The neural networks of touch observation

Abstract Studies have consistently demonstrated that the mere observation of touch engages our own somatosensory cortices. However, a systematic evaluation of the involved networks is missing. Here, we present results of a meta-analytic connectivity modeling (MACM) approach based on clusters revealed by activation likelihood estimation (ALE) combined with resting-state analysis to detect networks subserving our ability to empathize with tactile experiences of other people. ALE analysis revealed 8 clusters in frontal, temporal, and parietal brain areas, which behavioral domain profiles predominantly refer to cognition and perception. The MACM analysis further identified distinct networks that are subserved by subcortical structures, revealed that all clusters involved in touch observation are connected to dorso-medial frontal and anterior cingulate cortex control regions, and showed that medial temporal lobe memory structures do not inform network activation during touch observation (confirmed by post hoc resting-state connectivity analyses). Our data highlight the importance of higher-level control areas and suggest only a minor role for past bodily experiences in the ad hoc perception of other people’s experiences.