Premium
P4‐210: Neural‐Net Model for Predicting Clinical Symptom Scores in Alzheimer's Disease
Author(s) -
Bhagwat Nikhil,
Pipitone Jon,
Park Min Tae M.,
Voineskos Aristotle N.,
Chakravarty Mallar
Publication year - 2016
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2016.06.2302
Subject(s) - neuroimaging , artificial intelligence , support vector machine , random forest , machine learning , artificial neural network , regression , computer science , medicine , pattern recognition (psychology) , psychology , neuroscience , psychoanalysis
transmission, is disrupted early in MCI relative to healthy counterparts (HC). However, there is little information about if, or how striatial dysfunction is related to AD pathobiology or cognitive capacity. Methods: Two-year resting-state fMRI data were available in 15 HC and 20 MCI older adults from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database. We analyzed the amplitude of low-frequency fluctuations (ALFF) and fractional ALFF (fALFF) in three frequency bands (typical bands: 0.01-0.08 Hz, slow-4: 0.027–0.073 Hz and slow-5: 0.01–0.027 Hz). We obtained measures of cognitive performance and cerebrospinal fluid levels of tau, ptau, and b-amyloid from the baseline visit. Results: ALFF and fALFF of bilateral putamen, but not caudate, declined significantly in MCI group compared to the HC group, especially for slow-4 band. Controlling for age and diagnosis, higher baseline tau and ptau significantly predicted greater 2-year decline of ALFF (especially for slow-4) in the right putamen, while poorer baseline cognitive measures significantly predicted greater decline of fALFF in the left putamen. Conclusions:Putamen function declines early in the AD neurodegenerative process. Putaminal ALFF or fALFF may be a sensitive early index of AD pathobiology regardless of clinical diagnosis.