Socially adaptive cognitive architecture for human-robot collaboration in industrial settings.

This paper introduces DAC-HRC, a novel cognitive architecture designed to optimize human-robot collaboration (HRC) in industrial settings, particularly within the context of Industry 4.0. The architecture is grounded in the Distributed Adaptive Control theory and the principles of joint intentionality and interdependence, which are key to effective HRC. Joint intentionality refers to the shared goals and mutual understanding between a human and a robot, while interdependence emphasizes the reliance on each other's capabilities to complete tasks. DAC-HRC is applied to a hybrid recycling plant for the disassembly and recycling of Waste Electrical and Electronic Equipment (WEEE) devices. The architecture incorporates several cognitive modules operating at different timescales and abstraction levels, fostering adaptive collaboration that is personalized to each human user. The effectiveness of DAC-HRC is demonstrated through several pilot studies, showcasing functionalities such as turn-taking interaction, personalized error-handling mechanisms, adaptive safety measures, and gesture-based communication. These features enhance human-robot collaboration in the recycling plant by promoting real-time robot adaptation to human needs and preferences. The DAC-HRC architecture aims to contribute to the development of a new HRC paradigm by paving the way for more seamless and efficient collaboration in Industry 4.0 by relying on socially adept cognitive architectures.

Patient-specific modeling for guided rehabilitation of stroke patients: the BrainX3 use-case.

BrainX3 is an interactive neuroinformatics platform that has been thoughtfully designed to support neuroscientists and clinicians with the visualization, analysis, and simulation of human neuroimaging, electrophysiological data, and brain models. The platform is intended to facilitate research and clinical use cases, with a focus on personalized medicine diagnostics, prognostics, and intervention decisions. BrainX3 is designed to provide an intuitive user experience and is equipped to handle different data types and 3D visualizations. To enhance patient-based analysis, and in keeping with the principles of personalized medicine, we propose a framework that can assist clinicians in identifying lesions and making patient-specific intervention decisions. To this end, we are developing an AI-based model for lesion identification, along with a mapping of tract information. By leveraging the patient's lesion information, we can gain valuable insights into the structural damage caused by the lesion. Furthermore, constraining whole-brain models with patient-specific disconnection masks can allow for the detection of mesoscale excitatory-inhibitory imbalances that cause disruptions in macroscale network properties. Finally, such information has the potential to guide neuromodulation approaches, assisting in the choice of candidate targets for stimulation techniques such as Transcranial Ultrasound Stimulation (TUS), which modulate E-I balance, potentiating cortical reorganization and the restoration of the dynamics and functionality disrupted due to the lesion.

Redefining stroke rehabilitation: Mobilizing the embodied goal-oriented brain.

Advancements in stroke rehabilitation remain limited and call for a reorientation. Based on recent results, this study proposes a network-centric perspective on stroke, positing that it not only causes localized deficits but also affects the brain's intricate network of networks, transiting it into a pathological state. Translating these system-level insights into interventions requires brain theory, and the Distributed Adaptive Control (DAC) theory offers such a framework. When applied in the rehabilitation gaming system, these principles demonstrate superior results over conventional methods. This impact stems from activating extensive brain networks, particularly the executive control network, focused motor learning, and maintaining excitatory-inhibitory balance, which is essential for neural repair and functional reorganization. The analysis stresses uniting preclinical and clinical research and placing the architecture of the embodied volitional brain at the centre of rehabilitation approaches.

Multiscale effects of excitatory-inhibitory homeostasis in lesioned cortical networks: A computational study.

Stroke-related disruptions in functional connectivity (FC) often spread beyond lesioned areas and, given the localized nature of lesions, it is unclear how the recovery of FC is orchestrated on a global scale. Since recovery is accompanied by long-term changes in excitability, we propose excitatory-inhibitory (E-I) homeostasis as a driving mechanism. We present a large-scale model of the neocortex, with synaptic scaling of local inhibition, showing how E-I homeostasis can drive the post-lesion restoration of FC and linking it to changes in excitability. We show that functional networks could reorganize to recover disrupted modularity and small-worldness, but not network dynamics, suggesting the need to consider forms of plasticity beyond synaptic scaling of inhibition. On average, we observed widespread increases in excitability, with the emergence of complex lesion-dependent patterns related to biomarkers of relevant side effects of stroke, such as epilepsy, depression and chronic pain. In summary, our results show that the effects of E-I homeostasis extend beyond local E-I balance, driving the restoration of global properties of FC, and relating to post-stroke symptomatology. Therefore, we suggest the framework of E-I homeostasis as a relevant theoretical foundation for the study of stroke recovery and for understanding the emergence of meaningful features of FC from local dynamics.

Drive competition underlies effective allostatic orchestration.

Living systems ensure their fitness by self-regulating. The optimal matching of their behavior to the opportunities and demands of the ever-changing natural environment is crucial for satisfying physiological and cognitive needs. Although homeostasis has explained how organisms maintain their internal states within a desirable range, the problem of orchestrating different homeostatic systems has not been fully explained yet. In the present paper, we argue that attractor dynamics emerge from the competitive relation of internal drives, resulting in the effective regulation of adaptive behaviors. To test this hypothesis, we develop a biologically-grounded attractor model of allostatic orchestration that is embedded into a synthetic agent. Results show that the resultant neural mass model allows the agent to reproduce the navigational patterns of a rodent in an open field. Moreover, when exploring the robustness of our model in a dynamically changing environment, the synthetic agent pursues the stability of the self, being its internal states dependent on environmental opportunities to satisfy its needs. Finally, we elaborate on the benefits of resetting the model's dynamics after drive-completion behaviors. Altogether, our studies suggest that the neural mass allostatic model adequately reproduces self-regulatory dynamics while overcoming the limitations of previous models.

From mechanisms to markers: novel noninvasive EEG proxy markers of the neural excitation and inhibition system in humans.

Brain function is a product of the balance between excitatory and inhibitory (E/I) brain activity. Variation in the regulation of this activity is thought to give rise to normal variation in human traits, and disruptions are thought to potentially underlie a spectrum of neuropsychiatric conditions (e.g., Autism, Schizophrenia, Downs' Syndrome, intellectual disability). Hypotheses related to E/I dysfunction have the potential to provide cross-diagnostic explanations and to combine genetic and neurological evidence that exists within and between psychiatric conditions. However, the hypothesis has been difficult to test because: (1) it lacks specificity-an E/I dysfunction could pertain to any level in the neural system- neurotransmitters, single neurons/receptors, local networks of neurons, or global brain balance - most researchers do not define the level at which they are examining E/I function; (2) We lack validated methods for assessing E/I function at any of these neural levels in humans. As a result, it has not been possible to reliably or robustly test the E/I hypothesis of psychiatric disorders in a large cohort or longitudinal patient studies. Currently available, in vivo markers of E/I in humans either carry significant risks (e.g., deep brain electrode recordings or using Positron Emission Tomography (PET) with radioactive tracers) and/or are highly restrictive (e.g., limited spatial extent for Transcranial Magnetic Stimulation (TMS) and Magnetic Resonance Spectroscopy (MRS). More recently, a range of novel Electroencephalography (EEG) features has been described, which could serve as proxy markers for E/I at a given level of inference. Thus, in this perspective review, we survey the theories and experimental evidence underlying 6 novel EEG markers and their biological underpinnings at a specific neural level. These cheap-to-record and scalable proxy markers may offer clinical utility for identifying subgroups within and between diagnostic categories, thus directing more tailored sub-grouping and, therefore, treatment strategies. However, we argue that studies in clinical populations are premature. To maximize the potential of prospective EEG markers, we first need to understand the link between underlying E/I mechanisms and measurement techniques.

Bringing rehabilitation home with an e-health platform to treat stroke patients: study protocol of a randomized clinical trial (RGS@home).

BACKGROUND: There is a pressing need for scalable healthcare solutions and a shift in the rehabilitation paradigm from hospitals to homes to tackle the increase in stroke incidence while reducing the practical and economic burden for patients, hospitals, and society. Digital health technologies can contribute to addressing this challenge; however, little is known about their effectiveness in at-home settings. In response, we have designed the RGS@home study to investigate the effectiveness, acceptance, and cost of a deep tech solution called the Rehabilitation Gaming System (RGS). RGS is a cloud-based system for delivering AI-enhanced rehabilitation using virtual reality, motion capture, and wearables that can be used in the hospital and at home. The core principles of the brain theory-based RGS intervention are to deliver rehabilitation exercises in the form of embodied, goal-oriented, and task-specific action. METHODS: The RGS@home study is a randomized longitudinal clinical trial designed to assess whether the combination of the RGS intervention with standard care is superior to standard care alone for the functional recovery of stroke patients at the hospital and at home. The study is conducted in collaboration with hospitals in Spain, Sweden, and France and includes inpatients and outpatients at subacute and chronic stages post-stroke. The intervention duration is 3 months with assessment at baseline and after 3, 6, and 12 months. The impact of RGS is evaluated in terms of quality of life measurements, usability, and acceptance using standardized clinical scales, together with health economic analysis. So far, one-third of the patients expected to participate in the study have been recruited (N = 90, mean age 60, days after stroke ≥ 30 days). The trial will end in July 2023. DISCUSSION: We predict an improvement in the patients' recovery, high acceptance, and reduced costs due to a soft landing from the clinic to home rehabilitation. In addition, the data provided will allow us to assess whether the prescription of therapy at home can counteract deterioration and improve quality of life while also identifying new standards for online and remote assessment, diagnostics, and intervention across European hospitals. TRIAL REGISTRATION: C linicalTrials.gov NCT04620707. Registered on November 3, 2020.

Saccade rate is associated with recall of items in working memory.

Working memory has been shown to rely on theta oscillations' phase synchronicity for item encoding and recall. At the same time, saccadic eye movements during visual exploration have been observed to trigger theta-phase resets, raising the question of whether the neuronal substrates of mnemonic processing rely on motor-evoked responses. To quantify the relationship between saccades and working memory load, we recorded eye tracking and behavioral data from human participants simultaneously performing an n-back Sternberg auditory task and a hue-based catch detection task. In addition to task-specific interference in performance, we also found that saccade rate was modulated by working memory load in the Sternberg task's preresponse stage. Our results support the possibility of interplay between saccades and hippocampal theta during working memory retrieval of items.

Rehabilitation Gaming System for Alcohol-Related Cognitive Impairment: A Pilot Usability Study.

AIMS: Cognitive impairment in patients with alcohol use disorder (AUD) is highly prevalent, and it negatively impacts treatment outcome. However, this condition is neither systematically assessed nor treated. Thus, we aimed to explore the usability of a virtual reality-based protocol ('Rehabilitation Gaming System', RGS) for patients with AUD. METHODS: Twenty AUD patients (50% also cognitive impairment) underwent a single session of the RGS protocol (four cognitive training tasks, 10 minutes each). System Usability Scale (SUS) and Post-Study System Usability Questionnaire (PSSUQ) were applied to assess the RGS usability and patients' satisfaction with it. Also, the Perceived Competence Scale was administered to assess the patients' feelings of competence when using the training protocol. Comparisons of the responses to these questionnaires were performed between AUD patients with cognitive impairment and those without cognitive impairment. RESULTS: RGS usability was very positively rated (median SUS score = 80, Interquartile Range, IQR = 68.13-86-88). No significant differences were found in the median SUS scores for any of the sociodemographic or clinical variables, excepting for gender (women median score = 85; IQR = 80-94.38 vs. men median score = 71.25; IQR = 61.25-89.25; P-value = 0.035). The quality of the information provided by the RGS training scenarios and the usability were positively rated (PSSUQ), and patients experienced high feelings of competence. CONCLUSIONS: The RGS has been found to be usable in the short term and patients with AUD stated to be satisfied with it. Future larger, randomized trials are needed to explore the effectiveness of this tool to help overcome the cognitive deficits in AUD patients.

Distinguishing Self, Other, and Autonomy From Visual Feedback: A Combined Correlation and Acceleration Transfer Analysis.

In cognitive science, Theory of Mind (ToM) is the mental faculty of assessing intentions and beliefs of others and requires, in part, to distinguish incoming sensorimotor (SM) signals and, accordingly, attribute these to either the self-model, the model of the other, or one pertaining to the external world, including inanimate objects. To gain an understanding of this mechanism, we perform a computational analysis of SM interactions in a dual-arm robotic setup. Our main contribution is that, under the common fate principle, a correlation analysis of the velocities of visual pivots is shown to be sufficient to characterize "the self" (including proximo-distal arm-joint dependencies) and to assess motor to sensory influences, and "the other" by computing clusters in the correlation dependency graph. A correlational analysis, however, is not sufficient to assess the non-symmetric/directed dependencies required to infer autonomy, the ability of entities to move by themselves. We subsequently validate 3 measures that can potentially quantify a metric for autonomy: Granger causality (GC), transfer entropy (TE), as well as a novel "Acceleration Transfer" (AT) measure, which is an instantaneous measure that computes the estimated instantaneous transfer of acceleration between visual features, from which one can compute a directed SM graph. Subsequently, autonomy is characterized by the sink nodes in this directed graph. This study results show that although TE can capture the directional dependencies, a rectified subtraction operation denoted, in this study, as AT is both sufficient and computationally cheaper.

Erratum: Entorhinal mismatch: A model of self-supervised learning in the hippocampus.

[This corrects the article DOI: 10.1016/j.isci.2021.102364.].

Entorhinal mismatch: A model of self-supervised learning in the hippocampus.

The hippocampal formation displays a wide range of physiological responses to different spatial manipulations of the environment. However, very few attempts have been made to identify core computational principles underlying those hippocampal responses. Here, we capitalize on the observation that the entorhinal-hippocampal complex (EHC) forms a closed loop and projects inhibitory signals "countercurrent" to the trisynaptic pathway to build a self-supervised model that learns to reconstruct its own inputs by error backpropagation. The EHC is then abstracted as an autoencoder, with the hidden layers acting as an information bottleneck. With the inputs mimicking the firing activity of lateral and medial entorhinal cells, our model is shown to generate place cells and to respond to environmental manipulations as observed in rodent experiments. Altogether, we propose that the hippocampus builds conjunctive compressed representations of the environment by learning to reconstruct its own entorhinal inputs via gradient descent.

Epistemic Autonomy: Self-supervised Learning in the Mammalian Hippocampus.

Biological cognition is based on the ability to autonomously acquire knowledge, or epistemic autonomy. Such self-supervision is largely absent in artificial neural networks (ANN) because they depend on externally set learning criteria. Yet training ANN using error backpropagation has created the current revolution in artificial intelligence, raising the question of whether the epistemic autonomy displayed in biological cognition can be achieved with error backpropagation-based learning. We present evidence suggesting that the entorhinal-hippocampal complex combines epistemic autonomy with error backpropagation. Specifically, we propose that the hippocampus minimizes the error between its input and output signals through a modulatory counter-current inhibitory network. We further discuss the computational emulation of this principle and analyze it in the context of autonomous cognitive systems.

Volitional learning promotes theta phase coding in the human hippocampus.

Electrophysiological studies in rodents show that active navigation enhances hippocampal theta oscillations (4-12 Hz), providing a temporal framework for stimulus-related neural codes. Here we show that active learning promotes a similar phase coding regime in humans, although in a lower frequency range (3-8 Hz). We analyzed intracranial electroencephalography (iEEG) from epilepsy patients who studied images under either volitional or passive learning conditions. Active learning increased memory performance and hippocampal theta oscillations and promoted a more accurate reactivation of stimulus-specific information during memory retrieval. Representational signals were clustered to opposite phases of the theta cycle during encoding and retrieval. Critically, during active but not passive learning, the temporal structure of intracycle reactivations in theta reflected the semantic similarity of stimuli, segregating conceptually similar items into more distant theta phases. Taken together, these results demonstrate a multilayered mechanism by which active learning improves memory via a phylogenetically old phase coding scheme.

Estimating upper-extremity function from kinematics in stroke patients following goal-oriented computer-based training.

INTRODUCTION: After a stroke, a wide range of deficits can occur with varying onset latencies. As a result, assessing impairment and recovery are enormous challenges in neurorehabilitation. Although several clinical scales are generally accepted, they are time-consuming, show high inter-rater variability, have low ecological validity, and are vulnerable to biases introduced by compensatory movements and action modifications. Alternative methods need to be developed for efficient and objective assessment. In this study, we explore the potential of computer-based body tracking systems and classification tools to estimate the motor impairment of the more affected arm in stroke patients. METHODS: We present a method for estimating clinical scores from movement parameters that are extracted from kinematic data recorded during unsupervised computer-based rehabilitation sessions. We identify a number of kinematic descriptors that characterise the patients' hemiparesis (e.g., movement smoothness, work area), we implement a double-noise model and perform a multivariate regression using clinical data from 98 stroke patients who completed a total of 191 sessions with RGS. RESULTS: Our results reveal a new digital biomarker of arm function, the Total Goal-Directed Movement (TGDM), which relates to the patients work area during the execution of goal-oriented reaching movements. The model's performance to estimate FM-UE scores reaches an accuracy of [Formula: see text]: 0.38 with an error ([Formula: see text]: 12.8). Next, we evaluate its reliability ([Formula: see text] for test-retest), longitudinal external validity ([Formula: see text] true positive rate), sensitivity, and generalisation to other tasks that involve planar reaching movements ([Formula: see text]: 0.39). The model achieves comparable accuracy also for the Chedoke Arm and Hand Activity Inventory ([Formula: see text]: 0.40) and Barthel Index ([Formula: see text]: 0.35). CONCLUSIONS: Our results highlight the clinical value of kinematic data collected during unsupervised goal-oriented motor training with the RGS combined with data science techniques, and provide new insight into factors underlying recovery and its biomarkers.

Excitatory-Inhibitory Homeostasis and Diaschisis: Tying the Local and Global Scales in the Post-stroke Cortex.

Maintaining a balance between excitatory and inhibitory activity is an essential feature of neural networks of the neocortex. In the face of perturbations in the levels of excitation to cortical neurons, synapses adjust to maintain excitatory-inhibitory (EI) balance. In this review, we summarize research on this EI homeostasis in the neocortex, using stroke as our case study, and in particular the loss of excitation to distant cortical regions after focal lesions. Widespread changes following a localized lesion, a phenomenon known as diaschisis, are not only related to excitability, but also observed with respect to functional connectivity. Here, we highlight the main findings regarding the evolution of excitability and functional cortical networks during the process of post-stroke recovery, and how both are related to functional recovery. We show that cortical reorganization at a global scale can be explained from the perspective of EI homeostasis. Indeed, recovery of functional networks is paralleled by increases in excitability across the cortex. These adaptive changes likely result from plasticity mechanisms such as synaptic scaling and are linked to EI homeostasis, providing a possible target for future therapeutic strategies in the process of rehabilitation. In addition, we address the difficulty of simultaneously studying these multiscale processes by presenting recent advances in large-scale modeling of the human cortex in the contexts of stroke and EI homeostasis, suggesting computational modeling as a powerful tool to tie the meso- and macro-scale processes of recovery in stroke patients.

Collaboration Variability in Autism Spectrum Disorder.

This paper addresses how impairments in prediction in young adults with autism spectrum disorder (ASD) relate to their behavior during collaboration. To assess it, we developed a task where participants play in collaboration with a synthetic agent to maximize their score. The agent's behavior changes during the different phases of the game, requiring participants to model the agent's sensorimotor contingencies to play collaboratively. Our results (n = 30, 15 per group) show differences between autistic and neurotypical individuals in their behavioral adaptation to the other partner. Contrarily, there are no differences in the self-reports of that collaboration.

Multisensory cueing facilitates naming in aphasia.

BACKGROUND: Impaired naming is a ubiquitous symptom in all types of aphasia, which often adversely impacts independence, quality of life, and recovery of affected individuals. Previous research has demonstrated that naming can be facilitated by phonological and semantic cueing strategies that are largely incorporated into the treatment of anomic disturbances. Beneficial effects of cueing, whereby naming becomes faster and more accurate, are often attributed to the priming mechanisms occurring within the distributed language network. OBJECTIVE: We proposed and explored two novel cueing techniques: (1) Silent Visuomotor Cues (SVC), which provided articulatory information of target words presented in the form of silent videos, and (2) Semantic Auditory Cues (SAC), which consisted of acoustic information semantically relevant to target words (ringing for "telephone"). Grounded in neurophysiological evidence, we hypothesized that both SVC and SAC might aid communicative effectiveness possibly by triggering activity in perceptual and semantic language regions, respectively. METHODS: Ten participants with chronic non-fluent aphasia were recruited for a longitudinal clinical intervention. Participants were split into dyads (i.e., five pairs of two participants) and required to engage in a turn-based peer-to-peer language game using the Rehabilitation Gaming System for aphasia (RGSa). The objective of the RGSa sessions was to practice communicative acts, such as making a request. We administered SVCs and SACs in a pseudorandomized manner at the moment when the active player selected the object to be requested from the interlocutor. For the analysis, we compared the times from selection to the reception of the desired object between cued and non-cued trials. RESULTS: Naming accuracy, as measured by a standard clinical scale, significantly improved for all stimuli at each evaluation point, including the follow-up. Moreover, the results yielded beneficial effects of both SVC and SAC cues on word naming, especially at the early intervention sessions when the exposure to the target lexicon was infrequent. CONCLUSIONS: This study supports the efficacy of the proposed cueing strategies which could be integrated into the clinic or mobile technology to aid naming even at the chronic stages of aphasia. These findings are consistent with sensorimotor accounts of language processing, suggesting a coupling between language, motor, and semantic brain regions. TRIAL REGISTRATION: NCT02928822 . Registered 30 May 2016.