Design of remote fault detection method for CT scanning equipment based on two-dimensional search model and ZigBee network / 中国医学装备

ABSTRACT

Objective:To design and construct a remote fault detection method for CT scanning equipment based on two-dimensional search model and ZigBee network to improve the accuracy of fault detection.Methods:ZigBee network technology was used to collect the operating status data of CT scanning equipment from three aspects:environmental factors,electrical factors and gas circuit factors.The time and frequency domain signals were extracted from the monitoring data,and the extracted fault signal feature parameters were fused.The fault data were transformed into stationary time series by inverse binary transformation to restrain the fluctuation of the data.The dependence degree of fault data was calculated by establishing corresponding two-dimensional search model and matching function.Based on this,Bayesian evaluation function was introduced to realize the remote detection of equipment fault.A large-scale clinical CT scanner in clinical use in the hospital was selected as the experimental prototype,and at the same time,550 real fault data generated from the 80 CT scanning scanners with the same model and configuration as the experimental prototype were randomly selected as training samples.The accuracy of CT scanning equipment faults remote detection method with that of the long short-term memory network(referred to as method 1)and deep belief network(referred to as method 2)were compared.Results:The designed remote fault detection method of CT scanning equipment only had detection error on status label 5,but the judgment error was in the controllable range.In the five fault detection experiments,the detection accuracy of the remote fault detection method of CT scanning equipment was more than 80%,which was much higher than that of method 1 and method 2.Conclusion:The fault detection method of CT scanning equipment has a high detection accuracy and can realize the practical requirement of remote detection of CT scanning equipment.