id	author	title	date	pages	extension	mime	words	sentences	flesch	summary	cache	txt
10_1101-2021_02_09_430363	Bayer, Johanna M. M.	Accommodating site variation in neuroimaging data using hierarchical and Bayesian models	2021	20	.pdf	application/pdf	13439	1891	70	Accommodating site variation in neuroimaging data using hierarchical and Bayesian models The potential of normative modeling to make individualized predictions has led to structural neuroimaging results that go beyond the case-control approach. in a similar way for multi-site modeling in a pooled neuroimaging data set, which contained 7499 participants that org/abide/) data set to compare a non-linear, Gaussian version of the model, to a linear hierarchical Bayesian version and mathematical description of our approach to include site as predictor in a normative hierarchical Bayesian model. With the aim to create reliable normative models in multi-site neuroimaging data, we developed and compared two model is also able to capture non-linear effects between age and thickness of the cortical region ("Hierarchical Bayesian Gaussian Process term, which allows to model non-linear association between age and cortical thickness measures. The only models that perform better for most regions than the mean of the training data set are the Hierarchical Bayesian	./cache/10_1101-2021_02_09_430363.pdf	./txt/10_1101-2021_02_09_430363.txt