Journal of Mechanics in Medicine and Biology

Background and Objective: Exposure to mental stress in everyday life leads to several diseases that deteriorate the quality of people’s lives, and hinders the development of society. The development of mobile technology and wearable devices has made it possible to monitor the mental stress in real time. This study aims to develop a single-channel EEG-based framework to identify the multiple types of mental stress. Methods: Four different tasks (i.e., arithmetic operations, successive subtraction, Stroop color-word test, and number memory) combining time pressure with negative feedback were designed to induce mental stress. The EEG data of 21 participants was recorded and a total of 338 multi-domain features, i.e., time domain, frequency domain, nonlinear and time–frequency domain, were extracted from five sub-frequency bands and full frequency band. After a one-way analysis of variance, recursive feature elimination and support vector machine were combined to classify the stress level of the participants. Results: For all four tasks, 67, 72, 20, and 74 features were statistically different, and they covered all the feature domains. Similarly, the highest classification accuracy of 97.78%, 97.50%, 95.00%, and 100% were achieved while combining features from all domains. Furthermore, the delta, theta, alpha, and full frequency band were more effective in quantifying the stress levels. Conclusion: Our proposed framework validates the effectiveness of a single-channel EEG in detecting mental stress and offers great promise for its application in clinical and portable devices in everyday life.

A key study area throughout the medical sector for image processing and analysis is medical image segmentation. The diagnosis and treatment strategies of doctors may have a solid foundation owing to accurate and effective medical image segmentation. Conventional approaches in this field rely on manual feature extraction, which makes segmentation complex, costs doctors’ time and energy, and involves a subjective evaluation that is readily susceptible to diagnostic errors. Researchers have applied convolutional neural network-based deep learning techniques to the segmentation of medical images as a result of their impressive advancements and successes in the field of computer vision. The research described here uses the U-Net network’s outstanding feature learning capabilities and end-to-end processing mode for lung CT image segmentation via fully convolutional network (FCN) research. However, focusing on valuable, crucial information aspects in the U-Net network is challenging. This study employed multilevel attention mechanisms on the basis of U-Net networks to enhance the model’s accuracy in lung CT image segmentation. These mechanisms were inspired by attention mechanisms. By improving the segmentation accuracy and optimizing the segmentation effect, the new model embeds a self-attention module in front of each upsampling layer in the U-Net model. This module provides more detailed information by stitching the self-attention module of the original image and then suppresses irrelevant and redundant information by using the effect of feature extraction of the upsampling layer. Several additional comparative experiments were conducted on the 2019nCoVR dataset. The outcomes demonstrate the efficacy of the optimized model described in this paper and its application results in improved segmentation effects in lung CT images. Additionally, the new model has distinct advantages over existing approaches that are typical of medical image segmentation, which represent its higher level of lung CT image segmentation.

The purpose of this study is to develop a near-infrared blood flow measurement algorithm and verify its performance. The gradient value of the measurement point was determined by using LED lights with wavelengths of 760nm and 850nm. This process involved measuring the gradient value for a significant number of regularly spaced cubic pixels centered around the chosen point. The measured gradient values for each pixel were then averaged to derive the final gradient value. Based on the measurement of the gradation value, when the concentration ratio of blood was altered by 20% between 0% and 100%, the gradation values at 760nm and 850nm changed significantly (more than $\pm 2.5\%$) in all cases. This study validated the efficacy of a self-developed near-infrared light system for detecting changes in blood flow. The ability to measure alterations in blood flow can be instrumental in predicting peripheral vascular disease (PVD) due to compromised circulatory conditions.

Objective: To investigate the effects of extremely low-frequency electromagnetic fields (ELF-EMF) on myogenesis during the early stage of denervated muscle atrophy in Sprague-Dawley (SD) rats. Methods: a total of 36 female SD rats were fed in cages and divided into three groups: the normal control group, the muscle atrophy group and the 10Hz/1.5mT magnetic field group. The rats in the 10Hz/1.5mT magnetic field group received a magnetic field intervention for 2h per day. The posterior muscle groups of the lower limbs of rats were sampled on the first, second, third and fourth weeks for evaluating the underlying changes. The harvested tissues were stained with hematoxylin–eosin (HE) dye to observe the morphology of the muscle atrophy and to measure the muscle fiber cross-sectional area ratio. A reverse transcription polymerase chain reaction was used to detect the expression of Myf5 mRNA in each group. The myogenesis-related proteins in the harvested muscle tissue were detected by western blot assay. Results: (1) HE staining showed that with the extension of denervation time, the structural disorder of muscle fibers was aggravated, and in the second week of the denervated muscle atrophy, muscle fiber cross-sectional area tended to increase in the 10Hz/1.5mT magnetic field group. Overall muscle atrophy was inevitable. (2) Results of RT-PCR showed that the expression of Myf5 mRNA in the 10Hz/1.5mT magnetic field group was significantly high in the second week of denervation, and the difference was statistically significant compared with the muscle atrophy group (*$p<0.05$). (3) Western blot analysis showed that the expression of MYF5 proteins in the 10Hz/1.5mT magnetic field group was significantly higher than that in the muscle atrophy group (**$p<0.01$) in the second week of denervation. Further, the expression of MYOG protein in the 10Hz/1.5mT magnetic field group was significantly higher than that in the muscle atrophy group (****$p<0.0001$). Conclusions: 10Hz/1.5mT ELF-EMF can significantly promote myogenesis during early denervated muscle atrophy in SD rats.

The classification of electroencephalogram (EEG) signals derived from motor imagery (MI) has always been a hot topic in the field of brain–computer interfaces. Due to its ability to handle the nonstationary and uncertain information contained in EEG signals, the Takagi–Sugeno–Kang fuzzy system (TSK-FS) has become an advantageous classification algorithm. To train a fuzzy system with strong discrimination capabilities from EEG data interspersed with redundant information, this paper proposes a TSK-FS modeling method based on low-rank sparse subspace learning (TSK-LSSL). This method focuses on consequent parameter learning, which transforms the traditional consequent parameter learning strategy into low-rank subspace and sparse subspace learning processes. Low-rank subspace learning is used to mine the global structural information of data and effectively reduce the number of fuzzy rules. During sparse subspace learning, ${\ell }_{2,1}$-norm regularization is used to constrain the consequent parameters and causes the number of redundant consequent parameters to be zero, thereby simplifying the fuzzy rules. In addition, a local boundary term based on graph matrices is embedded into the objective function to mine the local structural information of the given data. TSK-LSSL simplifies the number of rules and the consequent part of the fuzzy rules. It exhibits good classification performance on two BCI Competition databases.

Auditory binaural beats (BBs) stimulation is a method used to support the assumption that artificial external stimuli can generate specific brain waves. This study aimed to examine the effects of BBs that induce brain waves only based on a different frequency using an inaudible baseline frequency. The experiment was conducted among 18 adults in their 20s (11 male and 7 female). A baseline frequency of 18,000Hz and BBs with a difference frequency of 5Hz was presented, and the changes in electroencephalogram were observed. Power spectrum analysis was performed, and the absolute power of the theta-band frequency range (4–7Hz) was averaged. Absolute theta power was compared between phases without binaural beats (non-BBs) and with binaural beats ($\theta$-BBs) using the Wilcoxon signed-rank test. Theta waves in $\theta$-BBs significantly increased in most brain regions (central, parietal, temporal, and occipital) compared to those in non-BBs. This study is significant because it focused solely on the effects of the BBs-inducing theta waves using a frequency difference, excluding the effect of sound perception.

Mental health is critical to an individual’s life and social functioning and affects emotions, cognition and behavior. Mental health status assessments can help individuals understand their own psychological status, identify potential problems in real-time and implement effective interventions to promote favorable mental health. In this study, a deep learning approach was used to construct a simple-minded and flexible model for electroencephalogram (EEG)-based mental health status assessment to construct the corresponding M4EEG model. This model is suitable not only for supervised learning tasks containing a large amount of labeled data but also for few-shot classification tasks in special cases. During execution, certain components of a pretrained transformer model are utilized as the model’s foundation. After deriving feature values from different inputs, these features are decoupled by cross-connecting them into the relation module. Finally, the correlation between the outputs and the classification results are determined by a relation score. In experiments, the Database for Emotion Analysis using Physiological Signals (DEAP) and Affective Mood and Interpersonal Goals in the School Environment (AMIGOS) datasets were partitioned into K-Shot files as the input information, and the classification results were derived from the M4EEG model. These results showed that the M4EEG model is capable of assessing mental health status through EEG, and the model can obtain results that cannot be achieved by existing models that do not apply comparable data labeling.

To improve the feature extraction performance in vascular medical images, a multi-scale Retinex medical image enhancement algorithm based on a four-step method is proposed. This method is based on the multi-scale Retinex enhancement method, which divides the entire enhancement process into four steps: The first step is to reduce the overall complexity of the enhancement process through block processing; second, perform Retinex enhancement in each sub-block image; the third step is to optimize the connection area of each sub-block; and the fourth step is to highlight detailed features through gradient filtering. During the experiment, three different images were enhanced. The experimental results show that the proposed method can be used in enhancing the vascular features in medical images and be extended to other application fields. This method can effectively suppress the blurring and halo of the original image, while achieving enhancement effects and better protecting detailed features. From the four sets of indicators, the proposed method has shown more significant improvement than both reference methods. Regarding information entropy, the doubles have increased by a maximum of 96.2% and a minimum of 30.3%.

The need for sentiment analysis in the mental health field is increasing, and electroencephalogram (EEG) signals and music therapy have attracted extensive attention from researchers as breakthrough ideas. However, the existing methods still face the challenge of integrating temporal and spatial features when combining these two types, especially when considering the volume conduction differences among multichannel EEG signals and the different response speeds of subjects; moreover, the precision and accuracy of emotion analysis have yet to be improved. To solve this problem, we integrate the idea of top-k selection into the classic transformer model and construct a novel top-k sparse transformer model. This model captures emotion-related information in a finer way by selecting k data segments from an EEG signal with distinct signal features. However, this optimization process is not without its challenges, and we need to balance the selected k values to ensure that the important features are preserved while avoiding excessive information loss. Experiments conducted on the DEAP dataset demonstrate that our approach achieves significant improvements over other models. By enhancing the sensitivity of the model to the emotion-related information contained in EEG signals, our method achieves an overall emotion classification accuracy improvement and obtains satisfactory results when classifying different emotion dimensions. This study fills a research gap in the field of sentiment analysis involving EEG signals and music therapy, provides a novel and effective method, and is expected to lead to new ideas regarding the application of deep learning in sentiment analysis.

Chronic Ankle Instability (CAI) significantly affects athletic performance and increases the risk of recurrent ankle sprains. Stroboscopic Glasses (SG) may enhance proprioceptive feedback and postural control. We aimed to evaluate the augmented effects of Proprioceptive Feedback Training (PFT) combined with SG in athletes with CAI. We conducted a single-blinded randomized controlled trial with participants assigned to the PFT+SG or PFT alone groups. The interventions were administered over an eight-week period. Improvements in the Foot and Ankle Ability Measure Sport subscale (FAAM-S), Identification of Functional Ankle Instability questionnaire (IdFAI), Joint Position Sense (JPS) error value, postural stability via the Balance Velocity Test (BVT), Functional Performance Tests (FPTs), and giving-way episodes were assessed pre- and post-intervention. Substantial improvements in FAAM-S and IdFAI scores were observed in the PFT+SG group compared to those in the PFT group. JPS and BVT were significantly higher in the PFT+SG group than in the PFT group. FPTs and giving-way episodes were significantly higher in the PFT+SG group than in the PFT group. The inclusion of SG in PFT significantly enhanced rehabilitation outcomes in athletes with CAI. SG combined with PFT may address proprioceptive deficits and improve functional performance in patients with CAI.

The objective was to investigate the effects of running shoes with midsole hollow structure span and height on the biomechanics of the lower limbs during running. We collected 21 adults with running habits who wore two pairs of running shoes with different midsole hollow structures and ran at a speed of 3.3m/s on a force-measuring treadmill. The lower limb kinematics, ground reaction force (GRF) and lower limb muscle activation characteristics were simultaneously captured by a motion capture system, a 3D force treadmill, and a surface electromyography (sEMG) system. Paired t-tests were performed on data for the two shoe conditions that fit the normal distribution assumptions; otherwise, Wilcoxon signed-rank tests were used. The statistical parameter mapping (SPM) technology was used for the analysis of 1D parameters of kinematic, dynamic, and sEMG activation characteristics. The result showed that the time to the peak impact force at touchdown of Hollow shoe2 was significantly increased ($P<0.01$), the maximum loading rate ($P<0.01$) and average loading rate ($P<0.05$) were significantly reduced, braking time ($P<0.05$), push time ($P<0.05$), contact time ($P<0.01$) of Hollow shoe2 were significantly increased compared with Hollow shoe1. Hollow shoe2 push phase of the tibialis anterior muscle activation characteristics was significantly lower (SPM, $P<0.05$) than Hollow shoe1. Our conclusion is that running shoes offer the solution as they have the advantage of the complex structure of the hollow midsole.

The increased weight of multifunction and intelligent helmet systems can cause an exacerbation of the neck workload. Further, increased weight of the helmet can reduce the wearing comfort, and prolonged use may cause neck fatigue and even neck injury. This study evaluates the fatigue of neck muscles caused by wearing a helmet in motion using electromyography (EMG) measurement. Additionally, a comprehensive evaluation model for neck fatigue was established effectively. Twenty healthy subjects were selected to run for 10min continuously in each task by wearing two different helmets, respectively, during which the EMG signals of different neck muscles and subjective scores were recorded and analyzed. Two-way repeated measures analysis of variance (ANOVA) and Pearson correlation analysis were used to obtain indicators of neck fatigue, based on which the support vector machine (SVM) was used to construct a classification model of neck fatigue. By using integral EMG (IEMG) and root mean square (RMS) for assessing the neck injury in sternocleidomastoideus and trapezius muscles, it was observed that the mean power frequency (MPF) and median frequency (MF) in trapezius and splenius capitis could be used as effective evaluation indicators of neck fatigue. The EMG activity in trapezius was the highly sensitivity to fatigue when wearing helmets in motion. The classifier for distinguishing three different neck fatigue levels was 91.67% accurate. This study demonstrates the feasibility of an EMG-based method to evaluate the neck fatigue caused by wearing a helmet in motion, and also provides an evaluation method for the optimal design of helmets in the future.

Mental illnesses such as depression are typically neurologically related psychological disorders that affect people’s mood, thinking and behavior. As the number of students who are concerned about their mental health continues to rise, depression has emerged as a mental health concern that has a significant impact on both students’ academic performance and overall lives. To identify depression in students at an earlier stage, the purpose of this study was to provide a potentially unique approach. An approach to the detection of mental illnesses that is based on deep learning networks is proposed in this paper. First, facial expression and physical activity data are utilized for detecting depression. Second, the transformer model is utilized to extract the characteristics of the individual’s physical behavior, and the multiregional attention network (MRAN) is utilized to extract the characteristics of the individual’s emotions. The information that is obtained from the two modalities is complementary. Finally, at the fusion stage, this work applies the classification prediction of depression and nondepression (normal) at the decision level. This is done to ensure that the respective modal properties that were learned by the two channels are preserved in their entirety. We have demonstrated that our strategy is highly effective by performing experimental validation using a dataset that we developed ourselves. It is possible to identify depression in children at an earlier stage with the help of this effective remedy. It is anticipated that the findings of this study will provide an efficient screening tool for depression to educational institutions and organizations that focus on mental health, hence assisting students in receiving essential assistance and intervention at an earlier stage.

The bending stiffness of footwear impacts running efficiency and lower-limb work redistribution. However, studies integrating both aspects are scarce. This study aimed to examine the impact of footwear stiffness on lower-limb biomechanics, joint work, and overall metabolic efficiency. This study was performed on 12 male recreational runners to complete an experimental protocol while wearing two different running shoes with varying degrees of longitudinal bending stiffness. Paired-sample t tests were applied in this research. The stiffer footwear decreased the range of motion (ROM), angular velocity, negative and positive work of the metatarsophalangeal (MTP) joint, and ROM of the ankle joint. The running economy significantly improved with the stiffer footwear. However, no significant difference was observed in the joint work of other lower limbs. Shoes with increased bending stiffness significantly reduced the MTP joint ROM and angular velocity, thereby reducing negative power and work at the MTP joint and improving running economy for runners. This study provided relevant information for shoe designers, developers, and scientists conducting research on footwear midsole structures and designs.

The objective of this research is to investigate the potential mechanisms underlying electro-acupuncture intervention in chronic prostatitis (CP). Rats were randomly divided into five groups: a control group, a model group (suffering from CP), an electro-acupuncture group, an inhibitor group, and an activator group. One day post-modeling, electro-acupuncture intervention was conducted on the electro-acupuncture group, inhibitor group, and activator group at the acupoints “Baihuan Shu” (BL 30) and “Huiyang” (BL 35). Relevant indicators were used to evaluate the efficacy. Compared with models, the electro-acupuncture group showed significantly increased horizontal and vertical movement scores and sugar water intake, and significantly decreased prostate wet weight and prostate index. Compared with the electro-acupuncture group, the inhibitor group showed significantly increased horizontal and vertical movement scores and sugar water consumption. In contrast, for the activator group, these scores and consumption levels significantly decreased, while their prostate wet weight and prostate index increased significantly. Compared with models, the expression of TLR4 and p-NF-$\kappa$B/NF-$\kappa$B proteins in the prostate tissue of electro-acupuncture group was significantly lower. In terms of the inhibitor group, they also showed a decrease in TLR4 and p-NF-$\kappa$B/NF-$\kappa$B protein expression, whilst the activator group showed an increase. Therefore, electro-acupuncture at “Baihuan Shu” (BL 30) and “Huiyang” (BL 35) can significantly enhance movement scores, vertical movement scores, and sugar water consumption in rats with CP. Moreover, it can decrease the wet weight and index of the prostate, possibly by regulating the TLR4/NF-$\kappa$B pathway.

Depression is a prevalent mental condition, and it is essential to diagnose and treat patients as soon as possible to maximize their chances of rehabilitation and recovery. An intelligent detection model based on multimodal fusion technology is proposed based on the findings of this study to address the difficulties associated with depression detection. Text data and electroencephalogram (EEG) data are used in the model as representatives of subjective and objective nature, respectively. These data are processed by the BERT–TextCNN model and the CNN–LSTM model, which are responsible for processing them. While the CNN–LSTM model is able to handle time-series data in an effective manner, the BERT–TextCNN model is able to adequately capture the semantic information that is included in text data. This enables the model to consider the various features that are associated with the various types of data. In this research, a weighted fusion technique is utilized to combine the information contained within the two modal datasets. This strategy involves assigning a weight to the outcomes of each modal data processing in accordance with the degree of contribution that each modal data will make to produce the ultimate depression detection results. In regard to the task of depression identification, the suggested model demonstrates great validity and robustness, as demonstrated by the results of the experimental validation that we carried out on a dataset that we manufactured ourselves. A viable and intelligent solution for the early identification of depression is provided by the proposed model. This solution will likely be widely utilized in clinical practice and will provide new ideas and approaches for the growth of the field of precision medicine.

Background: Obesity, characterized by excess fat accumulation, is associated with heightened inflammation and infection risk. However, the relationship between body mass index (BMI) and wound healing remains unclear. Objective: The study aimed to explore the relationship between BMI and wound infection incidence, pathogenic bacterial strains in wounds, and wound healing promotion using different BMI values in conventional medical dressings. Methods: The study included 120 patients with traumatic and orthopedic wound infections and 40 patients with vertebral fractures to assess dressing efficacy and the influence of BMI index on healing outcomes. Wound infection incidence rates, pathogenic bacteria distribution within groups, and antibiotic resistance were identified, and their association with BMI was evaluated. Factors contributing to wound infection were analyzed. Results: The wound infection incidence rate was 30.36% without statistically significant correlation with BMI ($p>0.05$). A weak correlation ($0.3<r<0.5$) was established between the distribution of pathogenic bacterial strains with cause of trauma and BMI ($0.3<r<0.5$). Wound exposure time ($p=0.001$), wound depth ($p=0.044$), and hospital stay ($p=0.017$) were identified as independent wound infection risk factors. Conclusions: BMI exhibited a weak correlation with pathogenic bacteria distribution and negatively correlated with the degree of wound healing.

Objective: The objective of this study was to determine the characteristics that increase the likelihood of acute kidney injury (AKI) in patients with severe community-acquired pneumonia (SCAP) and to create a predictive nomogram for AKI. Methods: This study comprised individuals who received a diagnosis of SCAP over the period from January 01, 2019, to December 31, 2023. The patients were categorized into two groups: AKI and non-AKI. The clinical and demographic characteristics of the patients were extracted from their medical records. An analysis was conducted to compare the rates of survival at 30 and 90 days among various groups. A multivariate analysis was performed to discover the autonomous risk factors linked to SCAP. The nomogram was built based on these parameters. A receiver operating characteristics (ROC) curve study was performed to assess the predictive accuracy of the nomogram, namely by measuring the area under the curve (AUC). Results: Initial screening was conducted on a total of 1218 patients. After excluding 744 individuals who did not meet the exclusion criteria, a total of 474 patients, with an average age of $74.22\pm 15.16$ years and a female representation of 33.3%, were selected for inclusion in this study. The prevalence of AKI in the subjects with SCAP was 47.7%. Out of these instances, 39.8% were categorized as AKI stage 1, 31.0% as AKI stage 2, and 29.2% as AKI stage 3. Those diagnosed with AKI exhibited a significantly higher mortality rate at both the 30-day and 90-day marks in comparison to those who did not have AKI. The independent risk factors for AKI were determined to include age, male gender, chronic renal disease, diabetes, and the utilization of nonsteroidal anti-inflammatory medicines (NSAIDs). In addition, higher levels of baseline serum creatinine and uric acid were identified as risk factors for AKI. The final predictive nomogram achieved an AUC of 0.811, with a 95% confidence interval (CI) ranging from 0.773 to 0.849. Conclusion: Our nomogram can serve as a valuable tool for evaluating the progression of AKI in patients with SCAP.

The advanced technique comprising an enhanced multi-W-shaped gauge and Taguchi-based analytical program of gamma camera (GC) spatial resolution optimization was proposed and successfully applied to the routine diagnosis of nuclear examination. The indigenous gauge was revised from a V-shaped gauge for computer tomography (CT) scanning. The multi-W-shaped slit had a gradually narrowing gap between two adjacent V-slits; once the gap was beyond the GC recognizable limit, the acquired imaging blurred, and the discrimination between V-slits was lost. The five factors of GC data collecting were assigned as A (detector to target distance), B (total counts), C (zoom of ROI), D (matrix size), and E (energy width). Taguchi’s orthogonal array (${\text{L}}_{18}$) of factor combinations was adopted to arrange them into 18 groups. The reproducibility was ensured by the manual ranking of the acquired SPECT imaging quality by three seasoned radiology experts in three discrete periods. Next, Matlab and OriginPro software packages performed the SPECT analysis via the Poisson distribution program, identifying the peak center and FWHM of peaks of interest in SPECT imaging to quantify the spatial resolution and calculate the minimum detectable difference (MDD). The smallest MDD of 6.21mm revealed the optimal combination of factors for GC data collecting: A1 (DTD 5cm), B3 (TC 225), C3 (Zm 1.0), D1 (MS $256\times 256$), and E2 (EW 10%). The conventional preset and group 7 (with the highest signal-to-noise ratio) had MDD values of 8.19mm and 7.62mm.

In the field of medical image segmentation, deep convolutional neural networks have achieved satisfying performance in the past decade or so. However, there are some shortcomings. First, the convolutional neural network model cannot provide good insight into the remote dependencies in the image. Second, medical imaging datasets are typically small, which leads to a much higher risk of overfitting in model training. To address these limitations, we innovatively designed the skipped features enhancer (SFE) to enhance the impact of preserved details. To gain insight into remote dependencies in images, this model (SFE-TransUNet) is based on transformer. Additionally, a different scale convolutional layer (additional information capturer) before and after the Transformer Encoder to fuse the features to retain more information of the original data. In addition, a gate mechanism was introduced in the multi-head self-attention (MHSA), and. Finally, an attention block with residuals. SFE-TransUNet was evaluated on two public medical image segmentation datasets. Experimental results show that it achieves better performance than other related Transformer-based architectures. Code available at https://github.com/xackz/SFE-TransUNet.

With the increase in academic pressure on college students, some students show signs of academic burnout, which may lead to mental health problems. This study aims to explore the mental health status of college students under the manifestation of academic burnout and to propose a mental health detection model based on electroencephalography (EEG) signals. In this study, college students with the manifestation of study burnout were selected as research subjects, and their EEG signals were collected to identify and analyze their psychological state using a mental health recognition model. In the selection of mental health recognition models, this study used a variety of classical models for comparative analysis, including support vector machine (SVM), k-nearest neighbor (KNN), logistic regression (LR), convolutional neural network (CNN), long and short-term memory network (LSTM), and convolutional neural network-long and short-term memory network (CNN-LSTM). Experimental results show that the CNN-LSTM model performs best in identifying students with mental health problems, and its accuracy and stability are better than other models. Through this study, we conclude that academic burnout is closely related to college students’ mental health problems, and EEG signals combined with the CNN-LSTM model can effectively identify the mental health status of these students. This finding provides a new method and tool for monitoring college students’ mental health, which helps detect and intervene in students’ mental health problems early and promote their healthy growth and development.

The end effector of a medical robotic arm based on human bionics is responsible for surgical operations, and reasonable end effector trajectory operation is significant to robots. A deep learning method based on Faster R-CNN network is proposed for optimizing the end trajectory of a seven-degrees-of-freedom biomimetic medical robotic arm. In this method, RPN replaces conventional search in deep learning, and RoI Pooling completes network partitioning, thereby improving the efficiency of deep learning. The experimental results show that, with four data types as inputs: velocity, angular velocity, acceleration, and angular acceleration, the Faster R-CNN network achieves effective optimization results. The end trajectory of the seven-degrees-of-freedom biomimetic medical robotic arm becomes smooth, and the impact force during the speed reversal process is significantly reduced.

Reducing the occurrence of diseases has become an important research direction in today’s social medicine. Accurate prediction of heart failure can greatly reduce the complication rate, allowing patients to know the condition in advance and receive treatment. To deal with heart failure, this paper proposes a heart failure prediction algorithm based on a reinforcement learning framework. This paper combines and improves the reinforcement learning algorithm and the swarm intelligence (SI) optimization algorithm, aiming to use the reinforcement learning algorithm to improve the global optimization capability of the SI optimization algorithm. In order to better improve the global search ability of the SI optimization algorithm, the penalty function will be replaced by a combination mode of dynamic and static reward. The improved reinforcement learning algorithm framework is significantly better than the previous algorithm framework and has been used to predict heart failure. The experimental results have proven the effectiveness of the algorithm and its accuracy in prediction of heart failure.

Based on the theory of multi-rigid body systems, a dynamic analysis model is established for human bones. For each joint in the human skeletal model, corresponding to the rotation joint, universal joint, and ball joint in the rigid body system, they are characterized to have single, double, and three degrees of freedom, respectively. Furthermore, the rotational changes between each joint are represented by the directional cosine matrix. The relationship is given between displacement and the bone center of mass. During the experiment, figure skaters were selected as the test subjects and waltz jumps were used as the test movements. Based on the human skeletal dynamics model, the angle change curves of the left hip joint, knee joint, and ankle joint were obtained, and a comparison with the standard action was formed. The conclusion shows that the proposed human body dynamics analysis method can assist in the scientific training of figure skating.

Background: To assess the effectiveness of moxibustion and fire needle acupuncture for the treatment of gouty arthritis (GA) by conducting a systematic review and meta-analysis of randomized controlled trials (RCTs). Methods: Seven electronic databases were searched from inception to August 2022 for all RCTs of moxibustion or fire needle acupuncture for GA treatment. Two investigators evaluated the quality of the included studies and extracted data independently. All statistical analyses were performed by RevMan 5.3. Results: Ten RCTs with 653 patients were included after the final screening. The meta-analysis showed that compared with Western medicine, both moxibustion and fire needle acupuncture significantly improved the total effective rate ($\mathrm{\text{RR}}=1.16$, 95% CI 1.10–1.23, $P<0.00001$) and the Comprehensive Joint Swelling and Pain (CSJSP) ($\mathrm{\text{RR}}=3.62$, 95% CI 2.99–4.25, $P<0.00001$) and significantly decreased the uric acid level ($\mathrm{\text{RR}}=19.34$, 95% CI 2.07–36.62, $P=0.03$). Subgroup analyses based on the intervention type showed that either moxibustion or fire needle acupuncture significantly improved the total effective rate and the CSJSP and reduced uric acid levels; however, moxibustion ($\mathrm{\text{MD}}=27.70$, 95% CI 19.85–35.54, $P<0.00001$) was significantly better than fire-acupuncture ($\mathrm{\text{MD}}=-1.05$, 95% CI −14.39–12.29, $P=0.88$) in lowering uric acid levels. Conclusion: This study showed that both moxibustion and fire needle acupuncture are effective for GA patients. Considering the limitations of the included studies such as language bias and small sample size, rigorously designed and high-quality RCTs should be conducted to further evaluate the effectiveness and safety of moxibustion and fire needle acupuncture in treating GA.

Electroencephalogram (EEG) is a pivotal diagnostic tool for epilepsy. EEG signals have complex sources, significant individual differences, and data distribution varies across different domains are also easily influenced by noise. This study develops a low-rank sparse representation-based transition subspace learning (LRSRTS) algorithm. Based on low-rank reconstruction transfer subspace learning, the LRSRTS algorithm uses subspace projection and low-rank constraints to ensure that multiple domains are aligned in the subspace and uses sparse constraints to ensure reconstruction from the source domain to the target domain. Separating the noise information in the subspace, the LRSRTS algorithm uses the source domain samples to establish a discriminative classifier. Since the label information is non-negative, the LRSRTS imposes non-negative constraints on the transition subspace. With the more flexible subspace and classifier, the LRSRTS algorithm demonstrates satisfactory recognition performance on the CHB-MIT dataset.

Stroke is a group of diseases that damage brain tissue caused by sudden local cerebral blood circulation disorder. If not diagnosed and treated in time, the patient will lose a large number of neurons and even die. Among the examination methods, using magnetic resonance scans is more accurate than CT images, but it also has other risk factors. In response to these existing problems, this paper attempts to use a stroke classification model based on pre-trained CNN networks, which can help doctors detect MRI images of ischemic stroke and reduce time delays. This paper uses two real datasets. Dataset 1 includes 1112 cases of ischemic stroke and 1202 normal MRI images, while dataset 2 includes 1008 cases of ischemic stroke and 1002 normal MRI images. After data processing and noise reduction, two processed datasets, which contain 4448 cases, 4808 cases and 4032 cases, 4008 cases of ischemic stroke and normal MRI images, respectively, were obtained. Features are obtained by extracting the last pooling layer and the last fully connected layer of the pre-trained CNN network ResNet50 and DenseNet201. After combining the features, the iterative mRMR feature selection method is used to select image features. A support vector machine classifier with 10-fold cross-validation is used to classify the data. Finally, IMV is applied to the prediction vectors to get the voting results, and the best results are selected from the results. The evaluation indicators of the final experimental results in dataset 1 yielded 94.74%, 93.56%, 96.22%, and 94.87% results for accuracy, recall, precision, and F1 score, respectively. The evaluation indicators in dataset 2 yielded 92.24%, 92.21%, 92.26%, and 92.23% results for accuracy, recall, precision, and F1 score, respectively. These results clearly demonstrate that ischemic stroke can be successfully classified based on MRI and confirm the success of the hybrid feature engineering approach.