Search

Birth weight is an important indicator of fetal development, which directly influencing the health and safety of both mother and child. However, accurately predicting fetal growth remains a challenging task due to complex influencing factors. To overcome this issue, this paper proposes a new framework called Multidirectional Perception Generative Adversarial Network with Rat Swarm Optimizer for Fetal Growth Prediction (MPGAN-RSO-FGP) to enhance birth weight predictions. The model integrates the capabilities of the Multidirectional Perception Generative Adversarial Network (MPGAN) with the Rat Swarm Optimizer (RSO) to optimize prediction accuracy. Input parameters, including gestational age and birth weight are categorized into three sets: (i) Small for Gestational Age (SGA), (ii) Appropriate for Gestational Age (AGA), (iii) Large for Gestational Age (LGA). In general, the MPGAN does not adopt any optimization strategy to determine the optimal parameters. That’s why, RSO is used to optimize the MPGAN for accurate fetal growth prediction. The proposed MPGAN-RSO-FGP is evaluated using performance metrics, such as Accuracy, Mean Relative Error (MRE), F-score, Precision, Sensitivity, Specificity, ROC, Computational time. The experimental results exemplify that the MPGAN-RSO-FGP outperforms existing models. The MPGAN-RSO-FGP attains 20.78%, 23.67%, and 17.98% higher accuracy, and 21.98%, 23.56%, and 30.78% higher precision compared to the existing LSTM-FBWP, SVM-PSGA, and RF-PLBW models. These findings demonstrate the model’s significant impact on decision-making systems, providing more reliable and efficient fetal growth predictions, which can aid in timely clinical interventions and improve maternal-infant outcomes. [ABSTRACT FROM AUTHOR]