Medical image synthesis is the use of generative models to produce artificial medical images, such as scans, for training data augmentation, modality translation, or privacy-preserving sharing.
Semantic Classification
Content
- Medical image synthesis generates images for modalities such as MRI, CT, and X-ray. One application translates between modalities, for example producing a synthetic CT image from an MRI scan to support planning without an additional acquisition.
- Synthetic images can augment limited training datasets and can be shared in place of patient scans to reduce privacy exposure, although clinical validation is required before synthetic data influences diagnosis.