Neurological disorders diagnosis makes use of several diagnostic examinations, including cerebral medical imaging. In this field, two of the most widely used examinations are brain PET (Positron Emission Tomography) and MR (Magnetic Resonance). We present a technique and a tool for the integration of morphologic and functional information from brain MR and PET images into a single fused image. The registration of the image is based on the elliptic Fourier descriptors, while the different images have been merged using the multiresolution analysis based on the wavelet transformation. The Graphic User Interface is implemented in the MATLAB® environment. Two experienced neuroradiologists evaluated the accuracy of our approach through visual inspection of contours continuity and using some mathematical indices computed, such as universal image quality index, root mean square error, and mutual information metrics. The tool aims at supporting the doctor during the investigation of metabolic activity within the anatomically defined abnormality. This is usefull in the diagnosis of many neurodegenerative diseases and in the planning and follow-up of radiation treatment.
A Tool for multimodal fusion of brain PET and MR
CAPOBIANCO, Giovanni;FASANO, Fausto;
2013-01-01
Abstract
Neurological disorders diagnosis makes use of several diagnostic examinations, including cerebral medical imaging. In this field, two of the most widely used examinations are brain PET (Positron Emission Tomography) and MR (Magnetic Resonance). We present a technique and a tool for the integration of morphologic and functional information from brain MR and PET images into a single fused image. The registration of the image is based on the elliptic Fourier descriptors, while the different images have been merged using the multiresolution analysis based on the wavelet transformation. The Graphic User Interface is implemented in the MATLAB® environment. Two experienced neuroradiologists evaluated the accuracy of our approach through visual inspection of contours continuity and using some mathematical indices computed, such as universal image quality index, root mean square error, and mutual information metrics. The tool aims at supporting the doctor during the investigation of metabolic activity within the anatomically defined abnormality. This is usefull in the diagnosis of many neurodegenerative diseases and in the planning and follow-up of radiation treatment.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
