A machine learning technique called "few-shot learning" enables artificial neural networks to learn and generate accurate predictions using just a few number of labelled instances. Due to the scarcity of labelled information in the medical field, few-shot medical picture classification is a difficult task. MedOptNet is a deep learning-based optimisation system designed especially for applications in medical imaging. It is specifically made to improve the performance and quality of magnetic resonance imaging (MRI). It reduces the time and resources needed for scanning while improving the presentation of medical images by using a neural network to solve issues like photo reconstruction and denoising. To overcome this difficulty, the MedOptNet system employs curvature-based optimisation and meta-learning, which produces better results than conventional techniques. The goal of this initiative is to improve the model's feature extraction and classification accuracy by addin