Deep learning based infant sucking simulation

The objective of this thesis is to develop a model of the vacuum dynamics of infants during the sucking process and to simulate this behavior using deep learning-based approaches. While breastfeeding plays an important role in the physical and emotional development of infants, difficulties experienced in this process can have a negative impact on nutritional production. In this context, pressure changes during the sucking process were modelled using Wasserstein Generative Adversarial Networks (WGAN) and the generated data were compared with the original data. The results of the research demonstrated that deep learning-based models can accurately reflect the dynamics of sucking behavior and produce consistent results supported by validation metrics. This study represents a novel approach to addressing neonatal health and breastfeeding problems by offering new perspectives for both clinical and technological applications.