Font Size: a A A

Design And Realization Of SPECT Thyroid ROI Extraction And Weight Measurement System Based On Image Technology

Posted on:2018-11-24Degree:MasterType:Thesis
Country:ChinaCandidate:S DengFull Text:PDF
GTID:2334330518999060Subject:Engineering
Abstract/Summary:PDF Full Text Request
As one of the representatives in nuclear medical imaging,Single-Photon Emission Computed Tomography(SPECT)has been widely applied in clinical diagnosis after thirty years of rapid development since it came into being.The related image processing technology for SPECT has also been paid more and more attention.SPECT is able to detect organ metabolism and diagnose the disease in the earlier stage.Affected by the imaging characteristics,equipment environment,patient's condition,particularity of medical images and other factors,image is attenuated and disturbed,which make it difficult to accurately segmenting the region of interest and extracting feature information.In clinical disease diagnosis,doctors need to manually segment the Region of Interest(ROI)by observing the SPECT image,so as to extract the characteristic information and carry out the diagnosis and treatment of the disease.The process is time-consuming and depends on the doctor's personal experience knowledge.This paper presents an optimal method of SPECT thyroid ROI extraction and weighing according to the clinical actual needs.Firstly,aiming at the shadow caused by gamma photon scattering at the edge of the region of interest,an ROI segmentation algorithm based on OTSU and morphology is proposed.The algorithm modifies the local gray level of ROI based on the OTSU segmentation theory,overcomes the border blurred problem,and uses the morphological operation to optimize the ROI structure so make it closer to the real situation at the same time.The characteristics information of thyroid is extracted on the basis of the above operation.In this paper optimized weight factor is presented based on the grayscale distribution of ROI so as to improve the existing weighing method,which overcomes the shortcomings of lack of spatial information.The experimental results show that the optimal method of SPECT thyroid ROI extraction and weighing can accurately locate the region of interest and provide a high quality reference image for clinical evaluation.The thyroid weight measured by this method is consistent with the gold standard in clinical diagnosis.This paper uses MATLAB graphical user interface development environment to design and implement SPECT thyroid weighing system which contains the original data input,preprocessing,ROI segmentation,information calculation,thyroid weighing and other processes,It helps to provide more comprehensive medical image information,identify pathological information and lesion site,improve the accuracy of disease diagnosis.It is also intuitive and easy to operate,which users can easily grasp,it greatly improve the current diagnostic efficiency of clinical thyroid disease.In addition,by compile the program into an executable file and integrated with the MATLAB Component Runtime,the applicability of the software is enhanced,so the system can be used in more platforms,which is of great significance to the application and popularization of the system.
Keywords/Search Tags:Single electron emission computed tomography, digital image processing, mathematical morphology, ROI segmentation, thyroid weight, Graphical User Interface
PDF Full Text Request
Related items