Spatial principal components analysis identifies ‘self-eigenmode’ response during social exchange (adapted from []). (A) Cingulate hemodynamic responses from the trustees’ brain during the ‘self’ and ‘other’ phases of the multi-round trust game. The spatially defined domains along the posterior to anterior axis were subjected to principal components analysis. Among the principal components identified was the ‘self-eigenmode’, which captures the dynamic agent-specific spatio-temporal activity in the cingulate as the trust game evolves. The self-eigenmode flips its sign as the game transitions from the self-phase to the other phase of the trust game. (B) Phase dynamics of cingulate response during multi-round trust game. The self-eigenmode response does not evolve in isolation. Three eigenvectors characterize the full set of the cingulate self-basis. Each of the red circles is a single TR in the measurement of the BOLD response during the self and other phases of the trust game in the trustee brain. These responses are plotted in the three dimensions that comprise the cingulate self-basis and shows how the cingulate response is dynamic, yet regular and throughout repeated trials of the self and other phases of the trust game. Increased self-responses are indicated by more positive values on the self-eigenmode axis (vertical axes); whereas non-self or ‘other’ responses are indicated by more negative values on the self-eigenmode. On the left we observe positive values on the self-eigenmode dimension leading up to an following the submission of a repayment by the trustee; and, on the right we observe negative self-eigenmode values in the trustee brain following the submission of an investment by the investor.