How to determine overfitting and underfitting based on loss graph?

Question

I'm currently doing my Final Year Project, the classification of Alzheimer's disease stage using a custom CNN model. However, I'm unsure whether my model is overfitting or underfitting according to my loss graph. I have trained the model with 50 epochs with a batch size of 16 and a learning rate of 0.0001 using the Adam optimizer. The confusion matrix seems alright to me. Does my loss graph show the sign of overfitting or underfitting? Here is my:

1. accuracy graph.

2. loss graph

3. confusion matrix

4. CNN Layer

5. Evaluation Metrics

If it shows any of these issues, how do I know, and what should I do to improve?