Deployment of Optimised Brain Tumour Segmentation Model on ZCU104 FPGA Using Xilinx Vitis AI

Brain tumour segmentation from magnetic resonance imaging requires the accurate delineation of the sub-regions under clinical constraints on computational resources. Deploying deep segmentation networks on edge hardware demands model optimisation that preserves accuracy while reducing complexity. This paper presents the deployment of a pruned RAAGR2-Net brain tumour segmentation model on the Xilinx ZCU104 FPGA using the Vitis AI 3.0 toolchain. The model was optimised through sensitivity-guided channel pruning, which suppresses the channels identified as redundant while maintaining segmentation stability. The pruned model was trained and evaluated on the BraTS 2019 dataset. At a pruning ratio of 0.1, the pruned model achieved Dice scores of 0.8578, 0.8131, and 0.8635 for tumour core, enhancing tumour, and whole tumour, respectively, compared to the scores of 0.8169, 0.7896, and 0.8424 achieved by the base model. The deployment further serves as a verification step, confirming that the optimised model successfully completed quantisation, compilation, and FPGA inference using the Vitis AI toolchain. Following INT8 post-training quantisation and compilation for the DPUCZDX8G accelerator, inference on the ZCU104 produced an average latency of 69.16 ms per sample and a throughput of 14.46 frames per second. These results demonstrate that sensitivity-guided channel pruning is compatible with FPGA deployment without requiring network restructuring.