Soroor Kalantari1, Fardin Samadi Khosh Mehr2, Mohammad Soltani1, Mehdi Maghbooli3, Zahra Rezaei4, Soheila Borji1, Behzad Memari1, Mohammad Bayat1, Behnaz Eslami5, and Hamidreza Saligheh Rad6
1Department of Radiology, Zanjan University of Medical Science, Zanjan, Iran (Islamic Republic of), 2Department of Medical Physics and Biomedical Engineering, Tehran University of Medical Sciences, Tehran, Iran (Islamic Republic of), 3Department of Neurology, Zanjan University of Medical Science, Zanjan, Iran (Islamic Republic of), 4Department of Computer and Electrical Engineering, University of Kashan, Kashan, Iran (Islamic Republic of), 5Tehran Islamic Azad University, Tehran, Iran (Islamic Republic of), 6Quantitative MR Imaging and Spectroscopy Group, Research Center for Molecular and Cellular Imaging, Department of Medical Physics and Biomedical Engineering, Tehran university of Medical Science, Tehran, Iran (Islamic Republic of)
This study aims to investigate the use of high-level de-noising
and machine-learning methods applied on ASL-MRI dataset acquired at 1.5T, and
in order to to find important regions in the brain for the classification of
patients with AD and MCI and normal aging. Automated classification and
prediction methods recognizing
perfusion changes in specific subregions of the brain are applied to pseudo-continuous
ASL-derived CBF-maps, predicting the
diagnosis of Alzheimer's disease, mild cognitive impairment, and normal
cognition. Due to alarming prevalence of AD, machine-learning approaches
for ASL- MRI are used to develop computer-aided diagnosis (CAD) tools for
clinical and screening targets, assisting early diagnosis of the AD process.
