INCREASED DEFAULT MODE NETWORK FUNCTIONAL CONNECTIVITY IN MILD COGNITIVE IMPAIRMENT: A DETRIMENTAL BRAIN MECHANISM ASSOCIATED WITH POOR SEMANTIC MEMORY PERFORMANCE

Poster Presentations: P4 Tammie L.S. Benzinger2, Randall Bateman3, John C. Morris3, Martin Rhys Farlow1, Bernardino Francesco Ghetti1, Andrew J. Saykin4, 1 Indiana University School of Medicine, Indianapolis, Indiana, United States; 2Washington University School of Medicine, St Louis, Missouri, United States; 3Washington University School of Medicine, St. Louis, Missouri, United States; 4Indiana University School of Medicine, Indianapolis, Indiana, United States. Contact e-mail: [email protected] Background: Carriers of PSEN1 and APP mutations associated with familial autosomal dominant Alzheimer’s disease (ADAD) have shown brain atrophy, amyloid deposition, and alterations in other biomarkers. However, the effect of these mutations on brain activation during memory processing has not been thoroughly explored. A visual scene encoding fMRI task was used to investigate activation changes in clinically affected and preclinical carriers of a PSEN 1 or APP mutation. Methods: 17 participants, including 5 clinically affected carriers of a PSEN1 or APP mutation (symptomatic mutation carrier (sMC); 4 PSEN1, 1 APP), 5 asymptomatic carriers (aMC; 4 PSEN1, 1 APP), and 7 healthy age-matched non-carriers (NC), were imaged using blood oxygen level dependent (BOLD) fMRI during a visual scene encoding task. After quality control, all fMRI scans were preprocessed using standard procedures in SPM8. Briefly, scans were spatially aligned, normalized to standard (MNI) space, resampled to 2mm isotropic voxels, and smoothed using a 6mm FWHM kernel. Contrast images were generated for each participant for the encoding > control conditions. These contrast images were then compared between groups using a one-way ANOVA in SPM8 and masked using a whole brain mask. Results were displayed at a significance threshold of p<0.01 (unc.) and a minimum cluster size of 25 voxels. Results: sMC showed significantly less activation than NC during visual scene encoding in the bilateral occipital, inferior temporal, and frontal lobes, as well as in the right medial temporal lobe (MTL; Figure 1A). sMC also showed significantly less activation than aMC in bilateral occipital, inferior temporal, frontal and MTL regions (Figure 1B). However, aMC showed greater activation than NC in the bilateral MTL (Figure 1C). Exclusion of the APP mutation carriers did not significantly alter the findings. Conclusions: Symptomatic carriers of a PSEN1 or APP mutation show reduced brain activation during visual encoding which suggests that the presence of clinical symptoms is associated with a decrease in brain activity during memory processing. Interestingly, asymptomatic carriers of these mutations showed increased activation in the bilateral MTL, which may represent a compensatory increase in activation in preclinical stage disease.

P823

Background: One of the primary goals of ADNI was to optimise the methods of MRI acquisition in order to ensure the standards of acquiring data were uniform over-time and across sites. ADNI sequences were also made available to non-ADNI studies that used either a 1.5T or 3T MR scanner. Non-ADNI sequences were designed to eliminate variability in data that arises from the use of different field strengths. The aim of the current study was to compare the ADNI-1.5T and ADNI-3T MR T1-weighted volumes using voxel-based morphometry (VBM). Methods: T1-weighted images of 30 subjects were acquired using a 1.5T and 3T GE Signa HDx scanner using the ADNI-1 volumes sequence. Images were pre-processed and analysed using SPM-8. Voxel-wise comparisons of grey matter (GM) and white matter (WM) were performed to compare whole-brain differences between sequences. Global GM, WM and cerebrospinal fluid (CSF) were assessed using SPSS using standard paired t-tests and the Intra-class correlation co-efficient (ICC) as a reliability measure to examine how strongly volumes between each coil and sequence were related. Results: Voxel-wise comparisons show widespread GM increases (Frontal and middle temporal gyrus), and localised decreases (Right Brainstem, Cingulate Gyrus, Cerebellar Tonsil) in the ADNI-3T relative to ADNI-1.5T images. Localised WM decreases (Medial Frontal Gyrus, Right/Left Cerebellum, Brainstem) and widespread increases (Precentral Gyrus, Fusiform Gyrus, Cerebellum) were observed in the ADI-1.5T compared to ADNI-3T images. Correlation coefficients and percentage differences for each tissue type between ADNI 1.5T and ADNI 3T were as follows: ((GM: ICC¼0.95, ADNI-3T 8.3% > ADNI-1.5T) (WM: ICC¼0.98, ADNI-1.5T 3.53% > ADNI-3T) (WM: ICC¼0.80, ADNI-1.5T 11.93% > ADNI-3T)). Conclusions: The global decrease of GM and the increase of WM in the ADNI-1.5T compared to the ADNI-3T MR volume sequence suggests that the two image acquisition protocols are not directly comparable using SPM-8. Reproducibility measures (ICC) and total volumes of GM, WM and CSF also differed between the two protocols in the following order of magnitude: CSF>GM>WM. This has implications for studies aiming to analyse images acquired using the ADNI-1.5T and ADNI-3T MRI scanning protocols under VBM. P4-107

INCREASED DEFAULT MODE NETWORK FUNCTIONAL CONNECTIVITY IN MILD COGNITIVE IMPAIRMENT: A DETRIMENTAL BRAIN MECHANISM ASSOCIATED WITH POOR SEMANTIC MEMORY PERFORMANCE

Simona Gardini1, Fabio Sambataro2, Fernando Cuetos3, Fabrizio Fasano4, Massimo Marchi5, Girolamo Crisi6, Annalena Venneri7, Paolo Caffarra5, 1 Department of Neuroscience, University of Parma, Parma, Italy; 2Center for Neuroscience and Cognitive Systems, Istituto Italiano di Tecnologia, Rovereto, Italy; 3Department of Psychology, University of Oviedo, Oviedo, Alberta, Spain; 4Department of Neuroscience, University of Parma, Parma, Italy; 5Department of Neuroscience, University of Parma, Parma, Italy; 6 Struttura Complessa Neuroradiologia, Department of Neuroscience, University of Parma, Parma, Italy; 7Department of Neuroscience, University of Sheffield, and Sheffield Teaching Hospital NHS Foundation Trust, Sheffield, United Kingdom and IRCCS, Fondazione Ospedale San Camillo, Venice, Italy. Contact e-mail: [email protected]

Figure 1. Altered brain activation during visual scene encoding in carriers of a PSEN1 or APP mutation P4-106

A VOXEL-WISE MORPHOMETRY COMPARISON OF THE ADNI 1.5T AND ADNI 3.0T VOLUMETRIC MRI PROTOCOLS

Simon Brunton1, Cerisse Gunasinghe1, Nigel Jones2, Matthew Kempton2, Eric Westman3, Andy Simmons4, 1King’s College London, London, England, United Kingdom; 2Kings College London, London, England, United Kingdom; 3Karolinska Institute, Stockholm, Sweden; 4King’s College London, Institute of Psychiatry, London, England, United Kingdom. Contact e-mail: [email protected]

Background: Semantic memory deficits and alterations of the Default Mode Network (DMN) connectivity have been described in Mild Cognitive Impairment (MCI). Nonetheless, the role of DMN changes in semantic memory impairments in this clinical condition is still unknown. The present study aimed at investigating the relationship between semantic memory performance and brain intrinsic connectivity within the DMN in MCI patients. Methods: Twenty-one MCI patients and twenty-one healthy elderly controls matched for demographics, took part in this study. All participants underwent a multi-tasking semantic battery involving tasks of category fluency, visual naming and naming from definition for objects, actions and famous people, a word-association task for early and late acquired words and a reading task. A subgroup of the original sample (s ixteen MCI patients and twenty healthy elderly controls) had an 8-min functional magnetic resonance imaging (fMRI) scan at rest with eyes closed. DMN functional connectivity was estimated using a seed-based approach with

P824

Poster Presentations: P4

seeds placed in the medial prefrontal cortex (mPFC) and posterior cingulate (PCC). Two sample t-tests were used to compare connectivity across diagnostic groups for each seed. The brain-behaviour relationship with indexes derived from the semantic battery was assessed using simple correlations. Results: Patients showed an extensive semantic memory impairment characterized by decreased performance in category fluency, visual naming, naming from definition, words-association and reading. Compared to healthy elderly, patients showed increased DMN connectivity between the mPFC and the PCC (BA 29/30), and between the PCC and the parahippocampus and anterior hippocampus, (BA 36). Patients showed also a significant negative correlation of mPFC connectivity with parahippocampus and posterior hippocampus (BA 27) and total scores on the visual naming task. Conclusions: Our findings suggest that DMN connectivity alterations may contribute to semantic memory deficits in MCI, specifically in visual naming. An increased DMN connectivity between the mPFC and the PCC, and between the PCC and the parahippocampus and anterior hippocampus, appears to cause relevant disadvantageous reorganization of brain functions in MCI. P4-108

APPLYING AUTOMATED MR-BASED DIAGNOSING IN A MEMORY CLINIC: A PROSPECTIVE STUDY WITH MILD COGNITIVE IMPAIRMENT

Ahmed Abdulkadir1, Stefan Kl€oppel2, Jessica Peter3, Anna Ludl4, Sabrina Maier4, Anne Pilatus4, Irina Mader4, Michael H€ull4, 1University of Freiburg, Computer Science, Freiburg, Germany; 2University Medical Center Freiburg, Freiburg, Germany; 3University of Freiburg, Freiburg, Germany; 4University Medical Center, Freiburg, Germany. Contact e-mail: [email protected] Background: Brain imaging is part of routine dementia workup and requires specialist’s knowledge for an accurate interpretation. Several studies have demonstrated that pattern recognition methods applied to MRI data can identify those with mild cognitive impairment (MCI) who will soon develop dementia. So far, these studies are limited by their application to a highly preselected sample that differs from that seen in a dementia clinic. Cases with co-morbidities or low scan quality are usually excluded. We performed a single centre prospective study to evaluate the diagnostic accuracy of automated methods in an routine clinical application. Methods: Over two years, cases with MCI for which structural MRI was requested as part of routine workup were included in the study. Clinicians had to indicate whether they expected conversion to dementia within 12 months together with their level of diagnostic confidence based on clinical interview and neuropsychology. These ratings were repeated after clinicians knew the outcome of MRI. Probability of conversion was also computed by a computer algorithm (support vector machine) trained with data from the ADNI study with a well established processing pipeline and without additional clinical data. Results: Over 50 subjects with MCI were included. Cases with existing CNS-disorders or low scan quality remained in the study. Confidence ratings by clinicians increased with the availability of MRI. Based on follow-up data from 15 subjects, the machine learning approach performed better than chance but was outperformed by prognostic estimations from clinicians with knowledge of MRI results. Conclusions: While automated MRI-based method have shown promise when applied to well standardized and quality-controlled datasets such as ADNI, their stability and diagnostic accuracy still requires improvement before applying it as a routine method in clinical practice. P4-109

EXISTING THRESHOLDS FOR PIB POSITIVITY ARE TOO HIGH

Sylvia Villeneuve1, Cindee Madison2, Nagehan Ayakta1, Renaud La Joie1, Brendan I. Cohn-Sheehy3, Jacob Vogel1, Shawn Marks1, Samia K. ArthurBentil1, Bruce Reed4, Charles DeCarli5, Gil Dan Rabinovici6, William Jagust1, 1University of California, Berkeley, California, United States; 2Helen Wills Neuroscience Institute, Berkeley, California, United States; 3Memory and Aging Center, University of California, San Francisco, California, United States; Jagust Lab, University of California, Berkeley, California, United States; 4Univeristy of California at Davis, Martinez, California, United States; 5University of California at Davis, Sacramento, California, United States; 6Memory and Aging Center, University of

California, San Francisco, California, United States. Contact e-mail: [email protected] Background: There is no consensus among researchers about the thresholds that define amyloid positivity and there is little known about the earliest phases of amyloid accumulation in older adults. The present study had two goals: first, to derive a cutoff that captures early accumulation using both DVR and SUVR data from our subjects and, secondly, to examine their pattern of amyloid accumulation. Methods: Amyloid accumulation was investigated in 152 cognitively normal older adults using: (1) a reference group of young adults, (2) Gaussian mixture modeling (GMM), (3) cluster analyses and (4) voxel-wise analyses. All analyses used DVR and SUVR data with a cerebellar gray reference ROI. For voxel-wise analyses, subjects were ranked based on their global DVR status. To track when and where amyloid starts accumulating we compared a group of 22 subjects with a mean DVR of 1 (control group) to the next 22 subjects (group of interest) and iteratively increased the mean DVR of the group of interest by dropping the subject with the lowest value and adding the subject with the next higher value. This procedure was repeated until the subject with the highest DVR was included in the group of interest. Results: The threshold 2 SD above the young subjects was a DVR of 1.07 (SUVR ¼ 1.19). Both the GMM and the cluster-derived thresholds were 1.09 (SUVR 1.22). The Figure shows that amyloid starts accumulating in the medial frontal cortex (mean DVR ¼ 1.07, SUVR ¼ 1.19), then spreads to the precuneus, the lateral frontal and parietal lobes, and finally the temporal lobe. Conclusions: Amyloid starts to accumulate long before individuals reach the widely used SUVR cutoffs of 1.4 and 1.5. These results support an SUVR cutoff of 1.21 (DVR ¼ 1.08) to capture early amyloid accumulation. This cutoff was confirmed by an autopsy study of 43 dementia cases (Rabinovici et al., submitted).

Figure. Pattern of amyloid accumulation in cognitively older adults. Each row of images reflects a voxel-wise contrast of 22 subjects with the mean value for global DVR/SUVR listed at left compared to a reference group (N¼22) with a global DVR¼1. Significant voxels first appeared when the group mean was DVR¼1.07. Threshold at p <.05 after family-wise error correction, k > 150. P4-110

RELATIONSHIP BETWEEN HIPPOCAMPAL VOLUME AND FREQUENCY OF EVENT RETRIEVAL USING THE HISTORIC EVENTS MEMORY TEST IN PEOPLE WITH ALZHEIMER’S DISEASE

Thomas Leyhe1, Stephan Mueller2, Christian Mychajliw2, Marko Wilke3, Andreas Fallgatter4, Michael Erb5, Cornelia Veil5, Klaus Scheffler6, Ralf Saur7, 1Psychiatric University Hospital, Basel, Switzerland;