Research

Resources

Our lab at Emory University has the following resources:

  • Cognitive neuropsychology
    • Participants with brain damage for single-case studies.
    • Software and training in using voxel-based lesion-symptom mapping (VLSM) to understand structure-function relations in the brain.
  • Cognitive neuroscience
  • Cognitive psychology
    • Access to a large pool of undergraduate students for cognitive testing.
    • Screened individuals with mirror-touch synesthesia, grapheme-color synesthesia, and other interesting phenomena.
    • Software for experimental presentation and statistical analysis (E-Prime, MATLAB, Psychtoolbox, SPSS, etc.)

Research Topics

Below is a list of some topics we are currently working on in the Cognitive Neuropsychology lab. If you are a prospective graduate student, and these sound like interesting topics to you – do not hesitate to contact us! We are happy to talk about our research, and what you could do at Emory.

Neural plasticity

Studies with non-human primates observed cortical plasticity in somatosensory maps after brain damage, such that lesioned areas that previously represented a region of skin surface would reemerge in neighboring cortex. However, little systematic work had examined the sensory consequences of such changes. We found that individuals with damage to primary somatosensory cortex (S1) or specific white matter tracts can detect, but not accurately localize, touch on the hand (Rapp, Hendel & Medina, 2002; Liu et al., 2020). Based on these results, we proposed a process in which information is converted from distorted somatosensory maps to a veridical representation of the skin surface (Medina & Coslett, 2010).

Brain damage also alters interhemispheric somatosensory processing that, surprisingly, leads to the creation of new percepts. In a series of case studies, we have examined individuals with synchiria – a condition in which contralesional stimulation results in sensation on both sides. In one case, we found that phantom synchiric percepts were modulated by body position (Medina & Rapp, 2008). In a second case study, we reported an individual who experienced phantom tactile and visual synchiric percepts – with visual phantoms occurring primarily when visual stimuli were presented on the body itself (Medina, Drebing, Hamilton & Coslett, 2016). From these studies, we developed a model for interhemispheric processing and the body, proposing that synchiria is caused by removal of inhibitory interhemispheric processing.

More recently, we have asked how these changes are reflected in the intact brain. Using functional neuroimaging with individuals who have somatosensory cortical damage, we found that motor processing recruits visuomotor areas differently after such damage: a reweighting of the networks supporting movement, rather than a simple loss of function (Liu et al., 2025). This work connects the behavioral consequences of somatosensory damage described above to the reorganization of processing across the wider network, and is a direction we are actively pursuing.

Embodiment and the sense of body ownership

Under the right conditions, people will experience objects that are not part of their body as though they were – and, conversely, can fail to experience their own limbs as their own. We use illusions to ask what processes underlie this sense of ownership, and what constrains it. Using the mirror box, we have examined which components of the illusion are necessary for embodiment and which merely accompany it, separating the experience of ownership from related changes in perceived hand position (Leach & Medina, 2022). We have also developed an illusion in which participants experience a sixth finger (the Anne Boleyn illusion) and used it to test the limits of what the body representation will accept as part of itself (Ambron & Medina, 2023).

A supernumerary limb is a useful test case because it cannot be accommodated by simply revising the perceived position of an existing body part. In recent work, we found that embodiment of an illusory sixth finger is facilitated when the additional finger is anchored to the hand, but not when it occupies a spatially distinct mapping (Folger & Medina, 2025). Together, this work suggests that embodiment is not simply a matter of multisensory correlation, but is constrained by stored knowledge about the structure of the body, a theme that runs throughout our research on multisensory integration and body representation.

Multisensory integration and the body

Dominant models of multisensory integration propose that unimodal information is weighted by the precision of each modality (Ernst & Banks, 2002). However, these models lack information regarding prior knowledge from stored body representations. I recently led a three-institution consortium in a collaborative effort to examine how knowledge influences perception. We developed a novel illusion to examine how prior knowledge of biomechanical constraints (potential range of body motion) influences multisensory integration. When placing the hands in a mirror box in opposing postures (i.e., hidden hand palm up, viewed hand palm down) while making synchronous bimanual hand movements, participants experienced an illusory “flip” of perceived hand position to the visually defined estimate (Liu & Medina, 2017). Even though the participants were restricted from rotating their hands, we found significantly less illusory rotation as biomechanical constraints between visually- and proprioceptively-defined hand positions increased. These findings provide novel evidence that multisensory integration considers not only unimodal precision, but also information from stored body representations (see also Liu & Medina, 2018, 2021).

Frames of reference

Tactile information is first represented in a somatotopic frame of reference – i.e., relative to other landmarks on the skin surface, with information next transformed into external spatial coordinates relative to the organism’s body. Following up on earlier work examining visual spatial representations in individuals with spatial neglect and brain stimulation (Medina et al., 2009; Khurshid, Medina, et al., 2012; Shah-Basak, Chen, Caulfield, Medina & Hamilton, 2018), we have studied spatial representations for touch. Past studies with the Simon effect in vision and audition (faster reaction times when the stimulus occurs on the same side as the response) showed that the effect operated based on external representations of space. We discovered that the tactile Simon effect is generated based on somatotopic, not external, spatial representations (Medina et al., 2014). Building on this paradigm, we also discovered evidence for a separate, “hand-centered” representation specific to touch (Medina et al., 2019). Evidence for this representation was confirmed in a case study of an individual with subcortical damage who localized tactile stimuli to the left side of the hand, regardless of hand orientation, posture, or modality of response (Liu et al., 2020).

We have also examined how these representations are instantiated neurally. In a high-field (7T) functional imaging study, we found that hand posture influences processing in primary somatosensory cortex (Ambron, Garcea, Cason, Medina, Detre & Coslett, 2024), indicating that even early somatosensory areas are sensitive to the current configuration of the body rather than representing touch in purely somatotopic terms.

Replicability and best practices in cognitive neuroscience

Along with my research on body representations, I also have a strong interest in developing best practices to ensure that research is robust and replicable. We found that the majority of published voxel-based lesion-symptom mapping (VLSM) studies – a common method for relating regions of brain damage with specific deficits – used statistical tests inappropriately (Medina, Kimberg, Chatterjee & Coslett, 2010), leading to substantial revisions in practice in the field. More recently, studies have claimed that transcranial direct current stimulation (tDCS) – a low-cost brain stimulation method – influences cognitive processing. In a p-curve meta-analysis, we found no evidential value in a representative sample of cognitive and working memory tDCS studies (Medina & Cason, 2017)