Study Of Head Motion Blur CT Image Restoration Technology Based On Lucy-Richardson Algorithm

Computed Tomography imaging(CT) has been widely used in medical diagnosis.Based on the CT images, doctors can intuitively learn the variation of the details in the patient's internal organs and the organization structure so as to make correct judgement on the patient's condition.However, in the process of CT scanning, the patient's body motion will destroy the consistency and the integrity of scanning data, which generates blurring and distortion of CT images.This kind of image blurring caused by body motion can reduce the accuracy of CT image so as to influence the doctor to diagnose the illness, which can also restrict the development of CT imaging technology. Eliminating the influence of motion blur and getting clear CT images plays an important role in the medical diagnosis and the perfecting and developing of CT technology.On the basis of detailly analyzing the existing restoring technology on motion blur CT image, this paper proposes a restoring method for head motion blur CT image based on the theory of motion blurring.First, a new mathematical model of motion blurring of CT image caused by the head motion is established,and a hypothesis is put forward,which assumes that the head motion is uniform linear motion during CT scan imaging, on the basis of that the point spread function of head motion is established;Then, the Lucy-Richardson algorithm with Gauss filter is used in the restoring of head CT image,so that the influence of the motion blur is eliminated and the clear head CT image is created. In addition, in order to obtain the motion parameters for establishing the point spread function of head motion, this paper proposes a calculation method of motion parameters based on the characteristic information of motion artifacts, first the Sobel algorithm is used in edge detection of head motion blur CT image, and then the key points of motion artifacts in the resulting binary image are marked in manual way, and then the key points are composited into straight lines by least squaremethod, from which the slope of straight line composited of artifact and the spacing of two adjacent artifacts is calculated, on this basis the motion angle and motion length corresponding to the point spread function is calculated. Compared with other calculation methods of the motion parameters that some other auxiliary equipments is used, the advantages of the methods that the motion parameters can be obtained according to the characteristic of CT image itself directly, which has high practicability.The restoring experiment and results analyzing for the clinical head motion blur CT image proves that the head motion blur CT image can be restored effectively in this paper's method, which has high stability and practicability.