Enhanced image quality and lesion detection in FLAIR MRI of white matter hyperintensity through deep learning-based reconstruction.

OBJECTIVE: To delve deeper into the study of degenerative diseases, it becomes imperative to investigate whether deep-learning reconstruction (DLR) can improve the evaluation of white matter hyperintensity (WMH) on 3.0T scanners, and compare its lesion detection capabilities with conventional reconstruction (CR). METHODS: A total of 131 participants (mean age, 46 years ±17; 46 men) were included in the study. The images of these participants were evaluated by readers blinded to clinical data. Two readers independently assessed subjective image indicators on a 4-point scale. The severity of WMH was assessed by four raters using the Fazekas scale. To evaluate the relative detection capabilities of each method, we employed the Wilcoxon signed rank test to compare scores between the DLR and the CR group. Additionally, we assessed interrater reliability using weighted k statistics and intraclass correlation coefficient to test consistency among the raters. RESULTS: In terms of subjective image scoring, the DLR group exhibited significantly better scores compared to the CR group (P < 0.001). Regarding the severity of WMH, the DL group demonstrated superior performance in detecting lesions. Majority readers agreed that the DL group provided clearer visualization of the lesions compared to the conventional group. CONCLUSION: DLR exhibits notable advantages over CR, including subjective image quality, lesion detection sensitivity, and inter reader reliability.

Gated Recurrent Neural Network for Predicting the Plasmonic Colloid Composition from Spectra.

In current research on the synthesis of colloidal nanostructures, the size and morphology of nanoparticles still exhibit certain dispersion and variation from batch to batch. Characterization of size distribution and morphology distribution of nanoparticles often requires techniques such as scanning electron microscopy or transmission electron microscopy, which involve high vacuum environments, are time-consuming, and costly. Experienced researchers can roughly estimate the size and distribution of nanostructure from spectra for a given synthetic route, but the accuracy is often limited. This paper reports the potential of using neural networks to accurately predict the composition of colloidal nanostructures from spectra. We address several fundamental issues in neural network prediction of colloidal composition. We first demonstrate the prediction of the composition of a colloidal binary mixture of gold nanoparticles using a gated recurrent neural network (GRU). The evolution of prediction errors for scattering, absorption, and extinction spectra of nanostructures with sizes ranging from 5 to 120 nm are analyzed. Furthermore, we demonstrate that the neural network model operates robustly under white noise in experimental testing scenarios. Compared to fully connected neural networks, the gated recurrent unit exhibits better testing accuracy in spectral prediction. When confronted with experimental data that deviates from simulation outputs, minor adjustments to the training set can allow the predictions to align closely with the experimental spectra, paving the way for the characterization of complex colloidal compositions with artificial intelligence.

PNBACE: an ensemble algorithm to predict the effects of mutations on protein-nucleic acid binding affinity.

BACKGROUND: Mutations occurring in nucleic acids or proteins may affect the binding affinities of protein-nucleic acid interactions. Although many efforts have been devoted to the impact of protein mutations, few computational studies have addressed the effect of nucleic acid mutations and explored whether the identical methodology could be applied to the prediction of binding affinity changes caused by these two mutation types. RESULTS: Here, we developed a generalized algorithm named PNBACE for both DNA and protein mutations. We first demonstrated that DNA mutations could induce varying degrees of changes in binding affinity from multiple perspectives. We then designed a group of energy-based topological features based on different energy networks, which were combined with our previous partition-based energy features to construct individual prediction models through feature selections. Furthermore, we created an ensemble model by integrating the outputs of individual models using a differential evolution algorithm. In addition to predicting the impact of single-point mutations, PNBACE could predict the influence of multiple-point mutations and identify mutations significantly reducing binding affinities. Extensive comparisons indicated that PNBACE largely performed better than existing methods on both regression and classification tasks. CONCLUSIONS: PNBACE is an effective method for estimating the binding affinity changes of protein-nucleic acid complexes induced by DNA or protein mutations, therefore improving our understanding of the interactions between proteins and DNA/RNA.

Thiacalix[4]arene-Stabilized Sb/Ag Bimetallic Nanoclusters: Elucidating the Effects of Sb Doping on Electrocatalytic CO2 Reduction in Ag Clusters.

Accurately identifying the metal doping effects within heterogeneous catalysts presents a formidable challenge due to the complex nature of controlling the interfacial chemistry at the molecular level. Herein, we use two sets of atomically precise nanoclusters to demonstrate the impact of Sb doping on the electrocatalytic CO2 reduction activity in Ag nanoclusters. Leveraging the unique properties of the thiacalix[4]arene, we have pioneered a methodology for incorporating catalytic Ag1+ and Sb3+ sites, culminating in the synthesis of the pioneering Sb-Ag bimetallic cluster, Sb2Ag11. We refined this structure by replacing the two Sb3+ sites with Na+ sites, resulting in a Na2Ag10 cluster. Broadening our investigative scope, we isolated the core components from both Sb2Ag11 and Na2Ag10 and obtained two clusters: Sb2Ag4 and Ag4. The subtle compositional variations between two pairs of structurally analogous clusters, Sb2Ag11 and Na2Ag10, as well as Sb2Ag4 and Ag4, create opportunities to investigate how the Sb doping impacts the catalytic activity of Ag clusters. Clearly, compared to the undoped clusters, those doped with Sb exhibit higher catalytic current densities and enhanced CO selectivity. The theoretical calculations suggest that Sb doping can enhance the adsorption barrier of *H, thereby inhibiting hydrogen evolution activity and conversely promoting eCO2RR to CO activity.

Continual learning in medical image analysis: A survey.

In the dynamic realm of practical clinical scenarios, Continual Learning (CL) has gained increasing interest in medical image analysis due to its potential to address major challenges associated with data privacy, model adaptability, memory inefficiency, prediction robustness and detection accuracy. In general, the primary challenge in adapting and advancing CL remains catastrophic forgetting. Beyond this challenge, recent years have witnessed a growing body of work that expands our comprehension and application of continual learning in the medical domain, highlighting its practical significance and intricacy. In this paper, we present an in-depth and up-to-date review of the application of CL in medical image analysis. Our discussion delves into the strategies employed to address specific tasks within the medical domain, categorizing existing CL methods into three settings: Task-Incremental Learning, Class-Incremental Learning, and Domain-Incremental Learning. These settings are further subdivided based on representative learning strategies, allowing us to assess their strengths and weaknesses in the context of various medical scenarios. By establishing a correlation between each medical challenge and the corresponding insights provided by CL, we provide a comprehensive understanding of the potential impact of these techniques. To enhance the utility of our review, we provide an overview of the commonly used benchmark medical datasets and evaluation metrics in the field. Through a comprehensive comparison, we discuss promising future directions for the application of CL in medical image analysis. A comprehensive list of studies is being continuously updated at https://github.com/xw1519/Continual-Learning-Medical-Adaptation.

Development and External Validation of a Motor Intention-Integrated Prediction Model for Upper Extremity Motor Recovery After Intention-Driven Robotic Hand Training for Chronic Stroke.

OBJECTIVE: To derive and validate a prediction model for minimal clinically important differences (MCIDs) in upper extremity (UE) motor function after intention-driven robotic hand training using residual voluntary electromyography (EMG) signals from affected UE. DESIGN: A prospective longitudinal multicenter cohort study. We collected preintervention candidate predictors: demographics, clinical characteristics, Fugl-Meyer assessment of UE (FMAUE), Action Research Arm Test scores, and motor intention of flexor digitorum and extensor digitorum (ED) measured by EMG during maximal voluntary contraction (MVC). For EMG measures, recognizing challenges for stroke survivors to move paralyzed hand, peak signals were extracted from 8 time windows during MVC-EMG (0.1-5s) to identify subjects' motor intention. Classification and regression tree algorithm was employed to predict survivors with MCID of FMAUE. Relationship between predictors and motor improvements was further investigated. SETTING: Nine rehabilitation centers. PARTICIPANTS: Chronic stroke survivors (N=131), including 87 for derivation sample, and 44 for validation sample. INTERVENTIONS: All participants underwent 20-session robotic hand training (40min/session, 3-5sessions/wk). MAIN OUTCOME MEASURES: Prediction efficacies of models were assessed by area under the receiver operating characteristic curve (AUC). The best effective model was final model and validated using AUC and overall accuracy. RESULTS: The best model comprised FMAUE (cutoff score, 46) and peak activity of ED from 1-second MVC-EMG (MVC-EMG 4.604 times higher than resting EMG), which demonstrated significantly higher prediction accuracy (AUC, 0.807) than other time windows or solely using clinical scores (AUC, 0.595). In external validation, this model displayed robust prediction (AUC, 0.916). Significant quadratic relationship was observed between ED-EMG and FMAUE increases. CONCLUSIONS: This study presents a prediction model for intention-driven robotic hand training in chronic stroke survivors. It highlights significance of capturing motor intention through 1-second EMG window as a predictor for MCID improvement in UE motor function after 20-session robotic training. Survivors in 2 conditions showed high percentage of clinical motor improvement: moderate-to-high motor intention and low-to-moderate function; as well as high intention and high function.

Capture of RNA-binding proteins across mouse tissues using HARD-AP.

RNA-binding proteins (RBPs) modulate all aspects of RNA metabolism, but a comprehensive picture of RBP expression across tissues is lacking. Here, we describe our development of the method we call HARD-AP that robustly retrieves RBPs and tightly associated RNA regulatory complexes from cultured cells and fresh tissues. We successfully use HARD-AP to establish a comprehensive atlas of RBPs across mouse primary organs. We then systematically map RNA-binding sites of these RBPs using machine learning-based modeling. Notably, the modeling reveals that the LIM domain as an RNA-binding domain in many RBPs. We validate the LIM-domain-only protein Csrp1 as a tissue-dependent RNA binding protein. Taken together, HARD-AP is a powerful approach that can be used to identify RBPomes from any type of sample, allowing comprehensive and physiologically relevant networks of RNA-protein interactions.

Exploring the use of bimodal transcranial direct current stimulation to enhance movement in individuals with patellofemoral pain-A sham-controlled double blinded pilot study.

Introduction: In individuals with patellofemoral pain (PFP), addressing increased knee valgus during weight-bearing activities typically involves strengthening weak hip muscles. However, recent literature highlights the role of altered descending central control in abnormal movements associated with PFP. While transcranial direct current stimulation (tDCS) has demonstrated the capacity to enhance neuroplasticity, its application targeting the corticomotor function of gluteal muscles in PFP remains unexplored. This study aimed to investigate the effects of combining bimodal tDCS with exercise on frontal plane kinematics in individuals with PFP. The hypothesis was that bimodal tDCS, specifically targeting the corticomotor function of the gluteal muscles, would augment the effectiveness of exercise interventions in improving frontal plane kinematics compared to sham stimulation. Methods: Ten participants with PFP participated in two sessions involving either bimodal tDCS or sham stimulation, concurrently with hip strengthening exercises. Weight-bearing tasks, including single leg squat, single leg landing, single leg hopping, forward step-down, and lateral step-down, were performed and recorded before and after each session. Pain visual analog scale (VAS) scores were also documented. A one-way ANOVA with repeated measures was employed to compare kinematics, while a Friedman test was used to compare VAS across the three conditions (pre-test, post-tDCS, and post-Sham). Results: We observed no significant differences in trunk lean angle, hip and knee frontal plane projection angles, or dynamic valgus index among the three conditions during the five weight-bearing tasks. VAS scores did not differ across the three conditions. Discussion and conclusion: A single session of tDCS did not demonstrate immediate efficacy in enhancing frontal plane kinematics or relieving pain in individuals with PFP. Considering observed positive outcomes in other neurological and orthopedic populations with multi-session tDCS applications, suggesting potential cumulative effects, further research is essential to explore the effects of multi-session tDCS on weight-bearing movement and underlying neurophysiology in individuals with PFP.

Effects of high-definition tDCS targeting individual motor hotspot with EMG-driven robotic hand training on upper extremity motor function: a pilot randomized controlled trial.

BACKGROUND: Delivering HD-tDCS on individual motor hotspot with optimal electric fields could overcome challenges of stroke heterogeneity, potentially facilitating neural activation and improving motor function for stroke survivors. However, the intervention effect of this personalized HD-tDCS has not been explored on post-stroke motor recovery. In this study, we aim to evaluate whether targeting individual motor hotspot with HD-tDCS followed by EMG-driven robotic hand training could further facilitate the upper extremity motor function for chronic stroke survivors. METHODS: In this pilot randomized controlled trial, eighteen chronic stroke survivors were randomly allocated into two groups. The HDtDCS-group (n = 8) received personalized HD-tDCS using task-based fMRI to guide the stimulation on individual motor hotspot. The Sham-group (n = 10) received only sham stimulation. Both groups underwent 20 sessions of training, each session began with 20 min of HD-tDCS and was then followed by 60 min of robotic hand training. Clinical scales (Fugl-meyer Upper Extremity scale, FMAUE; Modified Ashworth Scale, MAS), and neuroimaging modalities (fMRI and EEG-EMG) were conducted before, after intervention, and at 6-month follow-up. Two-way repeated measures analysis of variance was used to compare the training effect between HDtDCS- and Sham-group. RESULTS: HDtDCS-group demonstrated significantly better motor improvement than the Sham-group in terms of greater changes of FMAUE scores (F = 6.5, P = 0.004) and MASf (F = 3.6, P = 0.038) immediately and 6 months after the 20-session intervention. The task-based fMRI activation significantly shifted to the ipsilesional motor area in the HDtDCS-group, and this activation pattern increasingly concentrated on the motor hotspot being stimulated 6 months after training within the HDtDCS-group, whereas the increased activation is not sustainable in the Sham-group. The neuroimaging results indicate that neural plastic changes of the HDtDCS-group were guided specifically and sustained as an add-on effect of the stimulation. CONCLUSIONS: Stimulating the individual motor hotspot before robotic hand training could further enhance brain activation in motor-related regions that promote better motor recovery for chronic stroke. TRIAL REGISTRATION: This study was retrospectively registered in ClinicalTrials.gov (ID NCT05638464).

Medial Prefrontal Cortex Stimulation Reduces Retrieval-Induced Forgetting via Fronto-parietal Beta Desynchronization.

The act of recalling memories can paradoxically lead to the forgetting of other associated memories, a phenomenon known as retrieval-induced forgetting (RIF). Inhibitory control mechanisms, primarily mediated by the prefrontal cortex, are thought to contribute to RIF. In this study, we examined whether stimulating the medial prefrontal cortex (mPFC) with transcranial direct current stimulation modulates RIF and investigated the associated electrophysiological correlates. In a randomized study, 50 participants (27 males and 23 females) received either real or sham stimulation before performing retrieval practice on target memories. After retrieval practice, a final memory test to assess RIF was administered. We found that stimulation selectively increased the retrieval accuracy of competing memories, thereby decreasing RIF, while the retrieval accuracy of target memories remained unchanged. The reduction in RIF was associated with a more pronounced beta desynchronization within the left dorsolateral prefrontal cortex (left-DLPFC), in an early time window (<500 ms) after cue onset during retrieval practice. This led to a stronger beta desynchronization within the parietal cortex in a later time window, an established marker for successful memory retrieval. Together, our results establish the causal involvement of the mPFC in actively suppressing competing memories and demonstrate that while forgetting arises as a consequence of retrieving specific memories, these two processes are functionally independent. Our findings suggest that stimulation potentially disrupted inhibitory control processes, as evidenced by reduced RIF and stronger beta desynchronization in fronto-parietal brain regions during memory retrieval, although further research is needed to elucidate the specific mechanisms underlying this effect.

Laparoscopy is non-inferior to open surgery for rectal cancer: A systematic review and meta-analysis.

BACKGROUND: Laparoscopic surgery has been endorsed by clinical guidelines for colon cancer, but not for rectal cancer on account of unapproved oncologic equivalence with open surgery. AIMS: We started this largest-to-date meta-analysis to comprehensively evaluate the safety and efficacy of laparoscopy in the treatment of rectal cancer compared with open surgery. MATERIALS & METHODS: Both randomized and nonrandomized controlled trials comparing laparoscopic proctectomy and open surgery between January 1990 and March 2020 were searched in PubMed, Cochrane Library and Embase Databases (PROSPERO registration number CRD42020211718). The data of intraoperative, pathological, postoperative and survival outcomes were compared between two groups. RESULTS: Twenty RCTs and 93 NRCTs including 216,615 patients fulfilled the inclusion criteria, with 48,888 patients received laparoscopic surgery and 167,727 patients underwent open surgery. Compared with open surgery, laparoscopic surgery group showed faster recovery, less complications and decreased mortality within 30 days. The positive rate of circumferential margin (RR = 0.79, 95% CI: 0.72 to 0.85, p < 0.0001) and distal margin (RR = 0.75, 95% CI: 0.66 to 0.85 p < 0.0001) was significantly reduced in the laparoscopic surgery group, but the completeness of total mesorectal excision showed no significant difference. The 3-year and 5-year local recurrence, disease-free survival and overall survival were all improved in the laparoscopic surgery group, while the distal recurrence did not differ significantly between the two approaches. CONCLUSION: Laparoscopy is non-inferior to open surgery for rectal cancer with respect to oncological outcomes and long-term survival. Moreover, laparoscopic surgery provides short-term advantages, including faster recovery and less complications.

Pedaling Asymmetry Reflected by Bilateral EMG Complexity in Chronic Stroke.

This study examines pedaling asymmetry using the electromyogram (EMG) complexity of six bilateral lower limb muscles for chronic stroke survivors. Fifteen unilateral chronic stroke and twelve healthy participants joined passive and volitional recumbent pedaling tasks using a self-modified stationary bike with a constant speed of 25 revolutions per minute. The fuzzy approximate entropy (fApEn) was adopted in EMG complexity estimation. EMG complexity values of stroke participants during pedaling were smaller than those of healthy participants (p = 0.002). For chronic stroke participants, the complexity of paretic limbs was smaller than that of non-paretic limbs during the passive pedaling task (p = 0.005). Additionally, there was a significant correlation between clinical scores and the paretic EMG complexity during passive pedaling (p = 0.022, p = 0.028), indicating that the paretic EMG complexity during passive movement might serve as an indicator of stroke motor function status. This study suggests that EMG complexity is an appropriate quantitative tool for measuring neuromuscular characteristics in lower limb dynamic movement tasks for chronic stroke survivors.

Patellar articular overlap is better associated with patellar alignment during weight-bearing than traditional measures of patellar height.

BACKGROUND: A consequence of a high riding patella is reduced osseous stability and malalignment of the patella (i.e., lateral patellar tilt and displacement). Although quantification of patellar height is a routine part of the radiographic examination of the patellofemoral joint, it is not clear which measure of patellar height is best associated with patella alignment. HYPOTHESIS/PURPOSE: To determine if patellar articular overlap (PAO) is better associated with lateral patellar tilt and lateral patellar displacement compared to traditional measures of patellar height. STUDY DESIGN: Cross-sectional. METHODS: Magnetic resonance images were obtained from 50 female participants (21 with patellofemoral pain and 29 healthy controls) under loaded conditions (25-35% bodyweight) at 15-20 degrees of knee flexion. Measurements of lateral patellar tilt and displacement as well as the PAO, Insall-Salvati ratio (ISV), Caton Deschamps-index (CD-index), or the Blacburn Peel-index (BP-index) were obtained from sagittal and axial plane images. RESULTS: The PAO was found to significantly correlated with lateral patellar tilt (r = -0.77, p < 0.001). In contrast, the ISV, CD-index, or the BP-index were not found to be associated with lateral patellar tilt (r = 0.13, p = 0.34; r = -0.14, p = 0.33; r = -0.08, p = 0.56, respectively). Both the PAO and ISV were found to be significantly correlated with lateral patellar displacement (r = -0.52, p < 0.001; r = 0.43, p = 0.002, respectively). Conversely, the CD-index and BP-index were not found to be associated with lateral patellar displacement (r = 0.03 p = 0.83; r = 0.05 p = 0.74, respectively). CONCLUSION: Of the measures of patellar height evaluated, the PAO was found to provide the greatest association with lateral patellar tilt and displacement.

Melatonin Ameliorates Depressive-Like Behaviors in Ovariectomized Mice by Improving Tryptophan Metabolism via Inhibition of Gut Microbe Alistipes Inops.

Melatonin (N-acetyl-5-methoxytryptamine) is reported to improve mood disorders in perimenopausal women and gut microbiome composition is altered during menopausal period. The possible role of microbiome in the treatment effect of melatonin on menopausal depression remains unknown. Here, it is shown that melatonin treatment reverses the gut microbiota dysbiosis and depressive-like behaviors in ovariectomy (OVX) operated mice. This effect of melatonin is prevented by antibiotic cocktails (ABX) treatment. Transferring microbiota harvested from adolescent female mice to OVX-operated mice is sufficient to ameliorate depressive-like behaviors. Conversely, microbiota transplantation from OVX-operated mice or melatonin-treated OVX-operated mice to naïve recipient mice exhibits similar phenotypes to donors. The colonization of Alistipes Inops, which is abundant in OVX-operated mice, confers the recipient with depressive-like behaviors. Further investigation indicates that the expansion of Alistipes Inops induced by OVX leads to the degradation of intestinal tryptophan, which destroys systemic tryptophan availability. Melatonin supplementation restores systemic tryptophan metabolic disorders by suppressing the growth of Alistipes Inops, which ameliorates depressive-like behaviors. These results highlight the previously unrecognized role of Alistipes Inops in the modulation of OVX-induced behavioral disorders and suggest that the application of melatonin to inhibit Alistipes Inops may serve as a potential strategy for preventing menopausal depressive symptoms.

18 F-FDG PET/CT of Benign Tracheal Schwannoma.

ABSTRACT: Schwannoma is a benign tumor originating from Schwann cells. It commonly occurs in the head, neck, and extremities, but rarely occurs in the trachea. Tracheal schwannoma is usually asymptomatic. We reported the 18 F-FDG PET/CT findings of a 61-year-old man with bronchoscopically biopsy-proven schwannoma, which presented challenges in differentiation from certain benign tumors and low-grade malignancies in the trachea.

Structural Variances in Curcumin Degradants: Impact on Obesity in Mice.

Some thermal degradants of curcuminoids have demonstrated moderate health benefits in previous studies. Feruloyl acetone (FER), recently identified as a thermal degradant of curcumin, has been previously associated with anticancer and antioxidative effects, yet its other capabilities remain unexplored. Moreover, earlier reports suggest that methoxy groups on the aromatic ring may influence the functionality of the curcuminoids. To address these gaps, an animal study was conducted to investigate the antiobesity effects of both FER and its demethoxy counterpart (DFER) on mice subjected to a high-fat diet. The results demonstrated the significant prevention of weight gain and enlargement of the liver and various adipose tissues by both samples. Furthermore, these supplements exhibited a lipid regulatory effect in the liver through the adiponectin/AMPK/SIRT1 pathway, promoted thermogenesis via AMPK/PGC-1α activation, and positively influenced gut-microbial-produced short-chain fatty acid (SCFA) levels. Notably, DFER demonstrated superior overall efficacy in combating obesity, while FER displayed a significant effect in modulating inflammatory responses. It is considered that SCFA may be responsible for the distinct effects of FER and DFER in the animal study. Future studies are anticipated to delve into the efficacy of curcuminoid degradants, encompassing toxicity and pharmacokinetic evaluations.

Task-Oriented Training by a Personalized Electromyography-Driven Soft Robotic Hand in Chronic Stroke: A Randomized Controlled Trial.

BACKGROUND: Intensive task-oriented training has shown promise in enhancing distal motor function among patients with chronic stroke. A personalized electromyography (EMG)-driven soft robotic hand was developed to assist task-oriented object-manipulation training effectively. Objective. To compare the effectiveness of task-oriented training using the EMG-driven soft robotic hand. METHODS: A single-blinded, randomized controlled trial was conducted with 34 chronic stroke survivors. The subjects were randomly assigned to the Hand Task (HT) group (n = 17) or the control (CON) group (n = 17). The HT group received 45 minutes of task-oriented training by manipulating small objects with the robotic hand for 20 sessions, while the CON group received 45 minutes of hand-functional exercises without objects using the same robot. Fugl-Meyer assessment (FMA-UE), Action Research Arm Test (ARAT), Modified Ashworth Score (MAS), Box and Block test (BBT), Maximum Grip Strength, and active range of motion (AROM) of fingers were assessed at baseline, after intervention, and 3 months follow-up. The muscle co-contraction index (CI) was analyzed to evaluate the session-by-session variation of upper limb EMG patterns. RESULTS: The HT group showed more significant improvement in FMA-UE (wrist/hand, shoulder/elbow) compared to the CON group (P < .05). At 3-month follow-up, the HT group demonstrated significant improvements in FMA-UE, ARAT, BBT, MAS (finger), and AROMs (P < .05). The HT group exhibited a more significant decrease in muscle co-contractions compared to the CON group (P < .05). CONCLUSIONS: EMG-driven task-oriented training with the personalized soft robotic hand was a practical approach to improving motor function and muscle coordination. CLINICAL TRIAL REGISTRY NAME: Soft Robotic Hand System for Stroke Rehabilitation. CLINICAL TRIAL REGISTRATION-URL: https://clinicaltrials.gov/. UNIQUE IDENTIFIER: NCT03286309.