Text this: Optimizing Pre-Trained Models for Medical Dataset Classification with a Fine-Tuning Approach