This study presents a deep mathematical morphological neural network for the classification of periapical radiographs in the diagnosis of dental diseases. It extends a previous rule-based diagnostic system that utilized Bayes’ theorem and IF–THEN logic to diagnose common dental conditions using symptom-based expert knowledge. While earlier approaches relied on structured clinical inputs, the current work shifts toward image-based diagnosis using periapical radiographs. The proposed model integrates mathematical morphology with deep neural network architectures, leveraging convolutional neural network (CNN) principles to enhance feature extraction and pattern recognition in dental X-ray images. By combining morphological operators with deep learning, the system improves the ability to identify structural variations associated with dental pathologies. This hybrid approach aims to support clinicians by providing automated, objective, and consistent diagnostic assistance, thereby reducing subjectivity in radiographic interpretation. The framework demonstrates the potential of integrating mathematical morphology with artificial intelligence for improved dental disease classification and decision support in clinical practice
Müller et al. (Mon,) studied this question.