Research On Renal Image Segmentation And Functional Quantitative Assessment Algorithms

Kidney is one of the important organs of the human body,which maintains the metabolic balance of the human body.Kidney disease has become one of the major diseases that threaten human health.Nowadays,the number of kidney patients with chronic diseases in China has reached 120 million,and patients with renal tumors also occupy the second place in the tumors of urology in China,both of the two types can lead to a decrease in renal function.The best clinically evaluation index of renal function is the glomerular filtration rate（GFR）,the current clinical detection methods widely used is combining the single photon emission tomography ^99mTc-DTPA renal dynamic imaging method and the Gates method to evaluate the renal function,and the method is taken as the gold standard.However,clinically,most of the region of interest is obtained by manually delineated,it has no uniform standard and is inefficient.Especially when the kidney is severely damaged,the brightness of the radionuclide will not decline by filtering of the glomerular glomeruli,and the contour of the region of interest of the kidney is blurred and difficult to be segmented.Next,the paper will conduct research and innovation on the following aspects:（1）Aiming at the problem that the current renal dynamic imaging evaluation of renal functions rely too much on manual labeling and has low time efficiency,an automated quantitative evaluation method for renal dynamic imaging is proposed.The method can automatically obtain renal region of interest,background ROI and aortic ROI,then combines the Gates method to obtained the renal GFR and the total renal GFR.By using the method,renal function assessment will be obtained integratedly and automatically.The clinical trial result shows that the proposed automated evaluation method can improve the automation level and improve the evaluation accuracy in a short period of time.This method can give effective help on clinical diagnosis and adjuvant treatment.（2）For the severely damaged kidney,the development of the dynamic imaging of the kidney is weak and the contrast of the background area is decreased.The ROI contour of the kidney is difficult to distinguish.Aiming at the problem,this paper proposed a CV improved Level set model combined the prior shape knowledge don`t required to be reinitialized.First construct the normal kidney shape as a template,then contain the counted kidney-related parameters in the kidney template,then we can optimize the curve evolution performance by optimizing the parameter?of the adaptive shape constraint term in the energy function,finally,we can obtain ROI of secerely damaged kidney images.Experiments show that for severely damaged kidneys,the algorithm can get better ROI segmentation results.After calculating GFR by the Gates method,linear regression analysis and correlation analysis were performed between the calculated results and clinical reports,the statistical result shows that there has a high correlation between them.The solution of this problem can provide evidence for the diagnosis and treatment as well as and the changes of renal function of this type of patients,and can provide a uniform evaluation criteria so as to reduce the clinical inaccuracy of renal function caused by different subjective will.（3）Based on the algorithm proposed in this paper,an automated quantitative evaluation system for medical renal dynamic imaging was designed and implemented.The system includes four modules:input source image module,image preprocessing module,ROI acquisition module and renal function evaluation module.The system not only can replace the manual delineation with obtaining the ROI automatically to ensure its accuracy and timeliness,but also can obtain the kidney ROI from renal dynamic imaging images of different kidney injurie according to the pharmacokinetic principle.Then plot the time-radioactivity curve by combining the ROI parameters with the Gates method,and calculate the GFR values??of the kidney and total kidney respectively,so as to give an auxiliary support to the doctor`s diagnosis of renal function,and can give a guidance to the future treatment plans.Finally the experimental results show that the system is easy to operate,its`evaluation process of GFR value is complete and reusable,and has a good clinical application value.