FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2897601253> ?p ?o ?g. }

• W2897601253 abstract "Linear regression analyses are commonly utilized to assess relationships between neuroimaging and clinical variables. However, this traditional approach limits the number of questions one can ask because of the problem of multiple comparisons. An alternative approach involves fitting a single “comprehensive” hierarchical regression model that can answer multiple questions. We aimed to compare the power and precision of linear regression and hierarchical approaches evaluating the relationship between age and regional [18F]AV-1451 uptake in a cohort of atypical AD subjects. We compared these approaches using simulations based on tau PET findings in atypical AD in the context of realistic sample sizes. Thirty-seven PiB-positive atypical AD subjects underwent [18F]AV-1451 PET imaging and standard uptake value ratios (SUVRs) were calculated for ten regions-of-interest (ROIs) covering cortical and medial temporal regions. To compare statistical methods, many samples were simulated based on this cohort and modeled with a standard linear regression analysis. With log-SUVR as the response and age as the predictor, we obtained ten estimates and SEs of the region-specific age associations in each sample. P-values from the ten fits were examined in three ways: uncorrected, Bonferroni corrected, and false discovery rate (FDR) corrected. Also quantified was percent change in SUVR per decade difference. We compared these results to those from a hierarchical analysis in each sample by fitting a single linear mixed model with an overall age association, region-specific random slopes and intercepts, and subject-specific random intercepts. We compared statistical power, age association estimates, and precision in samples of 5 to 30 cases. Strong associations were observed between age and regional [18F]AV-141 uptake in both models, with youngest age associated with greater uptake in cortical regions. Power was greatest in hierarchical models (Figure 1) followed by unadjusted, FDR, and Bonferroni correction. The hierarchical approach provided estimates that were less susceptible to sampling variability at a given sample size (Figure 2). The probability of rejecting the null hypothesis in realized samples of size five to thirty from a population with known true effects. Hierarchical models outperform traditional linear regression with and without correction for multiple comparisons in almost every case. Estimated percent change in SUVR per 10-year increase in age in realized samples of size five to thirty from a population with known effects. Region specific estimates from hierarchical methods are more precise and are shrunk toward the overall age effect. Fitting a single hierarchical regression model to examine regional age associations provided appreciable gain in power and precision over fitting separate linear regression models in each region. Estimates based on hierarchical methods are likely more reproducible." @default.
• W2897601253 created "2018-10-26" @default.
• W2897601253 creator A5007824103 @default.
• W2897601253 creator A5022452703 @default.
• W2897601253 creator A5028578216 @default.
• W2897601253 creator A5040204986 @default.
• W2897601253 creator A5059557996 @default.
• W2897601253 date "2018-07-01" @default.
• W2897601253 modified "2023-10-16" @default.
• W2897601253 title "P3‐095: ADVANTAGES OF HIERARCHICAL MODELS FOR REGIONAL PET ANALYSES" @default.
• W2897601253 doi "https://doi.org/10.1016/j.jalz.2018.06.1451" @default.
• W2897601253 hasPublicationYear "2018" @default.
• W2897601253 type Work @default.
• W2897601253 sameAs 2897601253 @default.
• W2897601253 citedByCount "0" @default.
• W2897601253 crossrefType "journal-article" @default.
• W2897601253 hasAuthorship W2897601253A5007824103 @default.
• W2897601253 hasAuthorship W2897601253A5022452703 @default.
• W2897601253 hasAuthorship W2897601253A5028578216 @default.
• W2897601253 hasAuthorship W2897601253A5040204986 @default.
• W2897601253 hasAuthorship W2897601253A5059557996 @default.
• W2897601253 hasConcept C104317684 @default.
• W2897601253 hasConcept C105795698 @default.
• W2897601253 hasConcept C127808970 @default.
• W2897601253 hasConcept C129848803 @default.
• W2897601253 hasConcept C149769383 @default.
• W2897601253 hasConcept C152877465 @default.
• W2897601253 hasConcept C166957645 @default.
• W2897601253 hasConcept C183905921 @default.
• W2897601253 hasConcept C185592680 @default.
• W2897601253 hasConcept C193244246 @default.
• W2897601253 hasConcept C205649164 @default.
• W2897601253 hasConcept C2779343474 @default.
• W2897601253 hasConcept C33923547 @default.
• W2897601253 hasConcept C48921125 @default.
• W2897601253 hasConcept C53059260 @default.
• W2897601253 hasConcept C55493867 @default.
• W2897601253 hasConcept C83546350 @default.
• W2897601253 hasConcept C96608239 @default.
• W2897601253 hasConceptScore W2897601253C104317684 @default.
• W2897601253 hasConceptScore W2897601253C105795698 @default.
• W2897601253 hasConceptScore W2897601253C127808970 @default.
• W2897601253 hasConceptScore W2897601253C129848803 @default.
• W2897601253 hasConceptScore W2897601253C149769383 @default.
• W2897601253 hasConceptScore W2897601253C152877465 @default.
• W2897601253 hasConceptScore W2897601253C166957645 @default.
• W2897601253 hasConceptScore W2897601253C183905921 @default.
• W2897601253 hasConceptScore W2897601253C185592680 @default.
• W2897601253 hasConceptScore W2897601253C193244246 @default.
• W2897601253 hasConceptScore W2897601253C205649164 @default.
• W2897601253 hasConceptScore W2897601253C2779343474 @default.
• W2897601253 hasConceptScore W2897601253C33923547 @default.
• W2897601253 hasConceptScore W2897601253C48921125 @default.
• W2897601253 hasConceptScore W2897601253C53059260 @default.
• W2897601253 hasConceptScore W2897601253C55493867 @default.
• W2897601253 hasConceptScore W2897601253C83546350 @default.
• W2897601253 hasConceptScore W2897601253C96608239 @default.
• W2897601253 hasIssue "7S_Part_20" @default.
• W2897601253 hasLocation W28976012531 @default.
• W2897601253 hasOpenAccess W2897601253 @default.
• W2897601253 hasPrimaryLocation W28976012531 @default.
• W2897601253 hasRelatedWork W1434188134 @default.
• W2897601253 hasRelatedWork W2018697919 @default.
• W2897601253 hasRelatedWork W2042906314 @default.
• W2897601253 hasRelatedWork W2079415229 @default.
• W2897601253 hasRelatedWork W2102983641 @default.
• W2897601253 hasRelatedWork W2129571249 @default.
• W2897601253 hasRelatedWork W2166178455 @default.
• W2897601253 hasRelatedWork W2167660240 @default.
• W2897601253 hasRelatedWork W4214815066 @default.
• W2897601253 hasRelatedWork W4214917765 @default.
• W2897601253 hasVolume "14" @default.
• W2897601253 isParatext "false" @default.
• W2897601253 isRetracted "false" @default.
• W2897601253 magId "2897601253" @default.
• W2897601253 workType "article" @default.