基于图像处理的髌骨脱位辅助诊断方法研究

发布时间：2018-04-17 21:54

本文选题：髌骨脱位及半脱位 + 髌骨CT图像　；参考：《华北水利水电大学》2017年硕士论文

【摘要】：髌骨损伤常见于运动较多、量较大的群体当中,尤其是年轻人,易引起膝关节功能性障碍,而髌骨损伤最常见的临床表现则是髌骨脱位,若未得到及时正确的处理,则会带来髌骨不稳、复发性脱位等一系列后遗症,复发率高达20%-40%,给病人生活带来极大不便。计算机技术与医疗诊断相融合的CT(computerized tomography)、MRI(Magnetic resonance imaging)、PET(Positron emission tomography)及DSA(Digital subtraction angiography)等医学成像技术的出现及应用使得医生能够更加直观、清晰地看到病变部位,可极大地提高临床诊断正确率和病变治愈率,减少医生工作量,提高医生工作效率。本文设计了一个髌骨脱位辅助诊断系统。首先对比分析研究几种经典的图像滤波方法、对比度增强算法及几种常用图像分割方法的优缺点,选用高斯滤波、基于局部区域的对比度增强算法对髌骨CT图像进行预处理,以降低图像噪声、增强图像对比度;在此基础上,依次利用基于边缘检测的图像分割算法、包围盒算法精确分割出髌骨CT图像中目标区域的边缘轮廓;最后通过具体实验分析Harris角点检测算法、梯度检测方法选取边缘轮廓关键点的缺陷,提出了基于不同区域的关键点提取方法。利用该方法对图像分割阶段得到的边缘轮廓关键点进行提取,最终以关键点为基准采用平面拟合方法,通过数据量化及计算测量,得到髌骨股骨中轴面夹角、距离及髌骨倾斜角,为髌骨脱位诊断提供有效辅助。通过实验得到的测量数据与医学临床判断标准的对比。本文研究对于髌骨轮廓规则、特征点较为明显的髌骨CT图片,可获得较为精确的测量数据,能够有效辅助髌骨脱位诊断。最后,通过测量数据与判断标准的实验对比总结可知:对于髌骨轮廓较正、特征点较为明显的髌骨CT图片,本文研究设计的髌骨脱位计算机辅助诊断系统具有较高的正确性、实践意义及临床使用价值。
[Abstract]:Patellar injuries, especially among young people, tend to cause functional disorders of the knee joint, and the most common clinical manifestation of patellar injuries is dislocation of patella, if not properly dealt with in time, the most common clinical manifestation of patellar injury is dislocation of patella, especially among young people, and the most common clinical manifestation of patellar injury is dislocation of patella.Will bring patella instability, recurrent dislocation and other sequelae, recurrence rate up to 20-40, bring great inconvenience to the patient's life.The appearance and application of medical imaging techniques, such as CT(computerized tomphography and DSA(Digital subtraction emission, which combine computer technology with medical diagnosis, enable doctors to see the location of lesions more intuitively and clearly.It can greatly improve the correct rate of clinical diagnosis and cure rate of diseases, reduce the workload of doctors and improve the efficiency of doctors.An auxiliary diagnosis system for patellar dislocation is designed in this paper.Firstly, the advantages and disadvantages of several classical image filtering methods, contrast enhancement algorithm and several common image segmentation methods are compared and studied. Gao Si filter and contrast enhancement algorithm based on local region are used to preprocess patellar CT image.In order to reduce image noise and enhance image contrast, image segmentation algorithm based on edge detection and bounding box algorithm are used to accurately segment the edge contour of patellar CT image.Finally, the Harris corner detection algorithm and the gradient detection method are analyzed in detail to select the defects of the key points of the edge contour, and a key point extraction method based on different regions is proposed.This method is used to extract the key points of the edge contour in the image segmentation stage. Finally, the angle of the central axial plane of the patella femur is obtained by using the plane fitting method, which is based on the key points, and through the quantization of the data and the calculation and measurement.Distance and patellar angle provide effective assistance for diagnosis of patellar dislocation.The comparison between the measured data obtained from experiments and the standard of medical clinical judgment.In this paper, the patellar CT images with regular patellar contours and obvious characteristic points can obtain more accurate measurement data and can effectively assist in the diagnosis of patellar dislocation.Finally, by comparing the measured data with the standard of judgment, we can conclude that the computer aided diagnosis system for patellar dislocation is correct for the CT images of patella whose profile is more accurate and characteristic points are more obvious.Practical significance and clinical use value.
【学位授予单位】：华北水利水电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R684;TP391.41

【参考文献】