Hybrid Regression Models for Predicting Hydration: A Case Study in Pediatric Hemodialysis

With the rapid development of artificial intelligence (AI) and machine learning (ML), the medical field is experiencing significant advances that make it easier for doctors to monitor and personalize the hemodialysis process in pediatric patients. This case study focuses on the prediction of hydration in pediatric hemodialysis using hybrid ML models with hyperparameters for prediction accuracy. Data were collected from pediatric hemodialysis at the University Children's Hospital in Tiršova, and parameters such as weight, blood pressure, lean tissue index (LTI), fat tissue index (FTI), body mass index (BMI), total body water, extracellular water (ECW), adipose tissue mass, body cell mass, and bioimpedance were adjusted for training. The model was configured for each pediatric patient and retrained after each treatment to make individualized predictions with the highest accuracy. The proposed model uses measurable parameters to estimate hydration and provide better recommendations to the physician, leading to better results than commonly used state-of-the-art competing methods. The hybrid ML models with hyperparameters represent a novel, simplified, safe, and efficient method for predicting hydration in children, making it easier for doctors to monitor the hemodialysis process for children.