Impact of Synthetic Lesional MR Images in Automated Focal Cortical Dysplasia Detection in Low-Data Scenarios

This study investigates whether synthetic MRI data can enhance the automated detection of focal cortical dysplasia (FCD) in scenarios with limited real data. The authors generated realistic synthetic MRIs and found that using these images improved detection sensitivity and model confidence, while also reducing the need for manual annotations. However, they concluded that when sufficient real data is available, it remains more effective than synthetic augmentation for FCD detection.