[An optimal predicting method based on improved genetic algorithm embedded in neural network and its application to...

Area of Science:

• Biomedical Engineering
• Artificial Intelligence in Medicine
• Renal Medicine

Background:

• Peritoneal fluid absorption rate (PFAR) is a crucial index in peritoneal dialysis for renal failure patients.
• Accurate prediction of PFAR is complex due to the intricate nature of the dialysis process.
• Existing methods may lack the precision required for optimal treatment management.

Purpose of the Study:

• To develop an innovative predictive model for PFAR.
• To enhance the accuracy and efficiency of PFAR prediction in peritoneal dialysis.
• To leverage the strengths of genetic algorithms and neural networks for improved modeling.

Main Methods:

• An improved genetic algorithm (GA) was employed to optimize the initial weights and biases of a neural network (NN).
• The GA-NN hybrid model was designed to predict the PFAR index.
• The performance of the improved model was benchmarked against the standard GA-NN hybrid method.

Main Results:

• The improved GA-NN model demonstrated significantly enhanced prediction accuracy compared to the standard method.
• The optimized neural network model required less time for the learning process.
• Simulation results validated the superior performance of the proposed predictive model.

Conclusions:

• The developed improved GA-NN model offers a more accurate and efficient approach to predicting PFAR.
• This innovative method holds promise for optimizing peritoneal dialysis treatment for renal failure patients.
• The integration of GA's global search with NN's local search capabilities proved effective.