 The study aimed to develop a deep learning model for automatic measurement of various temporal parameters during swallowing using ResNet3D, which outperformed previous models in terms of accuracy, F1 score, and average precision values.