Neural Network Underlying Recovery from Disowned Bodily States Induced by the Rubber Hand Illusion

Figure 2

(a) Dynamic causal modeling (DCM) for the RHI (9 models). Blank arrow indicates endogenous connection, and solid arrow indicates modulatory effect. (b) DCM for recovery from the RHI by needle rotation in the real hand (R-TS, 3 models) and in the rubber hand (R-VS, 3 models). Winning models from the random-effects Bayesian model selection for each DCM analysis are marked with a box: Model 3 for the RHI, Model A for recovery from the RHI by tactile stimuli, and Model C for recovery from the RHI by visual stimuli. Winning Model 3 for the RHI was used in DCM analyses for R-TS and R-VS in which the driving inputs from mechanical stimulation were differentially defined as entering into the SII (Model A), into the LOC (Model B), and into the SII and the LOC (Model C). (c) Bayesian model selection (BMS), winning model, and parameter analysis.

(a) DCM models for RHI

(b) DCM models for recovery from RHI by tactile or by visual stimuli

(c) BMS (RFX)