Computer Methods in Biomechanics and Biomedical Engineering最新文献

Estimation of mental workload from electroencephalogram (EEG) signals aims to accurately measure the cognitive demands placed on an individual during multitasking mental activities. By analyzing the brain activity of the subject, we can determine the level of mental effort required to perform a task and optimize the workload to prevent cognitive overload or underload. This information can be used to enhance performance and productivity in various fields such as healthcare, education, and aviation. In this paper, we propose a method that uses EEG and deep neural networks to estimate the mental workload of human subjects during multitasking mental activities. Notably, our proposed method employs subject-independent classification. We use the "STEW" dataset, which consists of two tasks, namely "No task" and "simultaneous capacity (SIMKAP)-based multitasking activity". We estimate the different workload levels of two tasks using a composite framework consisting of brain connectivity and deep neural networks. After the initial preprocessing of EEG signals, an analysis of the relationships between the 14 EEG channels is conducted to evaluate effective brain connectivity. This assessment illustrates the information flow between various brain regions, utilizing the direct Directed Transfer Function (dDTF) method. Then, we propose a deep hybrid model based on pre-trained Convolutional Neural Networks (CNN) and Long Short-Term Memory (LSTM) for the classification of workload levels. The accuracy of the proposed deep model achieved 83.12% according to the subject-independent leave-subject-out (LSO) approach. The pre-trained CNN + LSTM approaches to EEG data have been found to be an accurate method for assessing the mental workload.

OpenSim Moco enables solving for an optimal motion using Predictive and Tracking simulations. However, Predictive simulations are computationally prohibitive, and the efficacy of Tracking in deviating from its reference is unclear. This study compares Tracking and Predictive approaches applied to the generation of morphology-specific motion in statistically-derived musculoskeletal shoulder models. The signal analysis software, CORA, determined mean correlation ratings between Tracking and Predictive solutions of 0.91 ± 0.06 and 0.91 ± 0.07 for lateral and forward-reaching tasks. Additionally, Tracking provided computational speed-up of 6-8 times. Therefore, Tracking is an efficient approach that yields results equivalent to Predictive, facilitating future large-scale modelling studies.

This study aimed to elucidate the vertebral bone density variations associated with adolescent idiopathic scoliosis (AIS), specifically examining the impact of unilateral muscle paralysis using an integrated approach combining Frost's Mechanostat theory, a three-dimensional subject-specific finite element model and a musculoskeletal model of the L2 vertebra. The findings revealed a spectrum of bone density values ranging from 0.29 to 0.31 g/cm3, along with vertebral micro-strain levels spanning from 300 to 2200, consistent with existing literature. Furthermore, the ratio of maximum von Mises stress between the concave and convex side in the AIS model with intact muscles was approximately 1.08, which decreased by 4% due following unilateral paralysis of longissimus thoracis pars thoracic muscle. Overall, this investigation contributes to a deeper understanding of AIS biomechanics and lays the groundwork for future research endeavors aimed at optimizing clinical management approaches for individuals with this condition.

Although the safety performance of guardrail end terminals is tested using crash tests in the U.S., occupant injury risks are evaluated based on the flail-space model. This approach developed in the early 1980s neglects the influence of safety features (e.g. seatbelt, airbags, etc.) installed in late model vehicles. In this study, a vehicle (sedan, 1100 kg), a guardrail end terminal (ET-Plus) and a human body model (Global Human Body Model Consortium, GHBMC) were integrated to simulate car-to-end terminal crashes. Five velocities, two offsets, and two angles were used as pre-impact conditions. In all the 20 simulations, kinematics and kinetic data were recorded in GHBMC and vehicle models to calculate the GHBMC injury probabilities and vehicle-based injury metrics, correspondingly. Pre-impact velocity was observed to have the largest effect on the occupant injury measures. All the body-region and full-body injury risks increased with the increasing velocity. Meanwhile, the angles had a larger effect than offset to the change of full-body injury risk (9.1% vs. 0.3%). All the vehicle-based metrics had good correlations to full-body injury probabilities. Occupant Impact Velocity (OIVx), Acceleration Severity Index (ASI), and Theoretical Head Impact Velocity (THIV) had a good correlation to chest, thigh, upper tibia, and lower tibia injuries. All the other correlations (e.g. brain/head injuries) were not statistically significant. The results pointed out that more vehicle-based metrics (ASI and THIV) could help improve the predictability in terms of occupant injury risks in the tests. Numerical methodology could be used to assess head and brain injury probabilities, which were not predictable by any vehicle-based metrics.

In this work, we investigate the effects of stent indentation on hemodynamic indicators in stented coronary arteries. Our aim is to assess in-silico risk factors for in-stent restenosis (ISR) and thrombosis after stent implantation. The proposed model is applied to an idealized artery with Xience V stent for four indentation percentages and three mesh refinements. We analyze the patterns of hemodynamic indicators arising from different stent indentations and propose an analysis of time-averaged WSS (TAWSS), topological shear variation index (TSVI), oscillatory shear index (OSI), and relative residence time (RRT). We observe that higher indentations display higher frequency of critically low TAWSS, high TSVI, and non-physiological OSI and RRT. Furthermore, an appropriate mesh refinement is needed for accurate representation of hemodynamics in the stent vicinity. The results suggest that disturbed hemodynamics could play a role in the correlation between high indentation and ISR.

Prediction of bone fracture risk is clinically challenging. Computational modeling plays a vital role in understanding bone structure and diagnosing bone diseases, leading to novel therapies. The research objectives were to demonstrate the anisotropic structure of the bone at the micro-level taking into consideration the density and subject demography, such as age, gender, body mass index (BMI), height, weight, and their roles in damage accumulation. Out of 438 developed 3D bone models at the micro-level, 46.12% were female. The age distribution ranged from 23 to 95 years. The research unfolds in two phases: micro-morphological features examination and stress distribution investigation. Models were developed using Mimics 22.0 and SolidWorks. The anisotropic material properties were defined before importing into Ansys for simulation. Computational simulations further uncovered variations in maximum von-Misses stress, highlighting that young Black males experienced the highest stress at 127.852 ± 10.035 MPa, while elderly Caucasian females exhibited the least stress at 97.224 ± 14.504 MPa. Furthermore, age-related variations in stress levels for both normal and osteoporotic bone micro models were elucidated, emphasizing the intricate interplay of demographic factors in bone biomechanics. Additionally, a prediction equation for bone density incorporating demographic variables was proposed, offering a personalized modeling approach. In general, this study, which carefully examines the complexities of how bones behave at the micro-level, emphasizes the need for an enhanced approach in orthopedics. We suggest taking individual characteristics into account to make therapeutic interventions more precise and effective.

Purpose: Currently, treating femoral neck fractures (FNFs) with the inverted triangle configuration requires alignment between the femoral neck's long axis and the axis of cannulated compression screws (CCS). To address whether the 'parallel' alignment is the most effective approach for fractures with varying Pauwels angles, we employed finite element analysis (FEA) to investigate how different angles between fracture line and CCS affect stability, based on various Pauwels angles. This study aims to offer improved guidance for treating FNFs with the inverted triangle configuration.

Methods: FNF models with Pauwels angles of 40°, 50°, 60°, and 70° were developed. The CCS were positioned in an inverted triangle configuration based on the angle between the fracture line and CCS. Using FEA, we compared the biomechanical properties of each model to evaluate the stability by evaluating five key parameters: maximal stress in the proximal femoral fracture fragment (MPFS) and implants (MIS), maximal displacement of the bone (MBD) and implants (MID), and maximal relative displacement of the fragments (MRD).

Results: For Pauwels angles of 40°, 50°, 60°, and 70° across different FNF models, various parameters exhibited similar results. The MPFS showed an upward trend with a decrease in the angle, whereas the MIS, MBD, MID, and MRD all exhibited downward trends.

Conclusion: The FEA results suggest that decreasing the angle between the fracture line and the CCS for the treatment of FNF can increase the tension resistance of the model, thus increasing the model's stability.

The classification of inter-patient ECG data for arrhythmia detection using electrocardiogram (ECG) signals presents a significant challenge. Despite the recent surge in deep learning approaches, there remains a noticeable gap in the performance of inter-patient ECG classification. In this study, we introduce an innovative approach for ECG classification in arrhythmia detection by employing a 1D convolutional neural network (CNN) to leverage both morphological and temporal characteristics of cardiac cycles. Through the utilization of 1D-CNN layers, we automatically capture the morphological attributes of ECG data, allowing us to represent the shape of the ECG waveform around the R peaks. Additionally, we incorporate four RR interval features to provide temporal context, and we explore the potential application of entropy rate as a feature extraction technique for ECG signal classification. Consequently, the classification layers benefit from the combination of both temporal and learned features, leading to the achievement of the final arrhythmia classification. We validate our approach using the MIT-BIH arrhythmia dataset, employing both intra-patient and inter-patient paradigms for model training and testing. The model's generalization ability is assessed by evaluating it on the INCART dataset. The model attains average accuracy rates of 99.13% and 99.17% for 2-fold and 5-fold cross-validation, respectively, in intra-patient classification with five classes. In inter-patient classification with three and five classes, the model achieves average accuracies of 98.73% and 97.91%, respectively. For the INCART dataset, the model achieves an average accuracy of 98.20% for three classes. The experimental outcomes demonstrate the superiority of the proposed model compared to state-of-the-art models in recognizing arrhythmias. Thus, the proposed model exhibits enhanced generalization and the potential to serve as an effective solution for recognizing arrhythmias in real-world datasets characterized by class imbalances in practical applications.

The S100 family proteins (S100s) participate in multiple stages of tumorigenesis and are considered to have potential value as biomarkers for detecting and predicting various cancers. But the role of S100s in lung adenocarcinoma (LUAD) prognosis is elusive. Transcriptional data of LUAD patients were retrieved from TCGA, and relevant literature was extensively reviewed to collect S100 genes. Differential gene expression analysis was performed on the LUAD data, followed by intersection analysis between the differentially expressed genes (DEGs) and S100 genes. Unsupervised consensus clustering analysis identified two clusters. Significant variations in overall survival between the two clusters were shown by Kaplan-Meier analysis. DEGs between the two clusters were analyzed using Lasso regression and univariate/multivariate Cox regression analysis, leading to construction of an 11-gene prognostic signature. The signature exhibited stable and accurate predictive capability in TCGA and GEO datasets. Subsequently, we observed distinct immune cell infiltration, immunotherapy response, and tumor mutation characteristics in high and low-risk groups. Finally, small molecular compounds targeting prognostic genes were screened using CellMiner database, and molecular docking confirmed the binding of AMG-176, Estramustine, and TAK-632 with prognostic genes. In conclusion, we generated a prognostic signature with robust and reliable predictive ability, which may provide guidance for prognosis and treatment of LUAD.

Total Knee Replacement (TKR) effectively improves mobility and reduces pain in patients with severe arthritis. Polyethylene wear limits implant longevity by inducing osteolysis and aseptic loosening. Therefore, this study evaluates the lubrication performance of four clinically relevant UHMWPE formulations representing distinct combination of crosslinking dosage and antioxidant. An ellipsoidal on-plane model of an artificial knee joint is considered and all governing equations are solved using Newton Raphson method. The results indicate that, vitamin E blended UHMWPE without cross linking exhibits highest contact pressure with only a marginal reduction in film thickness, while achieving the lowest coefficient of friction under EHL conditions.