基于Otsu法的内耳MRI图像三维快速分割的研究

3D Visualization of Inner Ear Based on Otsu Method

目的为了获取独立的内耳三维模型,探讨一种半自动、快速的内耳分割方法。方法将MRI图像导入3D slicer软件,基于Otsu法,采用半自动分割内耳,并结合meshlab软件进行内耳后期精细处理。结果对30例MRI影像数据进行分割,一侧内耳完成半自动分割仅需3钟左右,内耳三维模型图像清晰,内耳信息保留完整。结论Otsu法进行内耳半自动分割操作简单、快捷,内耳模型结构完整,三维成像效果好。

Objective To report a semi-automatic and fast segmentation method for the purpose of obtaining an independent 3 D model of the inner ear. Methods MRI images were imported and the inner ear was processed by semi-automatic segmentation using the 3 D Slicer software and then combined with the Otsu software. The inner ear was finally fine-tuned with the Mesh Lab software. Results In processing 30 images, segmentation of the inner ear on one side took about 3 minutes using the semi-automatic method, yielding clear three-dimensional images of the inner ear with no loss of information. Conclusion The Otsu method is a simple and fast inner ear semi-automatic segmentation tool for constructing and the inner ear model with good three-dimensional effects.