The Deep Neural Network helps to understand the interaction between the human brain and the simulated computational studies. Making best decision and gives an explicit result with an algorithm solves the limitation of input data, the hidden layers, and the overlapping problem between the connection layers. On the other side, estimating the central corneal thickness post- operatively during the laser eye surgery are mostly important parameter that may play the main roles in clinical decisions. Decisions like fully, under or over correction of refractive error of human eye. This decision is related to a number of clinical measurements (19 sets of inputs) that may be interconnected in complex form. In the present work the Deep Belief Neural Network have been modeled, capable of estimating the weight of the input clinical parameters relative to the final central depth for the laser eye surgery. Estimating these interconnection weights will help to correct amount of dose of laser to return the eye to its normal state. A gradient descent and back propagation technique algorithm through the training test will also help to correct the overlapping problem and how much the data and the hidden layers rise the machine will still stable. Result shows how much this model is stable in compare with the standard practical and theories and it gives the best and accurate decision which can depend on it within the medical diagnosis and treatment.
How to Cite This Article
Jaffer, Sumia H.; Ghaeb, Nebras H.; and Baker, Sahar S.
"Estimating the Corneal Thickness for post-operative Laser Eye Surgery using Deep Neural Network,"
Polytechnic Journal: Vol. 8:
2, Article 6.