基于GPUs可视化技术的心脏辅助诊断系统研究

Research of Cardiac Aided Diagnosis System Based on Visualization of GPUs

目的:实现基于GPUs的心脏断层图像的精确分割和三维可视化,完成心脏辅助诊断系统的设计。方法:结合临床专家诊断经验、心脏CT图像先验特征和图像分割算法模型,采用GPUs并行数据处理技术实现心脏结构的分割和三维可视化。结果:完成了CT心脏序列图像的精确、快速、鲁棒分割和三维可视化,初步实现了基于GPUs的可视化技术的心脏辅助诊断系统。结论:研究充分利用计算机图形处理单元GPU强大的并行计算能力,解决了医学图像处理和分割中的问题,提高了程序的运行效率,改善了用户体验。

Objective To develop a real-time CAD system of cardiac medical image.Methods In combination with diagnosis experience of clinical experts and cardiac CT images and image segmentation model,GPUs parallel computing capacity was employed to realize cardiac structure segmentation and 3D visualization.Results Cardiac CT images could be segmented accurately,precisely and robustly in 3D visualization.Cardiac diagnosis-assisted system was realized based on GPUs visualization.Conclusion GPUs parallel computing capacity of CAD system is utilized to process and segment medical images,which raise the running efficiency.