Computer Methods in Biomechanics and Biomedical Engineering最新文献

We propose a geometric framework for robust ECG classification by combining phase space reconstruction (PSR) and the distance-to-measure (DTM) function to capture intrinsic cardiac dynamics. One-dimensional ECG signals are embedded into high-dimensional phase spaces, forming point clouds that reflect rhythm characteristics. Using median-of-means kernel density estimation, we identify high-density regions and construct DTM-based pseudo-distances, enhancing noise resilience and discriminative power. Evaluated on MIT-BIH and PTB datasets, our method achieves 89.08% accuracy with logistic regression and 91.42% with Gradient Boosting Trees under Gaussian noise. DTM features complement traditional statistical ones, demonstrating strong potential for nonlinear, interpretable, and noise-robust ECG analysis.

Introduction: Cisplatin resistance remains a major cause of treatment failure in advanced bladder cancer. Moreover, growing evidence implicates ferroptosis in the development of this resistance.

Methods: We analyzed transcriptomic data from 44 cisplatin-treated bladder cancer patients to identify cisplatin-ferroptosis-related genes (CFRGs). Using machine learning and Cox regression, we developed a model and profiled the tumor microenvironment.

Results: The three gene signature stratified patients into high-risk and low-risk groups. SLC1A4 was markedly upregulated in tumors.

Discussion: Our data suggest SLC1A4, TXNIP, and PLIN4 are involved in ferroptosis mediated cisplatin resistance in BLCA, findings which merit further study.

Combining perspectives of infectious disease dynamics and evolutionary adaptive dynamics, we reveal the evolving patterns of SARS-CoV-2 virulence with the use of data-driven analysis from 22 January 2020 to 21 October 2022 and focus primarily on the Wild-type virus, D614G, Alpha, Delta, Omicron and XBB. We discuss conditions between transmission rate ($\beta \left(x\right),\beta \left(y\right)$) and the sum of natural death rate ($\mu$), mortality rate due to disease ($\alpha \left(x\right)\text{,}$$\alpha \left(y\right)$), and recovery rate ($\gamma$) for virus substitution and coexistence. (i) When $\frac{\beta \left(x\right)}{\beta \left(y\right)}<\frac{\mu +\alpha \left(x\right)+\gamma }{\mu +\alpha \left(y\right)+\gamma }\text{,}$ mutant y replaces the resident virus x; (ii) when $\frac{\beta \left(x\right)}{\beta \left(y\right)}=\frac{\mu +\alpha \left(x\right)+\gamma }{\mu +\alpha \left(y\right)+\gamma }\text{,}$ the mutant coexists with the resident virus and they are homogeneous virus. Further we adopt a segmented classification method to find that the transmission rate is gradually increasing and the mortality rate due to disease shows a significant increase at first, with the increasing virulence of the virus, there is a trend of gradual decline in the later period of the epidemic. Additionally, it turns out that reducing personnel mobility is conducive to the retention of virulent viruses and the symptoms of the patients tend to abate. To investigate the impact of dynamic changes in infectious disease systems links to SARS-CoV-2 on the evolution of its virulence over time scales, the coupling nesting of evolutionary dynamics and transmission dynamics is carried out. It suggests the results of the adaptive evolution of the virus have been verified and the increased speed of evolutionary adaptation shorts the time for viruses to peak.

To address longstanding challenges in evaluating hip protectors and compliant flooring for preventing hip fractures in the elderly, test systems should replicate hip loading and accurately assess their performance. This study evaluated a thigh impact test system (drop tower) against human body model (HBM) simulations of a sideways fall. A pelvis spring-damper model representing pelvic compliance was incorporated into the system. The test system with the pelvic model reproduced HBM-derived femoral neck forces, whereas systems without it overestimated them. These results highlight the importance of incorporating a pelvic model to improve the biofidelity of thigh impact test systems.

This study proposes a GRU-TCN model with Temporal-Channel Attention (GT-TCA) for dose-time-concentration prediction under data scarcity and multicollinearity. TimeCVAE augments limited pharmacokinetic data with distribution-consistent sequences. GRU captures temporal dependencies, TCN extracts multi-scale features, and attention emphasizes informative time steps and analytes. Experiments on Buyang Huanwu Decoction (normal/inflammatory) and simulations (RG1678, RIF) show GT-TCA reduces MAE by 22.7% and improves R² by 4% versus baselines (p < 0.05). Ablation confirms attention lowers MAE and RMSE by 6% and 5%. The model demonstrates robustness and provides more precise quantitative evidence to support precision dosing.

Accurately predicting the future cancer status of prostate cancer patients is critical for treatment. Studies show a strong link between prostate cancer and genetic mutations. To better predict a patient's cancer status and identify key mutated genes during metastasis, we propose a convolutional network with parallel structure (CNPS). Our approach includes a mutation data preprocessing method for easier feature extraction, followed by parallel convolutional networks to capture gene mutation features across multiple dimensions for more accurate predictions. Finally, CNPS is highly interpretable, allowing us to identify key genes involved in metastatic prostate cancer. After training, CNPS achieves higher accuracy on both the MPC and MSK-MET datasets.

This study aimed to simplify swimming simulation workflow using motion capture data and evaluate the resemblance of fluid force estimated from simulation and pressure distribution. Marker coordinate was collected during swimming using a motion capture system, while small pressure sensors attached to the hand and foot measured pressure distribution. The simulation used the coordinate data to reproduce relative swimming motions, with absolute angles of body segments input to calculate fluid force. The simplified workflow reduced the number of required inputs and demonstrated close resemblance in the temporal changes of hand and foot fluid force between actual swimming and the simulation.

Ribosome biogenesis (RB) is crucial for cell proliferation, but the role of RB-related genes (RBRGs) in gastric cancer remains unclear. Using TCGA and GEO data, we identified differentially expressed RBRGs and constructed a robust 7-gene prognostic model. High-risk patients exhibited activated oncogenic pathways, suppressed cell cycle/DNA repair, enriched M2 macrophages and Tregs, and increased sensitivity to various antitumor drugs. This RBRG-based model effectively predicts gastric cancer prognosis and reveals subtype-specific biological and immune characteristics.

This study investigated whether modes of variance in trial-to-trial whole-body kinematic variability identified by principal component analysis (PCA) were consistent across data pre-processing conditions generated from a common dataset. Comparisons made included 1) when trajectory data was expressed in a global vs. local reference frame; 2) when the number of landmarks used to represent whole-body motion differed, and; 3) whether input trajectory data were normalized to participant stature. Varying data pre-processing conditions prior to PCA does not bias the total variance identified. However, it can influence how modes of variance are dispersed across PCs, which in turn, can influence interpretation.

Automated and early detection of diabetics with polyneuropathy in an ambulatory health monitoring setup may reduce the major risk factors for diabetic patients. Increased and localized plantar pressure associated with impaired pain and temperature is a combination of developing foot ulcers in subjects with polyneuropathy. Although many interesting research works have been reported in this area, most of them emphasize on signal acquisition process and plantar pressure distribution in the foot region. In this work, a machine learning assisted low complexity technique was developed using plantar pressure and temperature signals which will classify between diabetic polyneuropathy and healthy subjects. Principal component analysis (PCA) and maximum relevance minimum redundancy (mRMR) methods were used for feature extraction and selection respectively followed by k-NN classifier for binary classification. The proposed technique was evaluated with 100 min of publicly available annotated data from 43 subjects and provides blind test accuracy, sensitivity, precision, F1-score, and area under curve (AUC) of 99.58%, 99.50%, 99.44%, 99.47% and 99.56% respectively. A low resource hardware implementation in ARM v6 controller required an average memory usage of 81.2 kB and latency of 1.31 s to process 9 s pressure and temperature data collected from 16 sensor channels for each of the foot region.