Supplemental Digital Content 2

Supplemental Digital Content 2 - Control experiment
The aMPFC, the vPMC and the pSTC were more active in the human hand condition than in the
rubber hand condition. Theoretically though, this difference can result from the physical
differences between the two conditions and not from the somatosensory-motor interactions
between the subject and the experimenter. To eliminate this option we constructed a control
experiment addressed to estimate the influence of temperature, texture and imagery on the
activation described above.
Methods
Of the twelve participants of the experiment seven subjects also participated in a control
experiment. The control experiment design was identical to the original experiment. In this
experiment the four objects were replaced with flat surfaced soft bottles containing water. Since
the bottles were flat in their shape, the resulting stroking motor movements resembled the
movements in the “texture” condition in the original experiment. The bottles contained either
room temperature water (20-25°C) or warm water (30-40°C) and were covered with one of two
kinds of textures (smooth or rough). This setup resulted in four different bottles: hot-smooth, hotrough, cold-smooth and cold-rough. To control for imagery we also instructed the subjects to
imagine that they are touching a hand rather than a bottle in the hot-smooth condition. We chose
this condition since it most resembled (in a somatosensory manner) the experimenter's hand.
Results
The aMPFC, the vPMC and the pSTC regions of interest in the control experiment were taken in
the same manner as in the original experiment using only the seven control subject participating
in the analysis. Areas with increased activation for touching a real hand resembled those of the
12 subjects (see figure 3 below, left). There were no voxels showing significant differences in
activation between different temperatures, surface textures or imagery versus no imagery
conditions anywhere near the aMPFC, the vPMC and the pSTC (see figure 3, right).
For a “region of interest” analysis, the same regions were used as in the original experiment, i.e.
the aMPFC, the left vPMC and the right pSTC, that were taken from a multi study of only the
seven subjects that participated in the control experiment. Figure 4 presents the averaged percent
signal change for all four conditions across the control subjects from 6 to 9 seconds after the
condition onset. Comparison of the different temperatures and surface textures reveals no
differences in the aMPFC, the vPMC and the pSTC. Furthermore no effect of imagery was
found. This suggests that the activation for the human hand condition in the original experiment
did not result from temperature, texture or visual imagery confounds. Furthermore, the percent
signal change for all control conditions in each region was similar to the activation in the same
area for the simple texture condition from the original experiment. Meaning, the hemodynamic
response in the areas we have investigated did not change with changes in surface texture or
temperature but rather with the change in the object's meaning.
These two analyses suggest that in these areas the preference for the human hand rather than the
rubber one did not derive from differences in temperature, surface texture or imagery.
human hand>rubber hand
hot>cold
smoothed>textured
imagery>no imagery
Z=11
Figure 3 – Activation in the original experiment (left) and in the control experiment (right). On
the left, areas of increased signal for touching a human hand versus touching a rubber hand for
the seven control subjects. Significantly active voxels are shown in orange using FDR correction,
q(FDR)<0.01, p<0.007. On the right, no active voxels were revealed using the control contrasts.
Z=11 in Talairach space.
Figure 4 – Averaged percent signal change for the control experiment within the three regions of
interest. The response was averaged across all subjects from 6 to 9 seconds after the condition
onset. Light and dark gray represent (respectively from left to right) warm and room-temperature
temperatures, smooth and rough texture surfaces and conditions with or without imagery.
Regions of interest taken from a "human hand" > "rubber hand" contrast. aMPFC-anterior medial
prefrontal cortex; vPMC-ventral premotor cortex; pSTC-posterior superior temporal cortex.