Premium
P1‐119: ENHANCING DEEP LEARNING MODEL PERFORMANCE FOR AD DIAGNOSIS USING ROI‐BASED SELECTION
Author(s) -
Qiu Shangran,
Heydari Megan S.,
Miller Matthew I.,
Joshi Prajakta S.,
Wong Benjamin C.,
Au Rhoda,
Kolachalama Vijaya B.
Publication year - 2019
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.2019.06.674
Subject(s) - artificial intelligence , convolutional neural network , computer science , region of interest , deep learning , neuroimaging , pattern recognition (psychology) , binary classification , sagittal plane , voxel , support vector machine , neuroscience , psychology , medicine , radiology
hypertension (SBP: 165616 vs. 14265 mmHg, 6-mo vs. 3-mo; mean6SD). Linear regression revealed a positive correlation between SBP and PWV (Figure 1). PWV was negatively correlated with cerebral perfusion in the hippocampus (Figure 2). Hippocampal CBF positively correlated with hippocampal NAA (Figure 3). Ratio of central rearing duration to total rearing duration positive correlated with hippocampal blood flow (Figure 4). Conclusions: In the Dahl-S rat model of hypertension, greater PWV, a marker for CAS, was associated with decreased hippocampal perfusion. Despite approaching statistical significance, the trend exhibits the theoretical relationship between PWV and hippocampal blood flow. Likewise, higher hippocampal blood flow is associated with a higher hippocampal NAA concentrations indicating that better blood supply would be resulted in a higher number of neurons. Lastly the positive correlation between hippocampal blood flow and increased central OF activity suggests that higher cerebral blood flow is associated with lower anxious behavior in Dahl-S rats. The work was supported by the Intramural Research Program of the NIA, NIH.