海量医学图像数据三维可视化表面重建方法

3-D Visualization Surface Reconstruction Method for Massive Medical Image Data

目的为了提高海量医学图像的识别能力,需要进行海量医学图像数据三维可视化表面重建。提出基于谱分析的海量医学图像数据三维可视化表面重建方法。方法构建海量医学图像数据的采集模型,提取海量医学图像数据三维图谱特征量,采用边缘轮廓检测的方法进行海量医学图像数据三维可视化特征分解,建立海量医学图像数据的模糊相关性特征分布结构模型。根据海量医学图像的色彩特征分布进行医学图像的图谱特征分析,建立海量医学图像数据的三维结构重组模型,根据对海量医学图像的三维结构分布式重组进行动态滤波检测,提取海量医学图像的谱特征量,根据谱特征分布进行海量医学图像的多维重建,结合三维可视化特征分析的方法,实现对海量医学图像数据的可视化表面重建和优化识别。结果采用该方法进行海量医学图像数据三维可视化表面重建的特征分辨能力较好,对图像数据的细节结构特征识别能力较强。结论研究方法提高了对海量医学图像的检测识别能力。

Objective To improve the recognition ability of mass medical images,it is necessary to reconstruct the 3-D visualization surface of mass medical image data.Based on spectral analysis,a 3-D visualization surface reconstruction method for massive medical image data is proposed.Methods The collection model of massive medical image data was constructed,and the feature quantity of 3-D Atlas of massive medical image data was extracted.The method of edge contour detection was used to decompose the 3-D visualization feature of massive medical image data,and the distribution structure model of fuzzy correlation feature of massive medical image data was established.Based on the color distribution of the mass medical image,the map features of the medical image were analyzed,and the mass medical map was established.The 3-D structure reconstruction model of massive medical image data was established.According to the 3-D structure distributed reconstruction of massive medical image,the dynamic filtering detection was carried out,and the spectral feature quantity of massive medical image was extracted.According to the spectral feature distribution,the multidimensional reconstruction of massive medical image was carried out.The visualization surface reconstruction and optimal recognition of massive medical image data were realized.Results The results showed that the proposed method had good feature resolution ability for 3-D visualization surface reconstruction of massive medical image data,and had strong recognition ability for the detailed structure features of image data.Conclusion The proposed method improves the detection and recognition ability of mass medical image.