Premium
IC‐P3‐193: Impact of apolipoprotein E gene polymorphism on cerebral white matter integrity in Alzheimer's disease: DTI study
Author(s) -
Lee Dong Young,
Choo I.L. Han,
Lee Jae Sung,
Oh Jungsu S.,
Lee Dong Soo,
Kim Ki Woong,
Jhoo Jin Hyeong,
Youn Jong Choul,
Kim Shin Gyeom,
Kim Ji Wook,
Woo Jong Inn
Publication year - 2008
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.2008.05.137
Subject(s) - white matter , apolipoprotein e , fractional anisotropy , diffusion mri , psychology , medicine , pathology , magnetic resonance imaging , disease , radiology
automated method, HV using semi-automated atlas-based warping method, and ERCV using manual-tracing [1-3]. Coefficients of variation (COVs) were calculated as a metric of measurement reproducibility. Ten AD subjects (73 9 years) were scanned using the same sequences (8-site/3platform) and processed using the same methods (four-operators). The most experienced operator’s output was considered as the gold-standard and intraclass correlation coefficients (ICC) were calculated for accuracy. Estimated sample-size was calculated to detect 20% treatment effects with consideration of normal-aging and measurement reproducibility (90%power/5%-level, rate (%) in placebo group as 2.4 1.1 (WBA), 4.5 2.9 (HV) and 7.1 3.2 (ERCV) [4-6]). Results: COVs (%) from single-/ multiple-operator for WBA, HV and ERCV were 0.3/0.3, 0.1/1.9 and 1.7/10.6, respectively. ICC between gold standard and each operator was 0.96 for both WBA and HV, and 0.89 0.94 for ERCV. The estimated sample-size using WBA and HV was 225/225, 270/2157 in each treatment arm for single-/multiple-operator set up, respectively, and 2000 for ERCV. Conclusions: It was shown that highly reproducible and accurate methods were implemented using multiple-site/platform MRI which can be applied to detect subtle changes in large scale clinical trials. Singleoperator analysis per subject is recommended for HV to minimize required sample-size due to higher inter-operator variability and subject anatomical heterogeneity. Despite previously reported high sensitivity for early changes using ERCV, larger number of samples is required due to high operator variability based on manual-tracing methods. References: [1] Freeborough PA et al, IEEE Tran Med Img, 1997, 16(5), 623-629. [2] Hsu R et al, JMRI, 2002,16, 305-310. [3] Insausti R et al, AJNR, 1998, 19, 659-671. [4] Fox NC et al, Arch Neurol, 2000, 57, 339-344. [5] Ridha BH et al, Arch Neurol, 2007, 64, 849-854. [6] Du AT et al, Neurology, 2004, 62, 422-427.