Ganglion cysts are the most common soft tissue tumors of the wrist, although accurate measurement and identification are still difficult to achieve during diagnosis. Due to their limited contrast with the surrounding soft tissues, single-sequence images frequently fail to clearly delineate cyst boundaries, even though conventional Magnetic Resonance Images (MRI) offers useful contrast information. FusionNet is a deep learning method that accurately segments wrist ganglion cysts by fusing multi-contrast MRI fusion with a modified U-Net architecture. The dataset is divided into training, validation, and testing subsets and includes clinical wrist MRI scans as well as artificially produced images with cyst annotations. Dice loss and cross-entropy loss are combined in the end-to-end training of FusionNet, which uses multi-contrast fusion to improve the visualization of cyst boundaries. With a Dice score of 0.92 and an Intersection-over-Union (IoU) of 0.89, the experimental evaluation shows a significant improvement over baseline U-Net and traditional segmentation techniques. Effective cyst border delineation and trustworthy volume change monitoring across timepoints are further validated by qualitative evaluation. This method is the first complete end-to-end solution that combines deep learning and MRI fusion for wrist ganglion cyst segmentation, enabling motion-aware dynamic imaging, longitudinal cyst volume monitoring, and improved clinical evaluation.