEG signal, and ischemic neuronal injury on histopathologic examination did not differ between groups. Animals in t-ECPR group did not have increased bleeding complications, including intracerebral hemorrhages. Conclusions In a porcine model of prolonged cardiac arrest, thrombolytic-enhanced ECPR improved cardiac resuscitability and reduced brain edema, without increasing bleeding complications. However, early EEG recovery and ischemic neuronal injury were not improved.

No abstract

No abstract

Learning by teaching is a broadly used methodology in human learning and shows great effectiveness in improving learning outcome: a learner deepens his/her understanding of a topic by teaching this topic to others. We are interested in investigating whether this powerful learning technique can be borrowed from humans to improve the learning abilities of machines. We propose a novel machine learning approach called learning by teaching (LBT). In our approach, the teacher creates a pseudo-labeled dataset and uses it to train a student model. Based on how the student performs on the validation dataset, the teacher re-learns its model and re-teaches the student until the student achieves great validation performance. We propose a multi-level optimization framework to formulate LBT which involves three learning stages: teacher learns; teacher teaches student; teacher and student validate themselves. We develop an efficient algorithm to solve the LBT problem. We apply our approach to neural architecture search on CIFAR-100, CIFAR-10, and ImageNet, where the results demonstrate the effectiveness of our method.