This pilot study explores the feasibility of using text from Electronic Health Record (EHR) at the University Medical Center Utrecht to classify patient eligibility for cardiovascular risk management. The primary goal is to improve the identification of at-risk patients and facilitate timely, personalized interventions. Through a ResNet architecture, which is used as a feature extractor, the outputs are then passed into a neural network classifier to predict patient eligibility for cardiovascular risk management. Despite excluding numerical data from EHR during model training and inference, and the imbalanced nature of the dataset, the model achieved promising results with an accuracy of 85% and an F1-score of 0.85. This preliminary analysis demonstrates the feasibility of the approach and establishes a solid foundation for further improvements. Future research directions include integrating Explainable AI techniques to enhance model transparency, expanding the dataset, and addressing class imbalance through data augmentation or resampling techniques.