Abstract

Info

Author(s):: YANG Yilong1; 2; FENG Yifei2; ZHU Liufeng1; LIU Yi1; HE Ying2; 1. School of Health Science?nd Engineering, University of Shanghai for Science and Technology, Shanghai 200093, China 2. Naval Specialty Medical Center, Shanghai 200433, China

Keywords:: Keywords: cerebral hemorrhage microwave detection system optimization XGBoost algorithm

Abstract:: Abstract: Objective To design and optimize the detection system for cerebral hemorrhage for overcoming the limitations in microwave detection system, such as expensive equipment, complex antenna gating switch system and redundant antenna channels. Methods The system adopted the simplest patch antenna structure, optimized the antenna gating switch system and reduced the number of channels by radiofrequency switch chip strategy. The simulated cerebral hemorrhage was sampled and detected through the optimized detection system and pattern recognition method was used to distinguish whether there was cerebral hemorrhage. Results XGBoost algorithm model showed superior performance in the task of cerebral hemorrhage recognition, with an accuracy of 1.000 on the test set, an average accuracy of 0.973 in K-fold cross-validation, and an accuracy of 0.996 on the training set. Conclusion The optimized cerebral hemorrhage detection system which has high detection accuracy and reliability has the potential to identify cerebral hemorrhage.