Application of BP Neural Network in Corneal Stromal Cutting Thickness Prediction in
SMILE Surgery
TANG Funana, YANG Chunhuaa, ZHANG Kea, ZHU Mingyuea, ZHANG Huia, WANG Yinga, YUAN Dongqingb
a. Department of Clinical Medical Engineering; b. Department of Ophthalmology, The First Affiliated Hospital of Nanjing Medical
University, Nanjing Jiangsu 210029, China
Abstract:Objective To construct the BP neural network model for accurately predicting the corneal stromal cutting thickness in
small-incision lenticule extraction (SMILE) surgery. Methods The standardized cutting thickness data table was obtained from
the manufacturer of full femtosecond surgical system, and a total of 12188 data were extracted as the research objects. Taking the
influencing factors of cutting thickness as input and the cutting thickness as output, the BP neural network model for predicting the
cutting thickness of corneal stroma in smile operation was constructed by MATLAB programming software, trained and verified.
Results For the BP neural network model, the maximum number of training iterations was 5000, the actual number of iterations was
2464, the setting error was 0.001 and the actual error was 0.112. The BP neural network model was used to simulate and verify the
data of 1038 clinical patients undergoing smile surgery in our hospital. After rounding the predicted data of cutting thickness and the
actual data, the error of about 99.81% of the data is within [-1,1] μm. Conclusion The BP neural network prediction model based
on Matlab realized the description of the nonlinear relationship between corneal stromal cutting thickness and spherical lens degree,
cylindrical lens degree, corneal curvature radius and lenticule diameter.By calling this model, the corneal stromal cutting thickness
can be accurately predicted and the residual corneal stromal thickness can be calculated, which provides a basis for smile operation
parameters and improves the diagnosis and treatment efficiency of SMILE operation.
汤福南a,杨春花a,张可a,竺明月a,张晖a,汪缨a,袁冬青b. BP神经网络在SMILE手术角膜基质切削厚度预测中的应用[J]. 中国医疗设备, 2022, 37(10): 41-45.
TANG Funana, YANG Chunhuaa, ZHANG Kea, ZHU Mingyuea, ZHANG Huia, WANG Yinga, YUAN Dongqingb. Application of BP Neural Network in Corneal Stromal Cutting Thickness Prediction in
SMILE Surgery. China Medical Devices, 2022, 37(10): 41-45.