Objective:Radiation exposure during paediatric cardiac catheterisation procedures should be minimised to “as low as reasonably achievable”. The aim of this study was to evaluate the effectiveness of a modified radiation safety protocol in reducing patient dose during paediatric interventional cardiac catheterisation.Methods:Radiation dose data were retrospectively extracted from January 2014 to December 2015 (Standard group) and prospectively collected from January 2016 to December 2017 (Low-dose group) after implementation of a modified radiation safety protocol. Both groups included five most common procedures: atrial septal defect closure, patent ductus arteriosus closure, perimembranous ventricular septal defect closure, pulmonary valvuloplasty, and supraventricular tachycardia ablation.Results:Median air Kerma was 48.4, 50.5, 29.75, 149, 218, and 12.9 mGy for atrial septal defect closure, pulmonary valvuloplasty, patent ductus arteriosus closure <20 kg, ventricular septal defect closure <20 kg, ventricular septal defect closure ≧20 kg, and supraventricular tachycardia ablation in Standard group, respectively, which significantly decreased to 18.75, 20.7, 11.5, 41.9, 117, and 3.3 mGy in Low-dose group (p < 0.05). This represents a reduction in dose to each patient between 46 and 74%. Among five procedural types in Low-dose group, dose of ventricular septal defect closure was the highest with median air Kerma of 62.5 mGy, dose area product of 364.7 μGy.m2, and dose area product per body weight of 21.5 μGy.m2/kg, respectively, along with the longest fluoroscopy time of 9.9 minutes.Conclusion:We provided a feasible radiation safety protocol with specific settings on a case-by-case basis. Increasing awareness and adequate training of a practical radiation dose reduction program are essential to improve radiation protection for children.

Blade icing is one of the common issues of large-scale wind turbines located in cold regions, which will affect the safety and efficiency of the whole turbine system. Currently, data-driven fault detection has gained increasingly interests due to the availability of a large volume of supervisory control and data acquisition (SCADA) data. However, SCADA data has complex time-varying characteristics and strong spatio-temporal correlations among different sensor variables, thus it is still challenging to extract effective fault features for accurate detection. To this end, this paper proposes an enhanced spatio-temporal feature learning approach, called multi-task temporal spatial attention network (MT-STAN). It contains two core modules: a feature extraction module and a multi-task learning module. For better spatio-temporal feature extraction, a spatio-temporal attention block is first developed to extract important variables in spatial dimension and temporal segments in temporal dimension via attention mechanism. Then, we design a multitask learning module, consisting of both deep metric learning and classification learning tasks, to further enhance the discriminative ability of the learned representations and improve the performance of fault detection. The proposed approach is evaluated on a real SCADA dataset, and the results show that our proposed MT-STAN model achieved better detection performance compared with several baseline models.