Stroke is a leading cause of long-term disability worldwide. With the advancements in sensor technologies and data availability, artificial intelligence (AI) holds the promise of improving the amount, quality and efficiency of care and enhancing the precision of stroke rehabilitation. We aimed to identify and characterize the existing research on AI applications in stroke recovery and rehabilitation of adults, including categories of application and progression of technologies over time. Data were collected from peer-reviewed articles across various electronic databases up to January 2024. Insights were extracted using AI-enhanced multi-method, data-driven techniques, including clustering of themes and topics. This scoping review summarizes outcomes from 704 studies. Four common themes (impairment, assisted intervention, prediction and imaging, and neuroscience) were identified, in which time-linked patterns emerged. The impairment theme revealed a focus on motor function, gait and mobility, while the assisted intervention theme included applications of robotic and brain–computer interface (BCI) techniques. AI applications progressed over time, starting from conceptualization and then expanding to a broader range of techniques in supervised learning, artificial neural networks (ANN), natural language processing (NLP) and more. Applications focused on upper limb rehabilitation were reviewed in more detail, with machine learning (ML), deep learning techniques and sensors such as inertial measurement units (IMU) used for upper limb and functional movement analysis. AI applications have potential to facilitate tailored therapeutic delivery, thereby contributing to the optimization of rehabilitation outcomes and promoting sustained recovery from rehabilitation to real-world settings.

Background: Chronic pain and somatosensory impairment are common following a stroke. It is possible that an interaction exists between pain and somatosensory impairment and that a change in one may influence the other. We therefore investigated the presence of chronic pain and self-reported altered somatosensory ability in individuals with stroke, aiming to determine if chronic pain is more common in stroke survivors with somatosensory impairment than in those without. Methods: Stroke survivors were invited to complete an online survey that included demographics, details of the stroke, presence of chronic pain, and any perceived changes in body sensations post-stroke. Results: Survivors of stroke (n = 489) completed the survey with 308 indicating that they experienced chronic pain and 368 reporting perceived changes in somatosensory function. Individuals with strokes who reported altered somatosensory ability were more likely to experience chronic pain than those who did not (OR = 1.697; 95% CI 1.585, 2.446). Further, this difference was observed for all categories of sensory function that were surveyed (detection of light touch, body position, discrimination of surfaces and temperature, and haptic object recognition). Conclusions: The results point to a new characteristic of chronic pain in strokes, regardless of nature or region of the pain experienced, and raises the potential of somatosensory impairment being a rehabilitation target to improve pain-related outcomes for stroke survivors.

Background: Chronic pain and body perception disturbance are common following stroke. It is possible that an interaction exists between pain and body perception disturbance, and that a change in one may influence the other. We therefore investigated the presence of body perception disturbance in individuals with stroke, aiming to determine if a perceived change in hand size contralateral to the stroke lesion is more common in those with chronic pain than in those without. Methods: Stroke survivors (N = 523) completed an online survey that included: stroke details, pain features, and any difference in perceived hand size post-stroke. Results: Individuals with stroke who experienced chronic pain were almost three times as likely as those without chronic pain to perceive their hand as now being a different size (OR = 2.895; 95%CI 1.844, 4.547). Further, those with chronic pain whose pain included the hand were almost twice as likely to perceive altered hand size than those whose pain did not include the hand (OR = 1.862; 95%CI 1.170, 2.962). This was not influenced by hemisphere of lesion (p = 0.190). Conclusions: The results point to a new characteristic of chronic pain in stroke, raising the possibility of body perception disturbance being a rehabilitation target to improve function and pain-related outcomes for stroke survivors.

Background: Chronic pain is common following stroke, yet its characteristics are poorly understood. Objectives: To characterize the beliefs and perceptions of people with stroke who experience chronic pain and compare these to a non-stroke population with chronic pain. Method: An online survey of the Pain Beliefs and Perceptions Inventory for stroke and non-stroke individuals with chronic pain. Pain beliefs and perceptions, including perceived causes of pain, were compared across the two groups. Results: A total of 223 participants completed all questions. The stroke group (N = 106) reported significantly higher pain intensities (p = <0.001) than the non-stroke group (N = 117). In addition, they identified the stroke as the primary cause of their pain, whereas the non-stroke group identified a body injury as the primary cause. When compared to the non-stroke group, people with stroke were more likely to report their pain as being constant (p = <0.001), demonstrate a lack of understanding regarding cause(s) of their pain (p = <0.001), believe less that their actions contribute to their pain (p = <0.001), and have greater belief that their pain is permanent (p = <0.001). Conclusion: These findings emphasize differences in the beliefs and perceptions associated with pain after stroke, supporting the value of targeted attention from research and therapy perspectives. Significant differences in the pain experienced by stroke and non-stroke groups may reflect a poor understanding of chronic pain following stroke. Future educational programs for stroke survivors should be tailored. Targets for future educational interventions are identified and include causes of pain, an understanding of its behavior and potential for improved prognosis.

Background and Purpose: Individuals with stroke often experience significant impairment of the upper limb. Rehabilitation interventions targeting the upper limb are typically associated with only small to moderate gains. The knowledge that body schema can be altered in other upper limb conditions has contributed to the development of tailored rehabilitation approaches. This study investigated whether individuals with stroke experienced alterations in body schema of the upper limb. If so, this knowledge may have implications for rehabilitation approaches such as motor imagery. Methods: An observational study performed online consisting of left/right judgment tasks assessed by response time and accuracy of: (i) left/right direction recognition; (ii) left/right shoulder laterality recognition; (iii) left/right hand laterality recognition; (iv) mental rotation of nonembodied objects. Comparisons were made between individuals with and without stroke. Secondary comparisons were made in the stroke population according to side of stroke and side of pain if experienced. Results: A total of 895 individuals (445 with stroke) participated. Individuals with stroke took longer for all tasks compared to those without stroke, and were less accurate in correctly identifying the laterality of shoulder (P < 0.001) and hand (P < 0.001) images, and the orientation of nonembodied objects (P < 0.001). Moreover, the differences observed in the hand and shoulder tasks were greater than what was observed for the control tasks of directional recognition and nonembodied mental rotation. No significant differences were found between left/right judgments of individuals with stroke according to stroke-affected side or side of pain. Discussion and Conclusions: Left/right judgments of upper limb are frequently impaired after stroke, providing evidence of alterations in body schema. The knowledge that body schemas are altered in individuals with longstanding stroke may assist in the development of optimal, well-accepted motor imagery programs for the upper limb. Video Abstract available for more insights from the authors (see the Video, Supplemental Digital Content 1, available at: http://links.lww.com/JNPT/A394).