A simple solution for model comparison in bold imaging: the special case of reward prediction error and reward outcomes

Abstract

Conventional neuroimaging techniques provide information about condition-related changes of the BOLD (blood-oxygen-level dependent) signal, indicating only where and when the underlying cognitive processes occur. Recently, with the help of a new approach called “model-based” functional neuroimaging (fMRI), researchers are able to visualize changes in the internal variables of a time varying learning process, such as the reward prediction error or the predicted reward value of a conditional stimulus. However, despite being extremely beneficial to the imaging community in understanding the neural correlates of decision variables, a model-based approach to brain imaging data is also methodologically challenging due to the multicollinearity problem in statistical analysis. There are multiple sources of multicollinearity in functional neuroimaging including investigations of closely related variables and/or experimental designs that do not account for this. The source of multicollinearity discussed in this paper occurs due to correlation between different subjective variables that are calculated very close in time. Here, we review methodological approaches to analyzing such data by discussing the special case of separating the reward prediction error signal from reward outcomes.

Reinforcement learning models and their neural correlates: An activation likelihood estimation meta-analysis

“…R. Roesch et al, 2010), although medial OFC activation has been shown to be coupled to RPE in some human fMRI studies. Our findings are consistent with the view that this is likely to be due to the correlation inherent between appetitive properties of the outcome and RPE in many of these designs (Erdeniz, Rohe, Done, & Seidler, 2013; Rohe et al, 2012). …”

Functional hemispheric asymmetries during the planning and manual control of virtual avatar movements

“…For this reason, three reduced general linear models were estimated and compared for each participant to decide, for each voxel separately, whether its BOLD response was overall better predicted by a model assuming activation only during visuomotor planning, only during online visuomotor processing or throughout both trial phases [ 50 ]. Each model comprised both sessions of the virtual avatar control task, high pass filtered with a high pass cut off at 128s to remove low frequency drifts, and an autoregressive model AR(1) to account for serial autocorrelations in the time series.…”

“…The three models were voxel-wise compared to reduce the risk of misattributing activation to a movement phase that contributed only little to the observed blood oxygen level dependent (BOLD) time-course. For each voxel and participant, we assessed whether log residual variance (ImCalc function of SPM) was smallest (greater model fit, more explained variance in the BOLD signal time-course) for the planning, online processing or combined model [ 50 ]. Afterwards, voxels whose BOLD signal changes were better described by the planning relative to the online processing and combined model (Planning log(residualVariance) < Execution log(residualVariance) ∩ Planning log(residualVariance) < Control log(residualVariance) , p < 0.05, uncorrected) were collected in a mask that was used for subsequent statistical interference in the planning model.…”

Value estimation and latent-state update-related neural activity during fear conditioning predict posttraumatic stress disorder symptom severity

“…To account for the potential impact of the electrotactile stimulation on neural activity, the design matrix also included a "shock" regressor. Because PE and the occurrence of the shock are highly correlated (Erdeniz, Rohe, Done, & Seidler, 2013), PE-related neural activity was not interpreted. Beta coefficients for the onset phase modulated by V and the outcome phase modulated by dB were included in the second-level analyses.…”