The Ganglionic Eminence (GE) is a transient structure of the fetal brain, which evolves in the ventral telencephalon from the fifth-week post-conception and plays a pivotal role in neural migration during development. Early detection of GE anomalies is crucial for identifying migration deficiencies that may lead to postnatal neurological or psychiatric disorders. Currently, no robust automatic GE segmentation approaches exist to enable its early analysis. Segmentation challenges arise from the transient nature of this structure, resulting in an increase of isolated components and a decrease in their volumes with age. Additional complexity is introduced by the inaccuracy of determining the gestational age, potential motion artifacts due to fetal movement, and intensity fluctuations in images for the same structure. In this work, we propose an automated GE segmentation method for fetal Magnetic Resonance Imaging (MRI) data by extending 3D UNets and introducing a novel registration-driven generative data augmentation technique to increase the number of MRI from 138 to more than 2,400 fetal scans with manually defined labels by an expert neuroradiologist. Our solution spans 19 to 38 weeks of gestation, achieving a mean Dice score of 0.76±0.05, Volume Similarity of 0.89±0.11 and Hausdorff Distance of 3.98±4.85. Additionally, we could show that the volumetric dynamics of predicted GE segmentations correlate with anatomical GE development patterns, indicating the model’s ability to learn GE dynamics over time and improving segmentation performance.