원문정보
초록
영어
The Mann-Whitney-Wilcoxon (MWW) test statistic, while distribution-free, suffers from a loss of efficacy for certain underlying distributions. In this manuscript, we instead use a data-adaptive weighted generalized Mann-Whitney-Wilcoxon (AWGMWW) test statistic, one that is optimal in the Pitman Asymptotic Efficacy (PAE) sense, to discern differences in hippocampus shape among twin populations with or without Major Depressive Disorder (MDD). Using this more powerful statistic, we verify, based on a previous study using the MWW statistic, that a high-risk group is more similar to the control group than the depressed group, the depressed group is more similar to the control group than the high-risk group, and the control group cannot be distinguished as more similar to one group or the other in terms of hippocampus shape.
목차
1. Introduction
1.1. Nonparametric Statistics
1.2. Major Depressive Disorder
1.3. Computational Anatomy
1.4. Goals
2. Data
3. Methods
3.1. Image Processing
3.2. Statistical Analysis
4. Results
5. Conclusions
References