It takes two to pantomime: Communication meets motor cognition

Abstract

For over a century, pantomime of tool use has been employed to diagnose limb apraxia, a disorder of motor cognition primarily induced by left brain damage. While research consistently implicates damage to a left fronto-temporo-parietal network in limb apraxia, findings are inconsistent regarding the impact of damage to anterior versus posterior nodes within this network on pantomime. Complicating matters is the fact that tool use pantomime can be affected and evaluated at multiple levels. For instance, the production of tool use gestures requires the consideration of semantic characteristics (e.g. how to communicate the action intention) as well as motor features (e.g. forming grip and movement). Together, these factors may contribute substantially to apparent discrepancies in previously reported findings regarding neural correlates of tool use pantomime.In the current study, 67 stroke patients with unilateral left-brain damage performed a classic pantomime task. In order to analyze different error characteristics, we evaluated the proper use of grip and movement for each pantomime. For certain objects, healthy subjects may use body parts as representative for the object, e.g. use of the fingers to indicate scissors blades. To specify the pathological use of body parts as the object (BPO) we only assessed pantomime items that were not prone to this response in healthy participants. We performed modern voxel-based lesion analyses on MRI or CT data to determine associations between brain injury and the frequency of the specific types of pantomime errors.Our results support a model in which anterior and posterior nodes of the left fronto-temporo-parietal network contribute differentially to pantomime of tool use. More precisely, damage in the inferior frontal cortex reaching to the temporal pole is associated with an increased frequency of BPO errors, whereas damage to the inferior parietal lobe is predominantly linked to an increased frequency of movement and/or grip errors. Our work suggests that the validity of attempts to specify the neural correlates of limb apraxia based on tool use pantomime depends on differentiating the specific types of errors committed. We conclude that successful tool use pantomime involves dissociable functions with communicative aspects represented in more anterior (rather ventral) regions and motor-cognitive aspects in more posterior (rather dorsal) nodes of a left fronto-temporo-parietal network.

Post‐stroke cognitive deficits rarely come alone: Handling co‐morbidity in lesion‐behaviour mapping

“…A plethora of VLBM studies utilised nuisance regression or similar approaches to control the variable of interest for the variance explained by additional behavioural variables (e.g., Almairac et al, ; Baldo et al, , ; Finkel et al, ; Gajardo‐Vidal et al, ; Jones et al, ; Lorca‐Puls et al, ; Martin et al, ; Moon et al, ; Pillay et al, , ; Schwartz et al, ; Walker et al, ; Wilson et al, ; Winder et al, ). While some previous studies provide a rationale for covariate control, others do not provide any rationale at all or simply resort to a data driven procedure, where all known variables that correlate with the target symptom are controlled for.…”

“…A plethora of VLBM studies utilised nuisance regression or similar approaches to control the variable of interest for the variance explained by additional behavioural variables (e.g., Almairac et al, 2015;Baldo et al, 2013Baldo et al, , 2018Finkel et al, 2018;Gajardo-Vidal et al, 2018;Jones et al, 2016;Lorca-Puls et al, 2018;Martin et al, 2017;Moon et al, 2018;Pillay et al, 2014Pillay et al, , 2017Schwartz et al, 2011;Walker et al, 2011;Wilson et al, 2015;Winder et al, 2016).…”

The network underlying human higher-order motor control: Insights from machine learning-based lesion-behaviour mapping in apraxia

“…First, the above reported findings from mass-univariate studies already identified different areas that are spread across the brain, and that might be cortical nodes of such network. This point is also supported by componential analyses, that investigated multiple different sub-symptoms of pantomime (Manuel et al, 2013;Finkel et al, 2018). These studies found different areas to underlie behavioural components of apraxia, which could be part of a network for motor control.…”

“…Goldenberg et al, 2007;Manuel et al, 2013;Hoeren et al, 2014), but also in single cases (Kertesz and Ferro, 1984) and by a fMRI-DTI study (Vry et al, 2015). Thus, damage to a brain network of multiple cortical regions that are connected by white matter fibres likely underlies apraxia (see also Niessen et al, 2014;Vry et al, 2015;Finkel et al, 2018;Buxbaum & Randerath, 2018), which already has been postulated in classical models of praxis skills by Liepmann and Geschwind (see Buxbaum & Randerath, 2018).…”

“…Another potential major source for heterogeneous results could be the general analysis approach of the above mentioned studies. While they used different analysis techniquessuch as voxel-based lesion behaviour mapping (VLBM; Manuel et al, 2013;Hoeren et al, 2014;Goldenberg & Randerath, 2015;Weiss et al, 2016;Finkel et al, 2018), subtraction plots (Goldenberg et al, 2007;Weiss et al, 2008;Hermsdörfer et al, 2013), or region-ofinterest analyses (Halsband et al, 2001)all studies followed a univariate approach.…”