Premium
IC‐P‐029: PATIENT‐TAILORED PREDICTIONS OF GRAY MATTER ATROPHY IN ALZHEIMER'S DISEASE BASED ON BASELINE FLORTAUCIPIR AND CONNECTOME‐BASED SPREAD
Author(s) -
Brown Jesse,
La Joie Renaud,
Visani Adrienne,
Rosen Howard J.,
Miller Bruce L.,
Jagust William J.,
Seeley William W.,
Rabinovici Gil D.
Publication year - 2018
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.2018.06.2093
Subject(s) - atrophy , medicine , nuclear medicine , dementia , alzheimer's disease , magnetic resonance imaging , neuroscience , psychology , disease , radiology
biomarker measures. A. An interaction term was added to the models described in Figure 2. The model predicting amyloid preformed significantly better with the interaction term (R0.249, p < 2310, ANOVA comparing models: F 970.58, p< 2310). Scatterplots were created using visreg in R in order to visualize the interaction between connectivity measures. In order to assess the effect of changes in pDMN variation values on the relationship between degree centrality and amyloid (left), the data were split into tertiles based on pDMN variation, and within each third pDMN variation was controlled for by setting it to the median value for that third. As is evident from the scatterplots, therewas a positive correlation between degree centrality and amyloid IC score within all tertiles, increasing from the bottom third of pDMN variation (red) to the top third (green). A similar analysis was done for pDMN variation (right), this time controlling for degree centrality, showing a similar positive correlation between pDMNvariation and amyloid IC, again increasingwith increasing degree centrality. B. The model predicting tau also preformed significantly better with the interaction term (R 0.363, p< 2310,ANOVA comparing models: F 2263.2, p< 2310). However, the interaction between connectivity measures were not congruent. There was a positive correlation between degree centrality and tau IC scores in the lower third of pDMN variation values, little correlation in the middle third and a negative correlation in the upper third (left). pDMN variation was positively correlated with tau scores in all three degree centrality tertiles, although the correlation was highest in the bottom third (right). Poster Presentations: Saturday, July 21, 2018 P34