结合Cine-和Tagged-MRI的发音器官运动观测与分析

发布时间：2018-06-11 11:43

本文选题：磁共振成像 + 轮廓信息　；参考：《天津大学》2016年硕士论文

【摘要】：在当今语音产生机理研究领域中,磁共振成像(MRI)技术是一种特别有前景的成像技术之一。静态磁共振成像和实时磁共振成像已经被广泛应用于语音研究领域以获得所有发音器官运动信息和测量发音过程中各个发音器官的形变信息。但是,目前多数的磁共振成像技术只是能够对发音器官的整个表面轮廓进行成像,与其他成像技术相比,如磁感应方法,它们的劣势在于缺乏特定的点的运动信息。对目前这一难题的一种可行性解决方案就是:结合两种现有的、不同特点的成像技术,使最终获得的数据既有发音器官的轮廓形变信息,又有发音器官上特定的点的运动信息。在本文中,我们提出了一种结合Cine-和Tagged-MRI两种磁共振成像技术的方法来实现上述方案。要做到这一点,两种成像技术就必须要在同一个扫描周期之间进行成像的同步。在Tagged-MRI数据中,舌头表面与标定线的交点是不随时间的变化而变化的,我们就在Tagged-MRI上手动的进行标记这些点的运动信息,并且将所得的所有点提取并映射到Cine-MRI数据上。在此过程中,我们最大限度地减少了映射所产生的误差。结果表明,通过这种方案来合成既有舌头轮廓形变信息,又有舌头上特定点的运动信息的数据是成功的。在得到合成的数据集之后,我们又分析了舌体的相对大小和发音过程中舌体变形的关系,发现相对舌体大小较大的被试,其舌体的运动速度相对较慢。
[Abstract]:Magnetic resonance imaging (MRI) is one of the most promising imaging techniques in the field of speech generation mechanism. Static magnetic resonance imaging and real-time magnetic resonance imaging have been widely used in the field of phonological research to obtain the movement information of all vocal organs and to measure the deformation information of each organ in the process of pronunciation. However, most magnetic resonance imaging techniques can only image the whole surface profile of the vocal organ. Compared with other imaging techniques, such as magnetic induction method, their disadvantage lies in the lack of specific point motion information. A feasible solution to this problem is to combine two existing imaging techniques with different characteristics, so that the final data can be obtained not only with the contour deformation information of the phonetic organ, but also with the movement information of the specific point on the phonetic organ. In this paper, we propose a method combining Cine- and Tagged-MRI to realize the above scheme. To do this, two imaging techniques must synchronize the imaging between the same scanning cycles. In the Tagged-MRI data, the intersection point between the tongue surface and the calibration line does not change with time. We manually label the motion information of these points on Tagged-MRI, and extract and map all the points to the Cine-MRI data. In the process, we minimize the error caused by the mapping. The results show that it is successful to synthesize both the deformation information of tongue contour and the motion information of specific points on tongue by this scheme. After obtaining the synthesized data set, we analyze the relationship between the relative size of the tongue body and the deformation of the tongue body during the process of pronunciation. It is found that the relatively large tongue body size of the subjects has a relatively slow tongue body movement speed.
【学位授予单位】：天津大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：R767.92;TP391.41

【相似文献】