Study On The Method Of Patellar Dislocation Assisted Diagnosis Based On Image Processing

Patella injury is much more common in young and the people who do more exercise,it easily leads to the functional disorders of the knee joints and the patella dislocation is one of the common clinical manifestation,if it cannot be treated timely and correctly,it will bring a series of sequel like patellar instability,recurrent dislocation,etc.Moreover,the recurrent rate is as high as 20%-40%,which can bring great inconvenience to the patients’ life.The appearance and application of the combination of computer technology and medical diagnostic CT(computerize tomography),MRI(Magnetic resonance imaging),PET(Positron emission tomography)and DSA(Digital subtraction angiography)and other medical imaging technology allows doctors to see the lesion more intuitively and clearly.It can greatly improve the accuracy and cure rate of clinical diagnosis,which can reduce the doctors’ workload and improve the doctors’ efficiency.In this paper,we propose an auxiliary diagnosis system for patellar dislocation based on Matlab.Firstly,Gauss filter,the local contrast enhancement algorithm are chose to preprocess the patella CT image to reduce image noise and enhance the image contrast though comparing analysis of several classic methods of image filtering,contrasting enhancement algorithm and the advantages and disadvantages of several commonly used methods of image segmentation.On this basis,the image segmentation algorithm based on edge detection and the bounding box algorithm are used to segment the contour of the target region in the CT image.Image segmentation algorithm and bounding box algorithm can segment the target counter of the CT image currently;finally,this paper proposes the extraction method based on key points of different areas through the analysis of the defect of the Harris corner detection algorithm and gradient detection method.The patella femoral axis face angle,distance and patellar tilt angle can be gained through the data and quantitative measurement which base on the plane fitting method with the key points that extracted through the image edge contour segmentation,and these datas can provide the effective support for the diagnosis of patellar dislocation.This study can obtain more accurate measurement data for patellar contour rule and the CT images whose feature points are obvious and also can effectively assist the diagnosis of patellar dislocation through the comparison between final measurement data and the clinical judgment.It also proves that this research has certain application value.Through the comparison between the measured data and the judgment standard,it can be known this computer aided diagnosis system researched and designed in this paper has high practical significance,and clinical value for the patellar outline that is positive and the feature points that are very obvious.