Prediction and Analysis of Volume Doubling Time of Sub-solid Nodules Based on Radiomics
TAO Xuemin1,FANG Rui1,JIN Xin1,LUO Chuncai1,YU Shilin2,HU Fei2,WANG Fang2,ZHAO Shaohong1
1. Department of Radiology, The First Medical Center, Chinese PLA General Hospital, Beijing 100853, China丨
2. Beijing Peredoc Technology Co., Ltd., Beijing 100043, China
Abstract:Objective To predict the volume doubling time of sub-solid pulmonary nodules by using radiomics to determine the
follow-up period and interval of patients. Methods The CT images of 45 patients with sub-solid pulmonary nodules in this study
were segmented by semi-automatic sketching, and the radiomic features of the lesions were extracted. LASSO regression and k-fold
cross-validation were used to screen two clinical features and radiomic features, and the features with zero regression coefficient were
eliminated, the remains were used to build the model of prediction doubling time. Results According to the minimum mean square
error (MMSE), five features were selected, including age, entropy of three-dimensional gray level co-occurrence matrix, small area
emphasis of three-dimensional gray level band matrix, morphological feature classification and density 2. The regression coefficients
were 0.024, 0.010, -0.029, -0.013 and 0.018, respectively. The multiple linear regression was used to establish the prediction model,
and the effectiveness of the model was verified in the test set. Conclusion The prediction model based on radiomics can quantify
the volume doubling time of the long-term sub-solid pulmonary nodules, which is critical for treatment suggestions and follow-up
interval decision.
陶雪敏1,方瑞1,金鑫1,罗春材1,余石林2,胡飞2,王方2,赵绍宏1. 基于影像组学的亚实性结节体积倍增时间预测分析[J]. 中国医疗设备, 2021, 36(8): 30-34.
TAO Xuemin1,FANG Rui1,JIN Xin1,LUO Chuncai1,YU Shilin2,HU Fei2,WANG Fang2,ZHAO Shaohong1. Prediction and Analysis of Volume Doubling Time of Sub-solid Nodules Based on Radiomics. China Medical Devices, 2021, 36(8): 30-34.