脑中风辅助诊断系统设计与实现

发布时间：2018-04-05 21:41

本文选题：计算机辅助诊断　切入点：脑中风　出处：《浙江大学》2016年硕士论文

【摘要】：随着现代人压力增大,生活节奏加快,脑中风的发病率越来越高,越来越成为人类的一大威胁。如何快速判断病人患病程度以确认合适的治疗方案至关重要,传统诊断方法依托于医生的经验,误差较大。虽然斯坦福大学开发出计算灌注弥散失调率的软件,但是该软件完全是一个黑盒,并没有提供修改一些经验常数的接口,医生无法对其中的一些常数根据地域以及人种差别进行修改；并且计算时间过长,无法应用于临床诊断。因此医生需要一个可调整经验常数,能获得可供研究的数据,并能够应用于临床诊断的脑中风辅助诊断系统。基于此,本文根据通过灌注弥散参数进行中风诊断的原理,依据黑盒的功能重新设计实现了一个透明白盒的脑中风辅助诊断系统,并为下一步将部分计算移植到GPU以实现临床实时诊断打下坚实基础。该系统通过磁共振设备拍摄出的数据,设计并实现了图像预处理操作、动脉输入函数获取、病变体积计算等几大模块。通过运动校正和时间校正完成图像预处理,改进AIF选取算法求取AIF位置,以及通过体积计算完成灌注弥散失调率的计算。本系统已经使用浙二医院提供的数据,得到了较为满意的结果。
[Abstract]:With the increasing pressure of modern people and the quickening pace of life, the incidence of cerebral apoplexy is becoming higher and higher, and it has become a great threat to human beings.It is very important to judge the patient's degree of illness quickly to confirm the appropriate treatment plan. The traditional diagnosis method relies on the doctor's experience and the error is great.Although Stanford University developed software to calculate perfusion dispersion misalignment, the software is completely black box and does not provide an interface to modify some empirical constants.Doctors are unable to modify some of these constants based on geographical and ethnic differences; and the calculation time is too long to be applied to clinical diagnosis.Therefore, doctors need an adjustable empirical constant, which can be used in the clinical diagnosis of stroke diagnosis system.Based on this, according to the principle of stroke diagnosis by perfusion dispersion parameters, according to the function of black box, a transparent white box stroke diagnosis system is redesigned.And lay a solid foundation for the next step to transplant part of the calculation to GPU to achieve clinical real-time diagnosis.The system designed and implemented several modules such as image preprocessing operation, arterial input function acquisition, lesion volume calculation and so on.The image preprocessing is completed by motion correction and time correction, the position of AIF is obtained by improved AIF selection algorithm, and the misadjustment rate of perfusion dispersion is calculated by volume calculation.This system has already used the data provided by Zhejiang No. 2 Hospital, and obtained more satisfactory results.
【学位授予单位】：浙江大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：R743.3;TP391.7

【相似文献】