Knowledge graphs make it possible for models to more precisely store contextual information by predicting the dependencies and relationships between items from structured data. Prediction techniques that are now in use, such as data-based clinical and neuroimaging methodologies, may have issues with interpretability, scalability, and capturing intricate relationships. By explicitly expressing links between things, enabling more interpretable models, enabling scale learning on interconnected data, and capturing complicated patterns in complex systems, knowledge graphs and graph neural networks address these restrictions. A knowledge graph is utilized by the Knowledge Graph-based Neural Network Model to Predict Neurological Disorder (KGNN-PND) system to uncover entity links, which are then transformed into embedding vectors. The model analyzes the vectors using a graph neural network, producing key-value pairs for further examination. Interestingly, the model examines relationships between neurological abnormalities and disease that have never been studied before. This novel approach improves interpretability and precision in the identification of patterns associated with neurological disorders. The proposed system provides healthcare professionals with valuable insights into the likelihood of particular disease manifestations, enabling early detection, personalized treatment strategies, and proactive health management.