人体颅脑MRI图像中下颌骨的分割及三维建模

目的:在目前的影像医疗诊断中,仅凭观察二维CT、MRI图像是很难实现准确的确定病变体的空间位置、大小、几何形状及与周围生物组织的空间关系的。本研究利用CT、MRI图像数据,帮助医生对病变体及其人体组织感兴趣的区域进行分割提取,用于医学图像的三维显示、三维重建。材料与方法:应用Mimics软件,对132层,层间距为1.0mm的MRI扫描图像进行图像去噪、分割和平滑等处理,最终创建三维模型。结果:建立了更为精确,应用更为广泛的下颌骨三维模型。结论:运用Mimics对MRI医学图像进行分割,重建三维模型,能全面显示病变的病理解剖改变,为临床治疗提供精细的影像学信息,从而使重建后的三维模型可清晰地再现病灶与周围组织的解剖关系,大大增强了医学图象分析系统的临床实用价值。

Mandible Segmentation and Three-Dimensional Modeling in Human Craniocerebral MRI Images

Objective： In current iconographic medical diagnoses, it is difficult to accurately determine the spatial location, size, geometric shape of a pathological body and its spatial relationship with surrounding biological tissues by merely observing two-dimensional CT and MRI images. In this study, by processing and CT or MRI data, pathological body and other parts of human body that interest doctors would be segmented and extracted for the three-dimensional demonstration and reconstruction. Methods：Build three-dimensional models after denoising, segmenting, smoothing MRI scanned images of 132 layers with interlamellar spacing of 1.0 mm, and so on by applying Mimics software. Results： The three-dimensional mandible models with more precise and wider application have been built. Conclusion： With the aid of Mimics software, a comprehensive image of a pathological body＇s anatomic pathological changes can be obtained easily, which provide precise iconographic information for clinical treatment. Then, the anatomic relationship between a lesion and its surrounding tissues can be clearly reshow, which can enhance the clinical application of medical image analytical system.