Computer Methods in Biomechanics and Biomedical Engineering最新文献

This paper proposes an optimized Long Short-Term Memory (LSTM+) model for predicting cumulative confirmed cases of COVID-19 in Germany, the UK, Italy, and Japan. The LSTM+ model incorporates two key optimizations: (1) fine-adjustment of parameters and (2) a 're-prediction' process that utilizes the latest prediction results from the previous iteration. The performance of the LSTM+ model is evaluated and compared with that of Backpropagation (BP) and traditional LSTM models. The results demonstrate that the LSTM+ model significantly outperforms both BP and LSTM models, achieving a Mean Absolute Percentage Error (MAPE) of less than 0.6%. Additionally, two illustrative examples employing the LSTM+ model further validate its general applicability and practical performance for predicting cumulative confirmed COVID-19 cases.

We scrutinized the impact of doping of X atoms (X = Fe, Co, Ni, Cu, and Zn) on the metformin (MF) drug delivery performance of a BP nanotube (BPNT) using density functional B3LYP calculations. The pristine BPNT was not ideal for the drug delivery of MF because of a weak interaction between the drug and nanotube. Doping of the Zn, Cu, Ni, Co, and Fe into the BPNT surface raised the adsorption energy of MF from -5.3 to -29.1, -28.7, -29.8, -32.1, and -26.9 kcal/mol, respectively, demonstrating that the sensitiveness of the metal-doped BPNT increased after increasing the radius atomic of metals. Ultimately, there was an increase in the adhesion performance and capacity of the MF after X (especially Co atom) doping, making the nanotube suitable for MF drug delivery. The mechanism of MF reaction with the BPNT changed from covalent bonding in the natural environment to hydrogen bonding in the cancerous cells with high acidity.

In order to comprehensively evaluate the driver's vibration comfort under different vibration conditions, eighteen subjects were required to drive a tractor at different speeds on field and asphalt roads respectively in the real vehicle experiment. The sEMG signals and vibration acceleration signals of the subjects were recorded. And the time-frequency domain analysis of sEMG signals and acceleration signals were used to determine the relationship among the characteristic indexes, tractor speed and road surfaces. The relevance analysis showed that there was a significant correlation between the integral electromyography (iEMG) and median frequency (MF) of the middle scalene muscle, erector spinae muscle and gastrocnemius muscle, the RMS of weighted acceleration (a_w) of the neck, waist and legs, and the subjective comfort feelings. It was proven that the tractor speed had a significant impact on human body vibration based on the ANOVA result (p < 0.05). With the increase of running speed, the time domain indexes of sEMG signals including iEMG, RMS and the vibration acceleration signals of the testing body parts increased significantly, while the amplitudes of frequency domain indexes decreased. Therefore, a quantitative regression evaluation model for the comfort of the neck, waist and legs integrating the sEMG and vibration signals was established, and its relative errors were 5.05, 4.38 and 6.12% respectively. This proposed assessment model can combine characteristics of the partial and overall vibration response of human body effectively, predict the tractor driver's vibration comfort accurately, provide a theoretical basis for the evaluation of tractor cab vibration comfort.

Hemodynamic analysis reveals a highly significant effect on the prevention, diagnosis, and treatment of human vascular diseases. This article goes deeply into the periodic pulsatile blood flow in the carotid artery with an elastic vessel wall. In view of blood rheological experimental data, the constitutive equation of fractional Maxwell power-law fluid with yield stress, which can describe the four characteristics of yield stress, viscoelasticity, shear thinning, and thixotropy is established. Meanwhile, drawing support from the data of pulsatile flow, the finite Fourier series of pressure gradient with a period of 1 s has been proposed. Leading into Hooke's law can build the fluid-structure coupling boundary condition of blood flow and elastic vessel wall. The numerical solutions are got hold of finite difference method integrated with the newly developed L1-algorithm, and their convergence and stability of which are verified. The axial velocities of blood under different constitutive relationships are compared. The results throw light that other constitutive relationships underestimate the velocity of blood. Furthermore, the flow rate and wall shear stress on different fluid are calculated. It can be concluded that compared with Bingham fluid, the maximum and minimum flow rate/wall shear stress of fractional Maxwell power-law fluid with yield stress increases by 19% and 32%, respectively. The flow rate lags behind the pressure gradient and has time delay effect, on the contrary, the velocity of blood vessel wall is keeping pace with the pressure gradient. The effects of relevant physical parameters on velocity are discussed. In addition, the spatiotemporal distribution of blood flow in cerebral artery and femoral artery are analyzed.

To analyze the fixation strength of cannulated screws fixation in the treatment of femoral neck fracture with posterior tilt due to insufficient reduction. Two sets of digital models of anatomical reduction and 15° tilting reduction were established by CT data. Each group of models was modeled with two different fixation methods. One fixation method was fixed according to the standard cannulated screws recommended by AO. Another fixation method is to tilt the screw posterior tilt 15°. The final four groups of models were obtained: AO principle nailing posterior tilt model (Group A), posterior direction nailing posterior tilt model (Group B), AO principle nailing anatomic reduction model (Group C) and posterior direction nailing anatomic reduction model (Group D). The maximum displacement of the fracture end, the maximum Von-Mises stress and the stress distribution of the internal fixation were compared among the four groups. Four groups of models were established on artificial bone by 3D printing guide plate technology. The 600 N pressure test and yield test were performed on a biomechanical machine. The finite element and biomechanical models showed that groups B and C were more stable than groups A and D. The stability of group B was not worse than that of group C. When the femoral neck fracture produces a posterior tilt, a posterior reduction is allowed. The change of AO screw to posterior tilting screw fixation has more powerful advantages. No posterior tilt or posterior reduction, AO screw placement is still required.

The practice of a musical instrument requires fine dexterity, repetitive, fast, and precise movements, as well as important efforts to set the instrument into vibration, while adopting postures often unnatural for the human body. As a result, musicians are often subject to pain and musculoskeletal disorders. In the case of plucked string instruments and especially the concert harp, the plucking force is directly related to the strings' tension. Consequently, the choice of the strings has to be made based on both, the musician feel while playing, and the musculoskeletal consequences. This paper investigates how the string properties and the playing dynamics affect the finger and wrist muscle activity during harp playing. This study first emphasized the noteworthy recruitment of the flexor and extensor muscles (42% and 29% of MVC, respectively). Findings outlined further that the fingering choice, the adopted playing dynamics and the string's material govern the muscular activity level and the playing control. Such results are a first step to better understand how the harp ergonomics may affect the player's integrity and help them decide the most suitable stringing for their practice.

Arrhythmic heartbeat classification has gained a lot of attention to accelerate the detection of cardiovascular diseases and mitigating the potential cause of one-third of deaths worldwide. In this article, a computer-aided diagnostic (CAD) approach has been proposed for the automated identification and classification of arrhythmic heartbeats from electrocardiogram (ECG) signals using multiple features aided supervised learning model. For proper diagnosis of arrhythmic heartbeats, MIT-BIH Arrhythmia database has been used to train and test the proposed approach. The ECG signals, extracted from sensor leads, have undergone pre-processing via discrete wavelet transform. Three sets of features, i.e. statistical, temporal, and spectral, are extracted from the processed ECG signals followed by random forest aided recursive feature elimination strategy to select the prominent features for proper classification of arrhythmic heartbeats by the proposed optimal extreme gradient boosting (O-XGBoost) classifier. Hyperparameters such as learning rate, tree-specific parameters, and regularization parameters have been optimized to improve the performance of the XGBoost classifier. Moreover, the synthetic minority over-sampling technique has been employed for balancing the dataset in order to improve the classification performance. Quantitative results reveal the remarkable performance over state-of-the-art methods. The proposed model can be implemented in any computer-aided diagnostic system with similar topological structures.

This study aimed to develop an algorithm for modelling bone loss defects in a given mandibular geometry, with a user-defined depth, width, place, and defect type. The algorithm was implemented using Grasshopper and models with different bone loss types and depths around a dental implant were built. The models were used in a finite element analysis (FEA) to predict the stresses in peri-implant bone. The FEA showed that the stresses in peri-implant bone depend primarily on the depth of bone loss, whereas the type of bone loss showed no major influence.

Adverse delivery outcomes is a major re-productive health problem that affects the physical and mental health of pregnant women. Obviously, obstetric clinical data has periodically time series characteristics. This paper proposed a three stage adverse delivery outcomes prediction model via the fusion of multiple time series clinical data. The first stage is data aggregation, in which the data set is collected from the obstetric clinical data and divided based on time series features. In the second stage, a multi-channel gated cycle unit is used to solve the calculation error caused by irregular sampling of time series data. The hidden layer feature vector is connected with the fully connected layer, reshaped into a new one-dimensional feature, and fused with the non-time series data into a new data set. The third stage is the prediction stage of adverse delivery outcomes. By connecting the multi-channel gated cycle unit with the extreme gradient lift, the data transmitted in the corresponding channel is used in the feature extraction stage, in which the weighted entropy-based feature extraction is adopted. With the help of the extracted features, a hybrid artificial neural network architecture (MGRU-XGB) was developed to predict adverse delivery outcomes. The experimental results showed that the hybrid model had the best prediction performance for adverse delivery outcomes compared with other single models in terms of sensitivity, specificity, AUC and other evaluation indexes.

Diabetes mellitus is a severe condition that has the potential to impair strength. The disease known as diabetes mellitus, which is a chronic condition, is brought on by a significant rise in blood glucose levels. The diagnosis of this condition is made using a variety of chemical and physical testing. Diabetes, however, can harm the organs if it goes undetected. This study develops a hybrid deep-learning technique to recognize Type 2 diabetes mellitus. The data is cleaned up at the pre-processing stage using a data transformation technique based on the Yeo-Jhonson transformation. The tanimoto similarity is used in the feature selection process to select the best features from the data. To prepare data for future processing, data augmentation is performed. The Deep Residual Network and the Rider-based Neural Network are recommended and trained separately for the T2DM identification using the Competitive Multi-Verse Rider Optimizer. The outputs generated by the RideNN and DRN classifiers are blended using correlation-based fusion. The suggested CMVRO-based NN-DRN has shown improved performance with the highest accuracy of 91.4%, sensitivity of 94.8%, and specificity of 90.1%.