论文部分内容阅读
目的依据中国虚拟人数据集对尾状核、豆状核和丘脑进行三维重建并与临床影像数据进行融合,为神经外科疾病的诊断与治疗提供形态学依据。方法以中国虚拟人数据集为依据,采用图像透明的方法对尾状核、豆状核和丘脑进行手工分割,在平滑处理中运用了腐蚀与膨胀的方法,并把分割后的神经结构与病人磁共振图像进行融合,最后用表面绘制与体绘制相结合的方法进行三维重建。结果重建后的图像可清晰显示尾状核、豆状核、丘脑各个神经结构的形态、位置和毗邻关系。可在三维空间中绕任意轴旋转任意角度,从不同的方向进行观察。融合后的图像能同时显示虚拟结构与病人MRI图像。结论用三维可视化与MRI图像融合方法获得的图像资料,对于解剖学的教学研究以及神经外科疾病的诊断与治疗具有重要的参考价值。
Objective To reconstruct the caudate nucleus, lenticular nucleus and thalamus according to the Chinese virtual human dataset and merge with the clinical imaging data to provide morphological evidence for the diagnosis and treatment of neurosurgical diseases. Methods Based on the Chinese virtual human data set, images were used to segment caudate nucleus, lentiform nucleus and thalamus by using the method of image transparency. In the smoothing process, the method of erosion and dilation was applied. The segmented neural structure and patient MRI image fusion, and finally with surface rendering and volume rendering method of combining three-dimensional reconstruction. Results The reconstructed images clearly show the morphological, positional and adjacent relationships of each neuronal structure of caudate nucleus, lentiform nucleus and thalamus. Rotate any angle around any axis in 3D, looking from different directions. The fused image shows both the virtual structure and the patient’s MRI image simultaneously. Conclusion The image data obtained by the method of 3D visualization and MRI image fusion have important reference value for anatomical teaching research and diagnosis and treatment of neurosurgical diseases.