Using multi-level Bayesian lesion-symptom mapping to probe the body-part-specificity of gesture imitation skills

The neural correlates of limb apraxia: An anatomical likelihood estimation meta-analysis of lesion-symptom mapping studies in brain-damaged patients

Limb apraxia is a motor disorder frequently observed following a stroke. Apraxic deficits are classically assessed with four tasks: tool use, pantomime of tool use, imitation, and gesture understanding. These tasks are supported by several cognitive processes represented in a left-lateralized brain network including inferior frontal gyrus, inferior parietal lobe (IPL), and lateral occipito-temporal cortex (LOTC). For the past twenty years, voxel-wise lesion symptom mapping (VLSM) studies have been used to unravel the neural correlates associated with apraxia, but none of them has proposed a comprehensive view of the topic. In the present work, we proposed to fill this gap by performing a systematic Anatomic Likelihood Estimation meta-analysis of VLSM studies which included tasks traditionally used to assess apraxia. We found that the IPL was crucial for all the tasks. Moreover, lesions within the LOTC were more associated with imitation deficits than tool use or pantomime, confirming its important role in higher visual processing. Our results questioned traditional neurocognitive models on apraxia and may have important clinical implications.

Visuo-spatial complexity potentiates the body-part effect in intransitive imitation of meaningless gestures

Recent studies on the imitation of intransitive gestures suggest that the body part effect relies mainly upon the direct route of the dual-route model through a visuo-transformation mechanism. Here, we test the visuo-constructive hypothesis which posits that the visual complexity may directly potentiate the body part effect for meaningless gestures. We predicted that the difference between imitation of hand and finger gestures would increase with the visuo-spatial complexity of gestures. Second, we aimed to identify some of the visuo-spatial predictors of meaningless finger imitation skills. Thirty-eight participants underwent an imitation task containing three distinct set of gestures, that is, meaningful gestures, meaningless gestures with low visual complexity, and meaningless gestures with higher visual complexity than the first set of meaningless gestures. Our results were in general agreement with the visuo-constructive hypothesis, showing an increase in the difference between hand and finger gestures, but only for meaningless gestures with higher visuo-spatial complexity. Regression analyses confirm that imitation accuracy decreases with resource-demanding visuo-spatial factors. Taken together, our results suggest that the body part effect is highly dependent on the visuo-spatial characteristics of the gestures.

Neuroanatomy of reduced distortion of body-centred spatial coding during body tilt in stroke patients

Awareness of the direction of the body's (longitudinal) axis is fundamental for action and perception. The perceived body axis orientation is strongly biased during body tilt; however, the neural substrates underlying this phenomenon remain largely unknown. Here, we tackled this issue using a neuropsychological approach in patients with hemispheric stroke. Thirty-seven stroke patients and 20 age-matched healthy controls adjusted a visual line with the perceived body longitudinal axis when the body was upright or laterally tilted by 10 degrees. The bias of the perceived body axis caused by body tilt, termed tilt-dependent error (TDE), was compared between the groups. The TDE was significantly smaller (i.e., less affected performance by body tilt) in the stroke group (15.9 ± 15.9°) than in the control group (25.7 ± 17.1°). Lesion subtraction analysis and Bayesian lesion-symptom inference revealed that the abnormally reduced TDEs were associated with lesions in the right occipitotemporal cortex, such as the superior and middle temporal gyri. Our findings contribute to a better understanding of the neuroanatomy of body-centred spatial coding during whole-body tilt.

Bayesian lesion-deficit inference with Bayes factor mapping: Key advantages, limitations, and a toolbox

Statistical lesion-symptom mapping is largely dominated by frequentist approaches with null hypothesis significance testing. They are popular for mapping functional brain anatomy but are accompanied by some challenges and limitations. The typical analysis design and the structure of clinical lesion data are linked to the multiple comparison problem, an association problem, limitations to statistical power, and a lack of insights into evidence for the null hypothesis. Bayesian lesion deficit inference (BLDI) could be an improvement as it collects evidence for the null hypothesis, i.e. the absence of effects, and does not accumulate α-errors with repeated testing. We implemented BLDI by Bayes factor mapping with Bayesian t-tests and general linear models and evaluated its performance in comparison to frequentist lesion-symptom mapping with a permutation-based family-wise error correction. We mapped the voxel-wise neural correlates of simulated deficits in an in-silico-study with 300 stroke patients, and the voxel-wise and disconnection-wise neural correlates of phonemic verbal fluency and constructive ability in 137 stroke patients. Both the performance of frequentist and Bayesian lesion-deficit inference varied largely across analyses. In general, BLDI could find areas with evidence for the null hypothesis and was statistically more liberal in providing evidence for the alternative hypothesis, i.e. the identification of lesion-deficit associations. BLDI performed better in situations in which the frequentist method is typically strongly limited, for example with on average small lesions and in situations with low power, where BLDI also provided unprecedented transparency in terms of the informative value of the data. On the other hand, BLDI suffered more from the association problem, which led to a pronounced overshoot of lesion-deficit associations in analyses with high statistical power. We further implemented a new approach to lesion size control, adaptive lesion size control, that, in many situations, was able to counter the limitations imposed by the association problem, and increased true evidence both for the null and the alternative hypothesis. In summary, our results suggest that BLDI is a valuable addition to the method portfolio of lesion-deficit inference with some specific and exclusive advantages: it deals better with smaller lesions and low statistical power (i.e. small samples and effect sizes) and identifies regions with absent lesion-deficit associations. However, it is not superior to established frequentist approaches in all respects and therefore not to be seen as a general replacement. To make Bayesian lesion-deficit inference widely accessible, we published an R toolkit for the analysis of voxel-wise and disconnection-wise data.

The relationship between motor pathway damage and flexion-extension patterns of muscle co-excitation during walking

Background

Mass flexion-extension co-excitation patterns during walking are often seen as a consequence of stroke, but there is limited understanding of the specific contributions of different descending motor pathways toward their control. The corticospinal tract is a major descending motor pathway influencing the production of normal sequential muscle coactivation patterns for skilled movements. However, control of walking is also influenced by non-corticospinal pathways such as the corticoreticulospinal pathway that possibly contribute toward mass flexion-extension co-excitation patterns during walking. The current study sought to investigate the associations between damage to corticospinal (CST) and corticoreticular (CRP) motor pathways following stroke and the presence of mass flexion-extension patterns during walking as evaluated using module analysis.

Methods

Seventeen healthy controls and 44 stroke survivors were included in the study. We used non-negative matrix factorization for module analysis of paretic leg electromyographic activity. We typically have observed four modules during walking in healthy individuals. Stroke survivors often have less independently timed modules, for example two-modules presented as mass flexion-extension pattern. We used diffusion tensor imaging-based analysis where streamlines connecting regions of interest between the cortex and brainstem were computed to evaluate CST and CRP integrity. We also used a coarse classification tree analysis to evaluate the relative CST and CRP contribution toward module control.

Results

Interhemispheric CST asymmetry was associated with worse lower extremity Fugl-Meyer score (p = 0.023), propulsion symmetry (p = 0.016), and fewer modules (p = 0.028). Interhemispheric CRP asymmetry was associated with worse lower extremity Fugl-Meyer score (p = 0.009), Dynamic gait index (p = 0.035), Six-minute walk test (p = 0.020), Berg balance scale (p = 0.048), self-selected walking speed (p = 0.041), and propulsion symmetry (p = 0.001). The classification tree model reveled that substantial ipsilesional CRP or CST damage leads to a two-module pattern and poor walking ability with a trend toward increased compensatory contralesional CRP based control.

Conclusion

Both CST and CRP are involved with control of modules during walking and damage to both may lead to greater reliance on the contralesional CRP, which may contribute to a two-module pattern and be associated with worse walking performance.

Indirect structural disconnection-symptom mapping

In vivo tracking of white matter fibres catalysed a modern perspective on the pivotal role of brain connectome disruption in neuropsychological deficits. However, the examination of white matter integrity in neurological patients by diffusion-weighted magnetic resonance imaging bears conceptual limitations and is not widely applicable, as it requires imaging-compatible patients and resources beyond the capabilities of many researchers. The indirect estimation of structural disconnection offers an elegant and economical alternative. For this approach, a patient's structural lesion information and normative connectome data are combined to estimate different measures of lesion-induced structural disconnection. Using one of several toolboxes, this method is relatively easy to implement and is even available to scientists without expertise in fibre tracking analyses. Nevertheless, the anatomo-behavioural statistical mapping of structural brain disconnection requires analysis steps that are not covered by these toolboxes. In this paper, we first review the current state of indirect lesion disconnection estimation, the different existing measures, and the available software. Second, we aim to fill the remaining methodological gap in statistical disconnection-symptom mapping by providing an overview and guide to disconnection data and the statistical mapping of their relationship to behavioural measurements using either univariate or multivariate statistical modelling. To assist in the practical implementation of statistical analyses, we have included software tutorials and analysis scripts.

Distinct cognitive components and their neural substrates underlying praxis and language deficits following left hemisphere stroke

Apraxia is characterised by multiple deficits of higher motor functions, primarily caused by left hemisphere (LH) lesions to parietal-frontal praxis networks. While previous neuropsychological and lesion studies tried to relate the various apraxic deficits to specific lesion sites, a comprehensive analysis of the different apraxia profiles and the related (impaired) motor-cognitive processes as well as their differential neural substrates in LH stroke is lacking. To reveal the cognitive mechanisms that underlie the different patterns of praxis and (related) language deficits, we applied principal component analysis (PCA) to the scores of sub-acute LH stroke patients (n = 91) in several tests of apraxia and aphasia. Voxel-based lesion-symptom mapping (VLSM) analyses were then used to investigate the neural substrates of the identified components. The PCA yielded a first component related to language functions and three components related to praxis functions, with each component associated with specific lesion patterns. Regarding praxis functions, performance in imitating arm/hand gestures was accounted for by a second component related to the left precentral gyrus and the inferior parietal lobule. Imitating finger configurations, pantomiming the use of objects related to the face, and actually using objects loaded on component 3, related to the left anterior intraparietal sulcus and angular gyrus. The last component represented the imitation of bucco-facial gestures and was linked to the basal ganglia and LH white matter tracts. The results further revealed that pantomime of (limb-related) object use depended on both the component 2 and 3, which were shared with gesture imitation and actual object use. Data support and extend the notion that apraxia represents a multi-componential syndrome comprising different (impaired) motor-cognitive processes, which dissociate - at least partially - from language processes. The distinct components might be disturbed to a varying degree following LH stroke since they are associated with specific lesion patterns within the LH.

Strategies for feature extraction from structural brain imaging in lesion-deficit modelling

High‐dimensional modelling of post‐stroke deficits from structural brain imaging is highly relevant to basic cognitive neuroscience and bears the potential to be translationally used to guide individual rehabilitation measures. One strategy to optimise model performance is well‐informed feature selection and representation. However, different feature representation strategies were so far used, and it is not known what strategy is best for modelling purposes. The present study compared the three common main strategies: voxel‐wise representation, lesion‐anatomical componential feature reduction and region‐wise atlas‐based feature representation. We used multivariate, machine‐learning‐based lesion‐deficit models to predict post‐stroke deficits based on structural lesion data. Support vector regression was tuned by nested cross‐validation techniques and tested on held‐out validation data to estimate model performance. While we consistently found the numerically best models for lower‐dimensional, featurised data and almost always for principal components extracted from lesion maps, our results indicate only minor, non‐significant differences between different feature representation styles. Hence, our findings demonstrate the general suitability of all three commonly applied feature representations in lesion‐deficit modelling. Likewise, model performance between qualitatively different popular brain atlases was not significantly different. Our findings also highlight potential minor benefits in individual fine‐tuning of feature representations and the challenge posed by the high, multifaceted complexity of lesion data, where lesion‐anatomical and functional criteria might suggest opposing solutions to feature reduction.

The challenge of apraxia: Toward an operational definition?

The diagnosis of limb apraxia relies mainly on exclusion criteria (e.g., elementary motor or sensory deficits, aphasia). Due to the diversity of apraxia definitions and assessment methods, patients may or may not show apraxia depending on the chosen assessment method or theory, making the definition of apraxia somewhat arbitrary. As a result, "apraxia" may be diagnosed in patients with different cognitive impairments. Based on a quantitative and critical review of the literature, it is argued that this situation has its roots in the evolution from a task-based approach (i.e., the use of gold standard tests to detect apraxia) toward a process-based approach, namely, the deconstruction of the conceptual or production systems of action into multiple cognitive processes: language, executive functions, working memory, semantic memory, body schema, body image, visual-spatial skills, social cognition, visual-kinesthetic engrams, manipulation knowledge, technical reasoning, structural inference, and categorical apprehension. The coexistence of both approaches in the current literature is a major challenge that stands in the way of a scientific definition of apraxia. As a step toward a solution, we suggest to focus on symptoms, and on two complementary definition criteria (in addition with traditional exclusion criteria): Specificity (i.e., is apraxia explained by the alteration of cognitive processes specifically dedicated to gesture production?), and consistency (i.e., is the gesture production impairment consistent across tasks?). Two categories of limb apraxia are proposed: symptomatic apraxia (i.e., gesture production deficits that are secondary to more general cognitive impairments) and idiopathic apraxia (i.e., gesture production deficits that can be observed in isolation). It turns out that the only apraxia subtype that fulfills exclusion, specificity, and consistency criteria is limb-kinetic apraxia. A century after Liepmann's demonstration of the autonomy of apraxia toward language, the autonomy of this syndrome toward the rest of cognition remains an open question, while it poses new challenges to apraxia studies.

Effect of body-part specificity and meaning in gesture imitation in left hemisphere stroke patients

Previous studies showed that imitation of finger and hand/arm gestures could be differentially impaired after brain damage. However, so far, the interaction between gesture meaning and body part in imitation deficits has not been fully assessed. In the present study, we aimed at filling this gap by testing 36 unilateral left brain-damaged patients with and without apraxia (20 apraxics), and 29 healthy controls on an imitation task of either finger or hand/arm meaningful (MF) gestures and meaningless (ML) movements, using a large sample of stimuli and controlling for the composition of the experimental list. Left-brain damaged patients imitated ML finger worse than hand/arm movements, whereas they did not show the same difference in MF gesture imitation. In addition, apraxic patients imitated finger movements worse than hand/arm movements. Furthermore, apraxic patients' imitation performance was equally affected irrespective of the action meaning, whereas non-apraxic patients showed better imitation performance on MF gestures. Results suggest that MF gestures are processed as a whole, as imitation of these gestures relies on the stored motor programs in long-term memory, independently of the body part involved. In contrast, ML movements seem to be processed through direct visuo-motor transformations, with left-brain damage specifically disrupting imitation performance of the more cognitive demanding finger movements.

Imitation of meaningless gestures in normal aging

While imitation of meaningless gestures is a gold standard in the assessment of apraxia in patients with either stroke or neurodegenerative diseases, little is known about potential age-related effects on this measure. A significant body of literature has indicated that different mechanisms (i.e., executive functioning, visuospatial skills, sensory integration, body knowledge, categorical apprehension) may underlie the performance depending on imitation conditions (i.e., finger/hand, uni-/bimanual, symmetric/asymmetric, crossed/uncrossed configurations). However, neither the effects of these conditions on performance, nor the contribution of the abovementioned mechanisms to imitation have been explored in normal aging. The aim of the present study was to fill this gap. To do so, healthy adults (n = 103) aged 50 to 89 were asked to imitate 45 meaningless gestures. The authors controlled for general cognitive function, motor function, visual-spatial skills, executive function, sensory integration, body knowledge, and mechanical problem-solving skills. The results showed that asymmetry, body-midline crossing and, to a lesser extent, bimanual activity added an additional layer of difficulty to imitation tasks. After controlling for motor speed and cognitive function, age had an effect on imitation skills after 70 years old. This may reflect a decline in body knowledge, sensory integration, and executive functions. In contrast, the visuospatial and mechanical problem-solving hypotheses were ruled out. An additional motor simulation hypothesis is proposed. These findings may prove useful for clinicians working in memory clinics by providing insights on how to interpret imitation deficits. Lower performance after 70 years old should not be considered abnormal in a systematic manner.

Neural correlates of differential finger gesture imitation deficits in left hemisphere stroke

Behavioural studies in apraxic patients revealed dissociations between the processing of meaningful (MF) and meaningless (ML) gestures. Consequently, the existence of two differential neural mechanisms for the imitation of either gesture type has been postulated. While the indirect (semantic) route exclusively enables the imitation of MF gestures, the direct route can be used for the imitation of any gesture type, irrespective of meaning, and thus especially for ML gestures. Concerning neural correlates, it is debated which of the visuo-motor streams (i.e., the ventral steam, the ventro-dorsal stream, or the dorso-dorsal stream) supports the postulated indirect and direct imitation routes.

To probe the hypotheses that regions of the dorso-dorsal stream are involved differentially in the imitation of ML gestures and that regions of the ventro-dorsal stream are involved differentially in the imitation of MF gestures, we analysed behavioural (imitation of MF and ML finger gestures) and lesion data of 293 patients with a left hemisphere (LH) stroke.

Confirming previous work, the current sample of LH stroke patients imitated MF finger gestures better than ML finger gestures. The analysis using voxel-based lesion symptom mapping (VLSM) revealed that LH damage to dorso-dorsal stream areas was associated with an impaired imitation of ML finger gestures, whereas damage to ventro-dorsal regions was associated with a deficient imitation of MF finger gestures.

Accordingly, the analyses of the imitation of visually uniform and thus highly comparable MF and ML finger gestures support the dual-route model for gesture imitation at the behavioural and lesion level in a substantial patient sample. Furthermore, the data show that the direct route for ML finger gesture imitation depends on the dorso-dorsal visuo-motor stream while the indirect route for MF finger gesture imitation is related to regions of the ventro-dorsal visuo-motor stream.

•

Identification of differential neural correlates for the imitation of meaningful and meaningless finger gestures.

•

Support for the dual-route model for gesture imitation in a substantial patient sample (n = 293).

•

Left hemispheric damage to dorso-dorsal stream areas is associated with an impaired imitation of meaningless finger gestures

•

Damage to ventro-dorsal regions is associated with a deficient imitation of meaningful finger gestures

Repetitive Transcranial Magnetic Stimulation Over the Left Posterior Middle Temporal Gyrus Reduces Wrist Velocity During Emblematic Hand Gesture Imitation

Results from neuropsychological studies, and neuroimaging and behavioural experiments with healthy individuals, suggest that the imitation of meaningful and meaningless actions may be reliant on different processing routes. The left posterior middle temporal gyrus (pMTG) is one area that might be important for the recognition and imitation of meaningful actions. We studied the role of the left pMTG in imitation using repetitive transcranial magnetic stimulation (rTMS) and two-person motion-tracking. Participants imitated meaningless and emblematic meaningful hand and finger gestures performed by a confederate actor whilst both individuals were motion-tracked. rTMS was applied during action observation (before imitation) over the left pMTG or a vertex control site. Since meaningless action imitation has been previously associated with a greater wrist velocity and longer correction period at the end of the movement, we hypothesised that stimulation over the left pMTG would increase wrist velocity and extend the correction period of meaningful actions (i.e., due to interference with action recognition). We also hypothesised that imitator accuracy (actor-imitator correspondence) would be reduced following stimulation over the left pMTG. Contrary to our hypothesis, we found that stimulation over the pMTG, but not the vertex, during action observation reduced wrist velocity when participants later imitated meaningful, but not meaningless, hand gestures. These results provide causal evidence for a role of the left pMTG in the imitation of meaningful gestures, and may also be in keeping with proposals that left posterior temporal regions play a role in the production of postural components of gesture.