European Radiology: Clinical application of MRI quantitative atrophy evaluation in patients with suspected Alzheimer's disease

Alzheimer's dementia (AD), the final clinical stage of Alzheimer's disease, is a progressive neurodegenerative disease
.
In patients presenting to neurology clinics, it is recommended that patients have at least one MRI to improve the sensitivity and specificity
of the diagnosis.
MRI of brain structures provides a non-invasive, reliable method to quantitatively assess the degree of atrophy in the brain by measuring the volume of the whole brain and regions, and is valuable
in identifying patients at risk of cognitive decline.
The latest clinical and research guidelines for the definition of Alzheimer's disease recommend including MRI in the evaluation of potentially at-risk individuals and quantifying
neurodegeneration.

The clinical evaluation of MRI scans relies primarily on detecting the manner and extent of temporoparietal lobe and overall cortical atrophy in the brain, usually supported
by visual scales.
Until now, methods to quantify (regional) atrophy have been mostly limited to research areas, and the widespread use of quantitative assessment in clinical settings has been hampered
by inadequate training of neuroradiology specialists, less demanding clinicians, and time issues.

Recently, a study published in the journal European Radiology evaluated the feasibility of quantifying brain parenchymal atrophy in AD patients in a real clinical environment, and clarified the difference between quantitative and visual measurement diagnostic performance, which provides a reference for
routine quantification of brain parenchyma in AD patients.

The review included 231 patients presenting to the Local Memory Clinic (LMC) in the Netherlands and 501 patients
from the Amsterdam Dementia Academic Cohort (ADC).
GMV and HCV were quantified
using FSL and FreeSurfer.
Use a visual score scale to assess medial temporal lobe atrophy and global atrophy
.
The ROC curve was used to assess which measure had the highest
distinction between cognitive normality (CN), mild cognitive impairment (MCI), and Alzheimer's dementia (AD).

Patients enrolled in LMC (age 70.
9± 8.
9 years; 47% were female; 19% CN; 34% MCI; 47% AD) and ADC patients (age 64.
9±8.
2 years; 42% were female; 35% CN, 43% MCI, 22% AD) compared to older age, more cerebrovascular lesions, lower GMV and HCV
.
While more than 95% of the scans in both cohorts can be visually scored, depending on the software, 94-98% of ADC scans can be quantified, while LMC scans are only 68-85%.

In both cohorts, visual scores and volumetric measurements performed similarly
in distinguishing CN from AD.

Figures for each diagnostic group (CN, MCI, AD) and each cohort (ADC, LMC) enable standardized HCV and GMV calculated using the FSL and FreeSurfer pipelines.

The measurements are displayed as residuals and adjusted for age and sex

The results of this study suggest that brain MRI scans from non-academic memory clinics have a high failure rate
for quantification of GMV and HCV without protocol optimization.
The volumetric quantitative output of automated software does not outperform visual scores by experienced radiologists, suggesting that the use of such software is not yet available for imaging of
patients with suspected Alzheimer's disease, given the time constraints and limited resources of real-life clinical settings.

Original source:

Silvia Ingala,Ingrid S van Maurik,Daniele Altomare,et al.
Clinical applicability of quantitative atrophy measures on MRI in patients suspected of Alzheimer's disease.
DOI:10.
1007/s00330-021-08503-7