by
Christina Hwang, Contributing Reporter | May 31, 2016
A new framework for the detection of Alzheimer’s using spherical brain mapping (SBM), a technique that allows physicians to visualize a 3-D MR image as a 2-D image, has been proposed by researchers from the University of Grenada, Spain.
In SBM, each pixel of the image represents a certain measurement in a specific direction, and in the study the researchers had focused on six types of measurement. Three statistical; average, entropy, and kurtosis, and three morphological; cortical thickness, topology of the brain’s surface, and the number of folds, were measured.
The researchers obtained structural images of the brain from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and developed the SBM framework. After establishing the framework, they then tested SBM on the images ADNI provided.
Ad Statistics
Times Displayed: 122196
Times Visited: 7083 MIT labs, experts in Multi-Vendor component level repair of: MRI Coils, RF amplifiers, Gradient Amplifiers Contrast Media Injectors. System repairs, sub-assembly repairs, component level repairs, refurbish/calibrate. info@mitlabsusa.com/+1 (305) 470-8013
“We analyzed the discriminative power of the projected maps using both visual and classification analysis,” Francisco Jesús Martínez-Murcia, a researcher from the study, department of signal theory, telematics and communications at the University of Granada, told HCB News.
The team had tested this technique against an MR database of 180 Alzheimer’s patients and 180 individuals without the disease, and found that the most appropriate maps to be used in Alzheimer’s were the ones that measured the average of the intensities of the MR image, which can be white, grey or black, and a textural measure called local binary patterns.
According to Martínez-Murcia, the average or local binary patterns maps have approximately 90 percent accuracy while other maps such as surface or kurtosis have a lower performance of about 70 percent.
“SBM provides visual maps that represent internal features of the brain, and therefore could be very useful to obtain that kind of high-level knowledge at a glance, without having to navigate the different coordinates of the 3-D images,” said Martínez-Murcia.
The maps also reduce computation load while maintaining a large amount of the original information provided by the 3-D image.
“[The map] is still a summary of all the information and [can] be used as a complementary tool to well-established methodology currently used in the diagnosis of Alzheimer’s,” Martínez-Murcia said.
