Hippocampal spatial position evaluation on MRI for research and clinical practice.

Abstract

In clinical practice as well as in many volumetric studies we use different reorientations of the brain position towards x and y axis on the magnetic resonance imaging (MRI) scans. In order to find out whether it has an overall effect on the resulting 2D data, manual hippocampal area measurements and rotation variability of the brain (in two reoriented axes) and the skull were performed in 23 Alzheimer's disease patients and 31 healthy controls. After the MRI scanning, native brain scans (nat) were reoriented into the two different artificial planes (anterior commissure-posterior commissure axis (AC-PC) and hippocampal horizontal long axis (hipp)). Hippocampal area and temporal horn of the lateral ventricle was measured manually using freeware Image J program. We found that 1) hippocampal area of nat images is larger compared to hipp images, area of the nat images is equal to the AC-PC images and area of the hipp images is smaller compared to AC-PC images, 2) hippocampal area together with the area of the temporal horn for nat images is larger compared to hipp images, area of the hipp images is smaller compared to the AC-PC images and area of the nat images is smaller compared to the AC-PC images. The conclusion is that the measured area of the hippocampus in the native MRI is almost the same as the area of MRI reoriented only into the AC-PC axis. Therefore, when performing 2D area studies of the hippocampus or in the clinical practice we recommend usage of not-reoriented MRI images or to reorient them into the AC-PC axis. Surprising finding was that rotation of both AC-PC and hipp line towards x-axis among patients varies up to 35° and the same is true for the skull rotation so that it is not only a matter of the brain position.