Publications (12) View all
-
Article: The Receiver Operational Characteristic for Binary Classification with Multiple Indices and Its Application to the Neuroimaging Study of Alzheimer's Disease.
[show abstract] [hide abstract]
ABSTRACT: Given a single index, the Receiver Operational Characteristic (ROC) curve analysis is routinely utilized for characterizing performances in distinguishing two conditions/groups in terms of sensitivity and specificity. Given the availability of multiple data sources (referred to as multi-indices), such as multi-modal neuroimaging datasets, cognitive tests, and clinical ratings and genomic data in Alzheimer's disease (AD) studies, the single-index based ROC under-utilizes all available information. For a long time, a number of algorithmic/analytic approaches combining multiple indices have been widely used to simultaneously incorporate multiple sources. In this study, we propose an alternative for combining multiple indices using logical operations, such as 'AND', 'OR', and 'at least n'; (where n is an integer), to construct multivariate ROC (multiV-ROC) and characterize the sensitivity and specificity statistically associated with the use of multiple indices. With and without the 'leave-one-out' cross-validation, we used two datasets from Alzheimer' disease studies to showcase the potentially increased sensitivity/specificity of the multiV-ROC in comparison to the single-index ROC and linear discriminant analysis (an analytic way of combining multi-indices). We conclude that, for the datasets we investigated, the proposed multiV-ROC approach is capable of providing a natural and practical alternative with improved classification accuracy as compared to univariate ROC and linear discriminant analysis.IEEE/ACM transactions on computational biology and bioinformatics / IEEE, ACM 11/2012; · 2.25 Impact Factor -
Article: Florbetapir PET analysis of amyloid-β deposition in the presenilin 1 E280A autosomal dominant Alzheimer's disease kindred: a cross-sectional study.
[show abstract] [hide abstract]
ABSTRACT: BACKGROUND: Fibrillar amyloid-β (Aβ) is thought to begin accumulating in the brain many years before the onset of clinical impairment in patients with Alzheimer's disease. By assessing the accumulation of Aβ in people at risk of genetic forms of Alzheimer's disease, we can identify how early preclinical changes start in individuals certain to develop dementia later in life. We sought to characterise the age-related accumulation of Aβ deposition in presenilin 1 (PSEN1) E280A mutation carriers across the spectrum of preclinical disease. METHODS: Between Aug 1 and Dec 6, 2011, members of the familial Alzheimer's disease Colombian kindred aged 18-60 years were recruited from the Alzheimer's Prevention Initiative's registry at the University of Antioquia, Medellín, Colombia. Cross-sectional assessment using florbetapir PET was done in symptomatic mutation carriers with mild cognitive impairment or mild dementia, asymptomatic carriers, and asymptomatic non-carriers. These assessments were done at the Banner Alzheimer's Institute in Phoenix, AZ, USA. A cortical grey matter mask consisting of six predefined regions.was used to measure mean cortical florbetapir PET binding. Cortical-to-pontine standard-uptake value ratios were used to characterise the cross-sectional accumulation of fibrillar Aβ deposition in carriers and non-carriers with regression analysis and to estimate the trajectories of fibrillar Aβ deposition. FINDINGS: We enrolled a cohort of 11 symptomatic individuals, 19 presymptomatic mutation carriers, and 20 asymptomatic non-carriers, ranging in age from 20 to 56 years. There was greater florbetapir binding in asymptomatic PSEN1 E280A mutation carriers than in age matched non-carriers. Fibrillar Aβ began to accumulate in PSEN 1E280A mutation carriers at a mean age of 28·2 years (95% CI 27·3-33·4), about 16 years and 21 years before the predicted median ages at mild cognitive impairment and dementia onset, respectively. (18)F florbetapir binding rose steeply over the next 9·4 years and plateaued at a mean age of 37·6 years (95% CI 35·3-40·2), about 6 and 11 years before the expected respective median ages at mild cognitive impairment and dementia onset. Prominent florbetapir binding was seen in the anterior and posterior cingulate, precuneus, and parietotemporal and frontal grey matter, as well as in the basal ganglia. Binding in the basal ganglia was not seen earlier or more prominently than in other regions. INTERPRETATION: These findings contribute to the understanding of preclinical familial Alzheimer's disease and help set the stage for assessment of amyloid-modifying treatments in the prevention of familial Alzheimer's disease. FUNDING: Avid Radiopharmaceuticals, Banner Alzheimer's Foundation, Nomis Foundation, Anonymous Foundation, Forget Me Not Initiative, Colciencias, National Institute on Aging, and the State of Arizona.The Lancet Neurology 11/2012; · 23.46 Impact Factor -
Article: Apolipoprotein E ε4 and age effects on florbetapir positron emission tomography in healthy aging and Alzheimer disease.
[show abstract] [hide abstract]
ABSTRACT: OBJECTIVES: Investigate apolipoprotein E ε4 (APOE4) gene and aging effects on florbetapir F18 positron emission tomography (PET) in normal aging and Alzheimer's disease (AD). Methods: Florbetapir F18 PET images were analyzed from 245 participants, 18-92 years of age, from Avid Radiopharmaceutical's multicenter registered trials, including 86 younger healthy control volunteers (yHC), 61 older healthy control volunteers (oHC), 53 mild cognitive impairment (MCI) patients, and 45 AD dementia patients (DAT). Mean florbetapir standard uptake value ratios (SUVRs) were used to evaluate the effects of APOE4 carrier status, older age, and their interaction in each of these groups. Results: In comparison with non-carriers, the APOE4 carriers in each of the oHC, MCI, and DAT groups had higher mean cortical-to-cerebellar florbetapir SUVRs, patterns of florbetapir PET elevations characteristic of DAT, and a higher proportion meeting florbetapir PET positivity criteria. Only the oHC group had a significant association between mean cortical florbetapir SUVRs and age. In cognitively normal adults, without regards to APOE4 genotype, amyloid began to increase at age 58 (95% confidence interval [CI]: 52.3-63.7), with a predicted typical age of florbetapir positivity occurring around age 71 years. Presence of the APOE4 gene reduced the age of predicted florbetapir positivity in normal aging to around age 56 years, approximately 20 years younger than non-carriers. Interpretation: Cerebral amyloid deposition is associated with APOE4 carrier status in older healthy control subjects and symptomatic AD patients, and increases with age in older cognitively normal individuals. Amyloid imaging positivity appears to begin near age 56 years in cognitively intact APOE4 carriers and age 76 years in APOE4 non-carriers.Neurobiology of aging 05/2012; · 5.94 Impact Factor -
Article: Summary metrics to assess Alzheimer disease-related hypometabolic pattern with 18F-FDG PET: head-to-head comparison.
[show abstract] [hide abstract]
ABSTRACT: In the recently revised diagnostic criteria for Alzheimer disease (AD), the National Institute on Aging and Alzheimer Association suggested that confidence in diagnosing dementia due to AD and mild cognitive impairment (MCI) due to AD could be improved by the use of certain biomarkers, such as (18)F-FDG PET evidence of hypometabolism in AD-affected brain regions. Three groups have developed automated data analysis techniques to characterize the AD-related pattern of hypometabolism in a single measurement. In this study, we sought to directly compare the ability of these three (18)F-FDG PET data analysis techniques--the PMOD Alzheimer discrimination analysis tool, the hypometabolic convergence index, and a set of meta-analytically derived regions of interest reflecting AD hypometabolism pattern (metaROI)--to distinguish moderate or mild AD dementia patients and MCI patients who subsequently converted to AD dementia from cognitively normal older adults. One hundred sixty-six (18)F-FDG PET patients from the AD Neuroimaging Initiative, 308 from the Network for Efficiency and Standardization of Dementia Diagnosis, and 176 from the European Alzheimer Disease Consortium PET study were categorized, with masking of group classification, as AD, MCI, or healthy control. For each AD-related (18)F-FDG PET index, receiver-operating-characteristic curves were used to characterize and compare subject group classifications. The 3 techniques were roughly comparable in their ability to distinguish each of the clinical groups from cognitively normal older adults with high sensitivity and specificity. Accuracy of classification (in terms of area under the curve) in each clinical group varied more as a function of dataset than by technique. All techniques were differentially sensitive to disease severity, with the classification accuracy for MCI due to AD to moderate AD varying from 0.800 to 0.949 (PMOD Alzheimer tool), from 0.774 to 0.967 (metaROI), and from 0.801 to 0.983 (hypometabolic convergence index). The 3 tested techniques have the potential to help detect AD in research and clinical settings. Additional efforts are needed to clarify their ability to address particular scientific and clinical questions. Their incremental diagnostic value over other imaging and biologic markers makes them easier to implement by other groups for these purposes.Journal of Nuclear Medicine 02/2012; 53(4):592-600. · 6.38 Impact Factor -
Article: Correlations between FDG PET glucose uptake-MRI gray matter volume scores and apolipoprotein E ε4 gene dose in cognitively normal adults: a cross-validation study using voxel-based multi-modal partial least squares.
[show abstract] [hide abstract]
ABSTRACT: We previously introduced a voxel-based, multi-modal application of the partial least square algorithm (MMPLS) to characterize the linkage between patterns in a person's complementary complex datasets without the need to correct for multiple regional comparisons. Here we used it to demonstrate a strong correlation between MMPLS scores to characterize the linkage between the covarying patterns of fluorodeoxyglucose positron emission tomography (FDG PET) measurements of regional glucose metabolism and magnetic resonance imaging (MRI) measurements of regional gray matter associated with apolipoprotein E (APOE) ε4 gene dose (i.e., three levels of genetic risk for late-onset Alzheimer's disease (AD)) in cognitively normal, late-middle-aged persons. Coregistered and spatially normalized FDG PET and MRI images from 70% of the subjects (27 ε4 homozygotes, 36 ε4 heterozygotes and 67 ε4 non-carriers) were used in a hypothesis-generating MMPLS analysis to characterize the covarying pattern of regional gray matter volume and cerebral glucose metabolism most strongly correlated with APOE-ε4 gene dose. Coregistered and spatially normalized FDG PET and MRI images from the remaining 30% of the subjects were used in a hypothesis-testing MMPLS analysis to generate FDG PET-MRI gray matter MMPLS scores blind to their APOE genotype and characterize their relationship to APOE-ε4 gene dose. The hypothesis-generating analysis revealed covarying regional gray matter volume and cerebral glucose metabolism patterns that resembled those in traditional univariate analyses of AD and APOE-ε4 gene dose and PET-MRI scores that were strongly correlated with APOE-ε4 gene dose (p<1 × 10(-16)). The hypothesis-testing analysis results showed strong correlations between FDG PET-MRI gray matter scores and APOE-ε4 gene dose (p = 8.7 × 10(-4)). Our findings support the possibility of using the MMPLS to analyze complementary datasets from the same person in the presymptomatic detection and tracking of AD.NeuroImage 02/2012; 60(4):2316-22. · 5.89 Impact Factor
