Abstract:It is subjective for clinicians to judge the common location of hepatic echinococcosis by experience. Therefore, the Faster
RCNN object detection model was applied to detect hepatic echinococcosis in this study, and the SCP method was combined to
estimate the location of hepatic echinococcosis in the liver region. In this study, the object detection model based on ResNet101
network was used to detect the lesions of hepatic echinococcosis, and then used the polar coordinate system based on SCP method to
calculate the common location of hepatic hydatid disease in the liver region. The experimental results showed that the object detection
model based on ResNet101 network can effectively extract the features of the target, and the detection accuracy rate reached 89.6%.
The frequent occurrence locations of hepatic echinococcosis based on SCP method are mainly in the IV and Ⅵ/Ⅶ regions of the
liver. In this study, the detection of liver hydatid lesions by object detection model can help doctors diagnose the disease and reduce
the occurrence of missed detection and wrong detection. For the statistics of liver hydatid disease in the liver region of the common
location, it can assist doctors in early detection, early diagnosis, and early treatment of the disease.
刘志华a,卡迪力亚·库尔班a,李丰军b,严传波b. Faster RCNN模型和SCP方法在肝包虫病病灶位置估计中的研究[J]. 中国医疗设备, 2021, 36(9): 91-94.
LIU Zhihuaa,KUERBAN Kadiliyaa,LI Fengjunb,YAN Chuanbob. Research on Faster Region-Convolutional Neural Networks Model and
SCP Method in Estimating the Location of Hepatic Echinococcosis. China Medical Devices, 2021, 36(9): 91-94.