Background:
Recently, efforts have been made to combine complementary perspectives in the assessment
of Alzheimer type dementia. Of particular interest is the definition of the fingerprints of an early
stage of the disease known as Mild Cognitive Impairment or prodromal Alzheimer's Disease. Machine learning
approaches have been shown to be extremely suitable for the implementation of such a combination.
Methods:
In the present pilot study we combined the machine learning approach with structural magnetic
resonance imaging and cognitive test assessments to classify a small cohort of 11 healthy participants and 11
patients experiencing Mild Cognitive Impairment. Cognitive assessment included a battery of standardised
tests and a battery of experimental visuospatial memory tests. Correct classification was achieved in 100% of
the participants, suggesting that the combination of neuroimaging with more complex cognitive tests is suitable
for early detection of Alzheimer Disease.
Results:
In particular, the results highlighted the importance of the experimental visuospatial memory test battery
in the efficiency of classification, suggesting that the high-level brain computational framework underpinning
the participant's performance in these ecological tests may represent a “natural filter” in the exploration
of cognitive patterns of information able to identify early signs of the disease.
Unified framework for early stage status prediction of autism based on infant structural magnetic resonance imaging
