Federated Learning- Based 3D Medical Image Compression

Key Points

• The central aim is to evaluate the performance of federated learning techniques in compressing 3D medical images while ensuring data privacy.
• Utilized federated learning algorithms for model training without data centralization.
• Conducted performance evaluations on 3D medical image datasets.
• Analysed compression ratios and quality of compressed images.
• Achieved significant compression rates compared to traditional methods.
• Maintained high image quality despite reduced file sizes.
• Demonstrated effective data privacy through decentralized learning.