MRI Based Pilot Study For3D Thickness Of Bladder Tumor

《Cancer facts&figures2013》indicates that bladder cancer is the fourth mostcommon malignancy in2013in the USA and the sixth most common malignancy in China.The incidence of bladder cancer and its fatality rate is climbing up rapidly. Statistics showthat bladder cancer patients are diagnosed while the tumor is in situ, for which the5-yearssurvival is96%and the5-years survival declines as the tumor developing. Therefor earlyscreening is important for improving the survival rate of patients. An estimated10%to70%of patients with a tumor confined to the mucosa will experience a recurrence within5years. In order to prevent the recurrence of bladder cancer, patients must have follow-uptimely.The cystoscopy is currently the gold standard for bladder cancer diagnosis, but itslimitation such as invasive and existing vision blind area make it hard to be widely used inearly screening and follow-up of bladder cancer. As mentioned above, it is significant tofind a rapid and noninvasive technique for the early screening and follow-up of bladdercancer. Recent research indicates that bladder cavity lesions can induce Morphologicalchanges of bladder wall, especially the bladder wall thickness, which is supposed to be theindicator of Bladder wall lesion. Research on bladder wall thickness is significant to checkout the bladder wall lesion and distinguish tumor tissue and the wall tissue. However thereis huge difference of bladder form in different people and different status, so it is difficultto do the quantitative comparison on bladder wall thickness.Aimed at above problem, in this research, which based on the MRI, we build up acomplete pipeline for bladder wall thickness measurement and analysis. Based on theabove, we compare and analyze the3D thickness between cancerous tissue and bladderwall tissue, using the T2-weighted MR image data of volunteers and patients. Theexperiment results are meaningful for the detection of bladder wall lesion.The major work of this research includes two large divisions:Part one, establishment of the measurement and analysis pipeline for the bladderwall thickness. Key links of this pipeline as follows:1. The appropriate MR series selection and data acquisition: Selected T2weightedfast echo sequenced according to image quality and acquisition time for bladder.2. Bladder inner and outer wall segmentation and extraction: Adopt the coupleddirectional level-set method, which is appropriate for T2weighted MRI, to segment thebladder wall from the MRI data we obtained before.3. Bladder wall3D thickness calculation: Adopt the Laplace partial differentialequation method to define and calculate the3D thickness of bladder wall.4. The thickness value standardization：Because the difference among differentsubject and the same subject with different filling state influences the3D thickness, it isdifficult to compare the thickness quantitatively. In this research, some valuestandardization is used to standardize the3D thickness value of bladder wall.5. Bladder form standardization: After the thickness value standardization,standardizing the bladder form and mapping the thickness value will be done, using theFFD parameter model, in order to achieving the quantitative comparison. Part two, imaging experimental analysis for the variation of bladder wall thickness.Collecting10cases of volunteer and bladder cancer patients respectively and there is nostatistical differences in age between the two groups. Based on the measurement andanalysis pipeline of bladder wall3D thickness, this study test and analyze the3D thicknessbetween bladder wall tissue and tumor tissue.5experiment is included as follow:1. Algorithms validation of the key links in measurement and analysis pipeline：Validate the coupled directional level-set method、the Laplace partial differentialequation method and the FFD parameter model. Results show that three algorithms aresuitable for this study to establish the bladder wall3D thickness measurement and analysispipeline.2. Significance analysis of bladder wall3D thickness between volunteers group andpatients group:In order to find whether there is a difference of the thickness between volunteersgroup and patients group, bilateral t test is used. Result shows that there is a statisticallysignificant difference of thickness between two groups. Patients group have a higheraverage compared with volunteers group.3. Value standardization analysis of bladder wall3D thickness:In order to eliminate the individual difference of bladder that influences the3Dthickness, NBWT and SBWT are used to standardize the3D thickness value of bladderwall. Analyze the differences between the two groups.(1) Using NBWT=1.5as the threshold value to distinguish tumor and wall tissue, canselect principal part of the tumor tissue.(2) Bilateral t test is used to find whether there is a difference of SBWT valuebetween volunteers group and patients group. Result shows that there is a statisticallysignificant difference of SBWT value between two groups. Patients group have a loweraverage compared with volunteers group.(3) In order to analyze whether urine will cause thickness different between the topand bottom of bladder wall, bilateral t test is used. Result shows that there is nostatistically significant difference. 4. Bladder form standardizing, thickness value mapping and the establishment of thehealthy bladder standardized thickness template:The difference of bladder form makes it hard to have quantitative comparisonbetween different bladders. To solve this problem, FFD parameter model is used tostandardize the different bladder form and map the thickness to the standardized model.(1)10cases of volunteer’s bladder data were normalized to ellipsoid. After scaletransform and coordinates translation, average the thickness values of each voxel, thehealthy bladder standardized thickness template is established.(2) In order to find whether there is a significant difference between the top and thebottom of the template, bilateral t test is used. Result shows that there is no statisticallysignificant difference.(3) Do Z-Score analysis with data of patients’ bladder using the template. Resultshows that using a threshold between2and2.5can select the tumor tissue.5. Analysis of the relationship between the bladder filling degree and the3Dthickness:(1) In order to find whether there is a significant difference of the volume of bladderbetween volunteers group and patients group, bilateral t test is used. Result shows thatthere is no statistically significant difference of volume between two groups.(2) In order to find whether there is a relationship between volume and thickness ofthe data of volunteers’, correlation analysis is used. Result shows that there is nostatistically significant correlation between volume and thickness.(3) In order to do a further research of the variation between volume and thickness,do the experiment to analyze the bladder filling dynamic process. Result show that there isa reverse relation between thickness and cubic root of volume. According to the existingdata, a curve of variation is fitted.As mentioned above, our research innovatively established a complete measurementand analysis pipeline of bladder wall3D thickness. On this basis, test and analyze the3Dthickness between bladder wall tissue and tumor tissue, and obtain multiple statistical analysis results, which provide a new way of thinking of detection of bladder wall lesionusing3D thickness.