Please use this identifier to cite or link to this item:
Author(s): Raquel S. Alves
Diogo Borges Faria
Durval Campos Costa
João Manuel R. S. Tavares
Title: Analysis of gated myocardial perfusion SPECT images based on computational image registration
Issue Date: 2015
Abstract: Myocardial perfusion is commonly studied based on the evaluation of the leftventricular function using stress-rest gated myocardial perfusion single photon emissioncomputed tomography (GSPECT), which provides a suitable identification of the myocardialregion, facilitating the localization and characterization of perfusion abnormalities. Theprevalence and clinical predictors of myocardial ischemia and infarct can be assessed fromGSPECT images.Here, techniques of image analysis, namely image segmentation and registration, areintegrated to automatically extract a set of features from myocardial perfusion SPECT imagesthat are automatically classified as related to myocardial perfusion disorders or not. Thesolution implemented can be divided into two main parts: 1) building of a template image,segmentation of the template image and computation of its dimensions; 2) registration of theimage under study with the template image previously built, extraction of the image features,statistical analysis and classification. It should be noted that the first step just needs to beperformed once for a particular population. Hence, algorithms of image segmentation,registration and classification were used, specifically of k-means clustering, rigid anddeformable registration and classification.The computational solution developed was tested using 180 3D images from 48 patients withhealthy cardiac condition and 72 3D images from 12 patients with cardiac diseases, whichwere reconstructed using the filtered back projection algorithm and a low pass Butterworthfilter or iterative algorithms. The images were classified into two classes: abnormalitypresent and abnormality not present. The classification was assessed using fiveparameters: sensitivity, specificity, precision, accuracy and mean error rate.The results obtained shown that the solution is effective, both for female and male cardiacSPECT images that can have very different structural dimensions. Particularly, the solutiondemonstrated reasonable robustness against the two major difficulties in SPECT imageanalysis: image noise and low resolution. Furthermore, the classifier used demonstrated goodspecificity and accuracy, Table 1.
Subject: Ciências Tecnológicas, Ciências da engenharia e tecnologias
Technological sciences, Engineering and technology
Scientific areas: Ciências da engenharia e tecnologias
Engineering and technology
Source: 2015 IEEE 4th Portuguese Meeting on Bioengineering (ENBENG)
Document Type: Artigo em Livro de Atas de Conferência Nacional
Rights: openAccess
Appears in Collections:FEUP - Artigo em Livro de Atas de Conferência Nacional

Files in This Item:
File Description SizeFormat 
99685.pdfAbstract135.45 kBAdobe PDFThumbnail

This item is licensed under a Creative Commons License Creative Commons