Capturing Expert Knowledge for the Personalization of Cognitive Rehabilitation: Study Combining Computational Modeling and a Participatory Design Strategy

NeuroAIreh@b: an artificial intelligence-based methodology for personalized and adaptive neurorehabilitation

Cognitive impairments are a prevalent consequence of acquired brain injury, dementia, and age-related cognitive decline, hampering individuals' daily functioning and independence, with significant societal and economic implications. While neurorehabilitation represents a promising avenue for addressing these deficits, traditional rehabilitation approaches face notable limitations. First, they lack adaptability, offering one-size-fits-all solutions that may not effectively meet each patient's unique needs. Furthermore, the resource-intensive nature of these interventions, often confined to clinical settings, poses barriers to widespread, cost-effective, and sustained implementation, resulting in suboptimal outcomes in terms of intervention adaptability, intensity, and duration. In response to these challenges, this paper introduces NeuroAIreh@b, an innovative cognitive profiling and training methodology that uses an AI-driven framework to optimize neurorehabilitation prescription. NeuroAIreh@b effectively bridges the gap between neuropsychological assessment and computational modeling, thereby affording highly personalized and adaptive neurorehabilitation sessions. This approach also leverages virtual reality-based simulations of daily living activities to enhance ecological validity and efficacy. The feasibility of NeuroAIreh@b has already been demonstrated through a clinical study with stroke patients employing a tablet-based intervention. The NeuroAIreh@b methodology holds the potential for efficacy studies in large randomized controlled trials in the future.

Toward Personalized Web-Based Cognitive Rehabilitation for Patients With Ischemic Stroke: Elo Rating Approach

Background

Stroke is a worldwide cause of disability; 40% of stroke survivors sustain cognitive impairments, most of them following inpatient rehabilitation at specialized clinical centers. Web-based cognitive rehabilitation tasks are extensively used in clinical settings. The impact of task execution depends on the ratio between the skills of the treated patient and the challenges imposed by the task itself. Thus, treatment personalization requires a trade-off between patients’ skills and task difficulties, which is still an open issue. In this study, we propose Elo ratings to support clinicians in tasks assignations and representing patients’ skills to optimize rehabilitation outcomes.

Objective

This study aims to stratify patients with ischemic stroke at an early stage of rehabilitation into three levels according to their Elo rating; to show the relationships between the Elo rating levels, task difficulty levels, and rehabilitation outcomes; and to determine if the Elo rating obtained at early stages of rehabilitation is a significant predictor of rehabilitation outcomes.

Methods

The PlayerRatings R library was used to obtain the Elo rating for each patient. Working memory was assessed using the DIGITS subtest of the Barcelona test, and the Rey Auditory Verbal Memory Test (RAVLT) was used to assess verbal memory. Three subtests of RAVLT were used: RAVLT learning (RAVLT075), free-recall memory (RAVLT015), and recognition (RAVLT015R). Memory predictors were identified using forward stepwise selection to add covariates to the models, which were evaluated by assessing discrimination using the area under the receiver operating characteristic curve (AUC) for logistic regressions and adjusted R² for linear regressions.

Results

Three Elo levels (low, middle, and high) with the same number of patients (n=96) in each Elo group were obtained using the 50 initial task executions (from a total of 38,177) for N=288 adult patients consecutively admitted for inpatient rehabilitation in a clinical setting. The mid-Elo level showed the highest proportions of patients that improved in all four memory items: 56% (54/96) of them improved in DIGITS, 67% (64/96) in RAVLT075, 58% (56/96) in RAVLT015, and 53% (51/96) in RAVLT015R (P<.001). The proportions of patients from the mid-Elo level that performed tasks at difficulty levels 1, 2, and 3 were 32.1% (3997/12,449), 31.% (3997/12,449), and 36.9% (4595/12,449), respectively (P<.001), showing the highest match between skills (represented by Elo level) and task difficulties, considering the set of 38,177 task executions. Elo ratings were significant predictors in three of the four models and quasi-significant in the fourth. When predicting RAVLT075 and DIGITS at discharge, we obtained R²=0.54 and 0.43, respectively; meanwhile, we obtained AUC=0.73 (95% CI 0.64-0.82) and AUC=0.81 (95% CI 0.72-0.89) in RAVLT075 and DIGITS improvement predictions, respectively.

Conclusions

Elo ratings can support clinicians in early rehabilitation stages in identifying cognitive profiles to be used for assigning task difficulty levels.

A comparison of two personalization and adaptive cognitive rehabilitation approaches: a randomized controlled trial with chronic stroke patients

BACKGROUND

Paper-and-pencil tasks are still widely used for cognitive rehabilitation despite the proliferation of new computer-based methods, like VR-based simulations of ADL's. Studies have established construct validity of VR assessment tools with their paper-and-pencil version by demonstrating significant associations with their traditional construct-driven measures. However, VR rehabilitation intervention tools are mostly developed to include mechanisms such as personalization and adaptation, elements that are disregarded in their paper-and-pencil counterparts, which is a strong limitation of comparison studies. Here we compare the clinical impact of a personalized and adapted paper-and-pencil training and a content equivalent and more ecologically valid VR-based ADL's simulation.

METHODS

We have performed a trial with 36 stroke patients comparing Reh@City v2.0 (adaptive cognitive training through everyday tasks VR simulations) with Task Generator (TG: content equivalent and adaptive paper-and-pencil training). The intervention comprised 12 sessions, with a neuropsychological assessment pre, post-intervention and follow-up, having as primary outcomes: general cognitive functioning (assessed by the Montreal Cognitive Assessment - MoCA), attention, memory, executive functions and language specific domains.

RESULTS

A within-group analysis revealed that the Reh@City v2.0 improved general cognitive functioning, attention, visuospatial ability and executive functions. These improvements generalized to verbal memory, processing speed and self-perceived cognitive deficits specific assessments. TG only improved in orientation domain on the MoCA, and specific processing speed and verbal memory outcomes. However, at follow-up, processing speed and verbal memory improvements were maintained, and a new one was revealed in language. A between-groups analysis revealed Reh@City v2.0 superiority in general cognitive functioning, visuospatial ability, and executive functions on the MoCA.

CONCLUSIONS

The Reh@City v2.0 intervention with higher ecological validity revealed higher effectiveness with improvements in different cognitive domains and self-perceived cognitive deficits in everyday life, and the TG intervention retained fewer cognitive gains for longer.

TRIAL REGISTRATION

The trial is registered at ClinicalTrials.gov, number NCT02857803. Registered 5 August 2016, .