Return to 8th International Conference Index

MRI in the Practical Evaluation of Dementia: ‘Beyond Exclusion’

View Philip Scheltens' presentation.

This paper reviews the use of imaging techniques to aid in the clinical diagnosis of dementia. Two approaches are distinguished.

The exclusionary approach is one in which imaging is used to rule out diseases that would mimic or cause dementia. Based on the literature this approach yields very little, if any, information that was not identified clinically.

The more positive approach uses imaging as a diagnostic tool to identify changes specific for causes of dementia. Any assessment of medial temporal lobe atrophy on Magnetic Resonance Imaging (MRI) will result in a reasonably high positive likelihood ratio distinguishing AD patients from non-demented individuals but fails to distinguish AD patients from patients with other dementias. For a diagnosis of vascular dementia, imaging is necessary, although not all vascular changes fulfill requirements of being relevant dementia. Potentially of more importance, given the higher prevalence of AD, is identifying concomitant vascular changes in AD that may be amenable to therapy and may be used to identify subgroups.

Serial imaging allows an assessment of progression. Subjects serve as their own reference and rates of change (e.g atrophy) can then be calculated. As long ago as the early 1980s, serial imaging studies (mainly CT) showed that when subjects are rescanned after an interval of more than one year, rates of ventricular enlargement were significantly greater in AD than in normal controls. On serial MR scans, hippocampal atrophy rates are significantly increased in AD, with a mean rate of 4% per year compared to age-matched elderly controls (1.6%). Manual measurements such as these are time-consuming and haven't entered routine clinical practice. Digital subtraction of serial registered (positionally matched) scans allows small amounts of diffuse atrophy to be detected and displayed automatically. Using such methods, rates of whole brain atrophy in AD (2.5 +/- 1.1 % per year) have been shown to be several times greater than in normal ageing (0.4 +/- 0.4 % per year). Furthermore rates of cerebral atrophy are already increased in asymptomatic individuals at risk of familial AD who subsequently become clinically affected. Analysis of the presymptomatic scans suggests atrophy rates increase first in medial temporal lobe and posterior cingulate regions but by the time a clinical diagnosis can usually be made atrophy is already generalized. Serial imaging is likely to be increasingly used in clinical trials in the search for treatments that slow diseases such as AD. Demonstration of a slowing in rates of cerebral atrophy would provide a strong suggestion that a therapy is actually slowing degeneration rather than simply providing short-term symptomatic relief. If and when such therapies are found there will be even greater interest in making an accurate diagnosis earlier in these diseases and MRI will be used to support these diagnoses.

View Philip Scheltens' presentation
To hear the associated audio, click on the "Play Audio" button on each slide.