<p>Intraocular pressure (IOP) is a key parameter to diagnose glaucoma disease and assess the treatment effect of cornea after refractive surgery. Current refractive surgeries inevitably change the configuration of the cornea, making it difficult to measure IOP accurately using conventional methods. The prediction method proposed in this article can accurately measure the IOP after refractive surgery. In this study, firstly, the finite element models of cornea free of IOP depicted by various vertex height, thickness, and radius are established, and the deformation of the cornea under different IOP is predicted. Based on the dataset obtained from the numerical simulations and the multi-layer perceptron neural network algorithm, two prediction models for the vertex height and thickness of the cornea free of IOP and for the configuration under IOP are developed, and a prediction method for IOP is then proposed by combining the two models. Following the similar way, two prediction models respectively for the parameters of the presumptive initial configuration of the cornea to undergo refractive surgery and for those of the cornea after surgery are constructed, and a prediction method for IOP of cornea after the surgery is presented. The validity of the prediction methods for regular IOP and that after refractive surgery is demonstrated using the clinical data from some volunteers. The proposed methods provide an efficient prediction method for regular IOP and that after refractive surgery.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Intraocular pressure prediction method combining finite element simulations and multi-layer perceptron neural network

  • Shi Yan,
  • Xiaocheng Hu,
  • Xiaohui Song,
  • Ke Yao,
  • Shaoxing Qu

摘要

Intraocular pressure (IOP) is a key parameter to diagnose glaucoma disease and assess the treatment effect of cornea after refractive surgery. Current refractive surgeries inevitably change the configuration of the cornea, making it difficult to measure IOP accurately using conventional methods. The prediction method proposed in this article can accurately measure the IOP after refractive surgery. In this study, firstly, the finite element models of cornea free of IOP depicted by various vertex height, thickness, and radius are established, and the deformation of the cornea under different IOP is predicted. Based on the dataset obtained from the numerical simulations and the multi-layer perceptron neural network algorithm, two prediction models for the vertex height and thickness of the cornea free of IOP and for the configuration under IOP are developed, and a prediction method for IOP is then proposed by combining the two models. Following the similar way, two prediction models respectively for the parameters of the presumptive initial configuration of the cornea to undergo refractive surgery and for those of the cornea after surgery are constructed, and a prediction method for IOP of cornea after the surgery is presented. The validity of the prediction methods for regular IOP and that after refractive surgery is demonstrated using the clinical data from some volunteers. The proposed methods provide an efficient prediction method for regular IOP and that after refractive surgery.