The central nervous system (CNS) controls movements and regulates joint stiffness with muscle coactivation, but until now few studies have examined muscle pairs during running. This study aims to investigate differences in lower limb muscle coactivation during gait at different speeds from walking to running. Nineteen healthy runners walked and ran at speeds ranging from 0.8 km/h to 9.3 km/h. Twelve lower limb muscles' coactivation was calculated using the time-varying multi-muscle coactivation function (TMCf) with global, flexor-extension, and rostro-caudal approaches. Spatiotemporal and kinematic parameters were also measured. We found that TMCf, spatiotemporal and kinematic parameters were significantly affected by gait speed for all approaches. Significant differences were observed in the main parameters of each coactivation approach, and in the spatiotemporal, and kinematic parameters at the transition between walking and running. Positive and negative strong correlations were found between global coactivation parameters and center of mass displacements, as well as some spatiotemporal parameters, regardless of gait speed. Our findings suggest that walking and running have different coactivation patterns and kinematic characteristics, with the whole-limb stiffness exerted more synchronously and stably in running. The coactivation indexes and kinematic parameters could be the result of global coactivation, which is a sensory-control integration process used by the CNS to deal with more demanding and potentially unstable tasks like running.

Axial spondyloarthritis (axSpA) is a chronic rheumatic disease characterised by the presence of inflammatory back pain. In patients with chronic low back pain, the lumbar flexion relaxation phenomenon measured by surface electromiography (sEMG) differs from that in healthy individuals. However, sEMG activity in axSpA patients has not been studied. The purpose of this study was to analyse the flexion relaxation phenomenon in axSpA patients. A study evaluating 39 axSpA patients and 35 healthy controls was conducted. sEMG activity at the erector spinae muscles was measured during lumbar full flexion movements. sEMG activity was compared between axSpA patients and the controls, as well as between active (BASDAI≥4) and non-active (BASDAI<4) patients. The reliability (using intraclass correlation coefficients (ICC)), criterion validity and discriminant validity using the area Under the curve (AUC) for the inverse flexion/relaxation ratio (1/FRR) were evaluated. Significant differences (p<0.05) were observed between axSpA patients and the control group in lumbar electric activity, especially during flexion, relaxation and extension and in FRR and 1/FRR (0.66±0.39 vs. 0.25±0.19, respectively). In addition, significant differences were found between active and non-active but also between non-active and healthy subjects. The sEMG showed good reliability (ICC>0.8 for 1/FRR) and criterion validity. ROC analysis showed good discriminant validity for axSpA patients (AUC=0.835) vs. the control group using 1/FRR. An abnormal flexion/relaxation phenomenon exists in axSpA patients compared with controls. sEMG could be an additional objective tool in the evaluation of patient function and disease activity status.

Recently, the integration of neural networks into robotic exoskeletons for physical rehabilitation has become popular due to their ability to interpret complex physiological signals. Surface electromyography (sEMG), electromyography (EMG), electroencephalography (EEG), and other physiological signals enable communication between the human body and robotic systems. Utilizing physiological signals for communicating with robots plays a crucial role in robot assisted neurorehabilitation. This systematic review synthesizes 44 peer-reviewed studies, exploring how neural networks can improve exoskeleton robot assisted rehabilitation for individuals with impaired upper limbs. By categorizing the studies based on robot assisted joints, sensor systems, and control methodologies, we offer a comprehensive overview of neural network applications in this field. Our findings demonstrate that neural networks, such as Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM), and Radial Basis Function Neural Networks (RBFNN), and other forms of neural network significantly contribute to patient specific rehabilitation by enabling adaptive learning and personalized therapy. CNNs improve motion intention estimation and control accuracy, while LSTM networks capture temporal muscle activity patterns for real-time rehabilitation. RBFNNs improve human-robot interaction by adapting to individual movement patterns, leading to more personalized and efficient therapy. This review highlights the potential of neural networks to revolutionize upper limb rehabilitation, improving motor recovery and patient outcomes in both clinical and home-based settings. It also recommends the future direction to customize existing neural networks for robot assisted rehabilitation applications.