Temporal data abstraction (TA) is a set of techniques aiming to abstract time-points into higher-level interval concepts and to detect significant trends in both low-level data and abstract concepts. Dynamic Bayesian networks (DBNs) are temporal probabilistic graphical models that model temporal processes, temporal relationships between events and state changes through time. In this paper, we propose the integration of TA methods with DBNs in the context of medical decision-support systems, by presenting an extended DBN model. More specifically, we demonstrate the derivation of temporal abstractions which are used for building the network structure. We also apply machine learning algorithms to learn the parameters of the model through data. The model is applied for diagnosis of coronary heart disease using as testbed a longitudinal dataset. The classification accuracy of our model evaluated using the evaluation metrics of Precision, Recall and F1-score, shows the effectiveness of our proposed system.