Alzheimer´s disease causes symptoms such as memory loss, disorientation or mood swings in those suffering from it, but, what exactly changes in the brain and how can we improve its detection?

Magnetic resonance imaging (MRI) opens the possibility of examining the structure, the metabolism or the functionality of the brain and thereby, numerous analysis tools have been developed to understand and extract the information they carry. One of the most common techniques is “Voxel-based morphometry” or VBM (Ashburner and Friston, 2000), used to identify brain regional atrophy by statistically comparing a group of patients with a group of controls. Figure 1 describes the steps involved in a VBM analysis.

Figure 1: Steps of a VBM analysis.

VBM studies to date have exploited T1-weighted MRIs, which were designed to enhance the MRI signal contrast between grey and white matter tissues (Figure 2). They are therefore suitable for detecting the grey matter reduction caused by Alzheimer’s in areas such as the hippocampus, the region involved in memory formation. However, the T1-weighted images are not able to detect the neurological changes occurring in an early state of the disease—such as deposits of a protein called amyloid—which are present almost 20 years before the first symptoms appear. On the other hand, a different type of MRI called T2-weighted (Figure 2)—not available in 3D until recently—is more sensitive to possible fluctuations of the magnetic field, which are usually caused by the presence of certain metals in the brain.

Figura 2: Imagen mRI: a) T1-weighted b) T2-weighted

Amyloid plaques are surrounded by iron and this could alter the signal intensity of the T2 scans. Consequently, the hypothesis examined in (Diaz-de-Grenu et al., 2011) is that a VBM analysis using T2-weighted images (T2-VBM) might be able to locate not only morphological changes—already detected by T1-VBM—but also amyloid deposition. Results show that the T2-VBM analysis outperforms T1-VBM (Figure 3) and that the pattern of neurological lesions revealed is decidedly concordant with the high amyloid areas detected by PET (Positron emission tomography) scans (Frisoni and Delacourte, 2009) (Figure 4), which goes along the line of the proposed hypothesis.

Figure 3: Comparison of the results from two VBM analyses: (up) using T1-weighted images and (down) using T2-weighted images. The yellow areas mark the regions where there was a statistical difference between the group of Alzheimer’s patients and that of the healthy controls.

 So, if amyloid can already be detected by PET, why do we want another way of doing so? MRI scans—compared to PET—are considerably cheaper, more accessible and do not involve ionising radiation. Although this methodology is not expected to challenge the sensitivity of PET amyloid imaging, at a group level T2-VBM could provide an important surrogate marker of amyloid deposition and could have considerable potential for both longitudinal observational studies and drug development.

Figure 4: (Left) PET scan showing in red the areas more severely affected by amyloid. (Right) areas detected by T2-VBM.


Ashburner, J., Friston, K.J., 2000. Voxel-based morphometry–the methods. Neuroimage. 11, 805-21.

Diaz-de-Grenu, L.Z., et al., 2011. MRI detection of tissue pathology beyond atrophy in Alzheimer’s disease: introducing T2-VBM. Neuroimage. 56, 1946-53.

Frisoni, G.B., et al., 2009. In vivo mapping of amyloid toxicity in Alzheimer disease. Neurology. 72,1504-11.