This monograph introduces a Magnetic Resonance Imaging (MRI) reconstruction, which gives rapid achievement and it is more advantageous for many clinical application. Functional brain imaging in humans as we presently know it began when the experimental strategies of cognitive psychology were combined with modern brain imaging techniques, first positron emission tomography (PET) and then functional magnetic resonance imaging (fMRI), to examine how brain function supports mental activities. This marriage of disciplines and techniques galvanized the field of cognitive neuroscience, which has rapidly expanded to include a broad range of the social sciences as well as basic scientists interested in the neurophysiology, cell biology and genetics of the imaging signals. While much of this work has transpired over the past couple of decades, its roots can be traced back more than a century.
In 1927 Egas Moniz, professor of neurology in Lisbon and Nobel Prize in Physiology or Medicine winner in 1949, introduced cerebral angiography, whereby both normal and abnormal blood vessels in and around the brain could be visualized with great accuracy. In its early days this technique likewise carried both immediate and long-term risks, many of them referable to deleterious effects of the positive-contrast substances that were used for injection into the circulation. Techniques have become very refined in the past few decades, with one in 200 patients or less experiencing ischemic sequelae from the procedure. As a result, cerebral angiography remains an essential part of the neurosurgeon's diagnostic imaging armamentarium and, increasingly, of the therapeutic armamentarium as well, in the neurointerventional management of cerebral aneurysms and other blood-vessel lesions and in some varieties of brain tumor.
This monograph introduces a Magnetic Resonance Imaging (MRI) reconstruction, which gives rapid achievement and it is more advantageous for many clinical application. This reduces the scanning cost as well as image reconstructed in very less time, this will be advantage for real time technology. This paper confer a deep learning based plan for reconstruction of MRI images. In our Generative Adversarial Network (GAN) designed a generator which gives the better enhancement like texture smoothness, and high resolution to MRI images, Also find the ` In addition including the frequency domain information to embed similarity in both the images using parameter Structural Similarity Index (SSIM). Also performed radon transform to find the structural similarity of images with enhancement, accuracy and test whether the images are real or fake. Compared to other method, our GAN method provides superior reconstruction.
Magnetic Resonance Imaging (MRI) is mostly used for scanned imaging application. MRI scans the body tissue and gives excellent contrast, which also includes the structural and functional information of whole body. The main drawback of MRI is slightly slow speed because data samples cannot directly collected in image, but rather than in specific area. This data has special time period information, it collects data serially. For high quality of image upto 512 lines of data needs. During MRI patients movement and other physiological motion gives slow speed. MRI data samples are obtain sequentially in k-space. K-space determined by Nyquist-Shannon sampling criteria. Under-sampled k-spaces, gives the result; acceleration rate is directly proportional to the under sampling ratio. In that principle only less amount of data required.
I would like to acknowledge the contribution of Priyanka Milind Shende, M.Tech student .I personally thank our Head of Department Dr.P.T.Karule who has given full support for completing this work. I also acknowledge the support provided by my Co-author Dr. N.P. Patidar for his contribution in this monograph. Last but not the least I thank my wife Mrs.Pushpalata Pawar for her support.
To read further you can click on Download link.