Objective: Dysphagia is a common clinical condition characterized by difficulty in swallowing. It is sub-classified into oropharyngeal dysphagia, which refers to problems in the mouth and pharynx, and esophageal dysphagia, which refers to problems in the esophageal body and esophagogastric junction. Dysphagia can have a significant negative impact one’s physical health and quality of life as its severity increases. Therefore, proper assessment and management of dysphagia are critical for improving swallowing function and preventing complications. Thus a guideline was developed to provide evidence-based recommendations for assessment and management in patients with dysphagia. Methods: Nineteen key questions on dysphagia were developed. These questions dealt with various aspects of problems related to dysphagia, including assessment, management, and complications. A literature search for relevant articles was conducted using Pubmed, Embase, the Cochrane Library, and one domestic database of KoreaMed, until April 2021. The level of evidence and recommendation grade were established according to the Grading of Recommendation Assessment, Development and Evaluation methodology.Results: Early screening and assessment of videofluoroscopic swallowing were recommended for assessing the presence of dysphagia. Therapeutic methods, such as tongue and pharyngeal muscle strengthening exercises and neuromuscular electrical stimulation with swallowing therapy, were effective in improving swallowing function and quality of life in patients with dysphagia. Nutritional intervention and an oral care program were also recommended. Conclusion: This guideline presents recommendations for the assessment and management of patients with oropharyngeal dysphagia, including rehabilitative strategies.

Kim DY, et al. Brain Neurorehabil. 2023 Jul;16(2):e18. https://doi.org/10.12786/bn.2023.16.e18

Background: We investigated the clinical course of patients with prolonged disorders of consciousness (PDoC), predictors of emergence from PDoC (EDoC), and the temporal dynamics of six neurobehavior domains based on the JFK Coma Recovery Scale-Revised (CRS-R) during the recovery. Methods: A total of 50 traumatic and non-traumatic patients with PDoC were enrolled between October 2014 and February 2017. A retrospective analysis of the clinical findings and neurobehavioral signs was conducted using standardized methodology such as CRS-R. The findings were used to investigate the incidence and predictors of EDoC and determine the cumulative pattern of neurobehavioral recovery at 6 months, 1 year, and 2 years post-injury. Results: The results showed that 46% of the subjects emerged from PDoC after 200 median days (64–1197 days) of injury onset. The significant predictors of EDoC included minimally conscious state (MCS) (vs. vegetative state), higher auditory, communication, arousal, total CRS-R scores, shorter lag time post-injury, and the absence of intra-axial lesions. In terms of cumulative recovery of motor and communication signs in patients who emerged from PDoC, 39 and 32% showed EDoC at 6 months post-injury, and 88 and 93% exhibited EDoC at 2 years post-injury, respectively. Conclusions: Nearly half of the patients with PDoC recovered consciousness during inpatient rehabilitation. MCS, shorter lag time, the absence of intra-axial lesions, higher auditory, communication, arousal, and total CRS-R scores were important predictors for EDoC. Motor scores in the early stage of recovery and communication scores after prolonged intervals contributed to the higher levels of cumulative EDoC.

Transcranial Direct Current Stimulation (tDCS) has benefits for motor rehabilitation in stroke patients, but its clinical application is limited due to inter-individual heterogeneous effects. Recently, optimized tDCS that considers individual brain structure has been proposed, but the utility thereof has not been studied in detail. We explored whether optimized tDCS provides unique electrode positions for each patient and creates a higher target electric field than the conventional approach. A comparative within-subject simulation study was conducted using data collected for a randomized controlled study evaluating the effect of optimized tDCS on upper extremity function in stroke patients. Using Neurophet tES LAB 3.0 software, individual brain models were created based on magnetic resonance images and tDCS simulations were performed for each of the conventional and optimized configurations. A comparison of electrode positions between conventional tDCS and optimized tDCS was quantified by calculation of Euclidean distances. A total of 21 stroke patients were studied. Optimized tDCS produced a higher electric field in the hand motor region than conventional tDCS, with an average improvement of 20% and a maximum of 52%. The electrode montage for optimized tDCS was unique to each patient and exhibited various configurations that differed from electrode placement of conventional tDCS. Optimized tDCS afforded a higher electric field in the target of a stroke patient compared to conventional tDCS, which was made possible by appropriately positioning the electrodes. Our findings may encourage further trials on optimized tDCS for motor rehabilitation after stroke.

Cerebrovascular accidents (CVA) cause a range of impairments in coordination, such as a spectrum of walking impairments ranging from mild gait imbalance to complete loss of mobility. Patients with CVA need personalized approaches tailored to their degree of walking impairment for effective rehabilitation. This paper aims to evaluate the validity of using various machine learning (ML) and deep learning (DL) classification models (support vector machine, Decision Tree, Perceptron, Light Gradient Boosting Machine, AutoGluon, SuperTML, and TabNet) for automated classification of walking assistant devices for CVA patients. We reviewed a total of 383 CVA patients’ (1623 observations) prescription data for eight different walking assistant devices from five hospitals. Among the classification models, the advanced tree-based classification models (LightGBM and tree models in AutoGluon) achieved classification results of over $90\%$ accuracy, recall, precision, and F1-score. In particular, AutoGluon not only presented the highest predictive performance (almost $92\%$ in accuracy, recall, precision, and F1-score, and $86.8\%$ in balanced accuracy) but also demonstrated that the classification performances of the tree-based models were higher than that of the other models on its leaderboard. Therefore, we believe that tree-based classification models have potential as practical diagnosis tools for medical rehabilitation.

Repetitive and specific verbal cues by a therapist are essential in aiding a patient’s motivation and improving the motor learning process. The verbal cues comprise various expressions, sentences, volumes, and timings, depending on the therapist’s proficiency. This paper proposes an AI therapist (AI-T) that implements the verbal cues of professional therapists having extensive experience with robot-assisted gait training using the SUBAR for stroke patients. The AI-T was developed using a neuro-fuzzy system, a machine learning technique leveraging the benefits of fuzzy logic and artificial neural networks. The AI-T was trained with the professional therapist’s verbal cue data, as well as clinical and robotic data collected from robot-assisted gait training with real stroke patients. Ten clinical data and 16 robotic data are input variables, and six verbal cues are output variables. Fifty-eight stroke patients wore the SUBAR, a gait training robot, and participated in the robot-assisted gait training. A total of 9059 verbal cue data, 580 clinical data of stroke patients, and 144 944 robotic data were collected from 693 training sessions. Test results show that the trained AI-T can implement six types of verbal cues with 93.7% accuracy for the 1812 verbal cue data of the professional therapist. Currently, the trained AI-T is deployed in the SUBAR and provides six verbal cues to stroke patients in robot-assisted gait training.