Medical & Biological Engineering & Computing最新文献

Surgical complications pose significant risks to patient safety and impose financial burdens, underscoring the need for reliable surgical skill training. Effective skill training requires accurate assessment. Conventional assessment methods are often subjective and labor-intensive. While motion metrics evaluate surgical performance, they provide limited insight into physiological mechanisms. This study assessed surgical proficiency through electromyography (EMG) during simulated laparoscopic tasks. Eighteen participants were recruited: five experts, five intermediates, and eight novices. EMG signals were recorded from Biceps Brachii, Triceps Brachii, Brachioradialis, Wrist Flexors, and Wrist Extensors of both arms. Root mean squared (RMS) values assessed muscle activity amplitude, mutual information (MI) quantified bimanual coordination, and instantaneous median frequency (IMDF) evaluated fatigue susceptibility. Higher skill levels, compared to lower levels, had significantly lower RMS EMG values in Biceps and Triceps, suggesting more relaxed muscle states. They exhibited significantly higher MI values, indicating superior bimanual coordination. Novices showed a significant decline in mean IMDF over trials, highlighting fatigue susceptibility, particularly in the Biceps and Triceps. These findings underscore EMG metrics' merit in objectively assessing surgical skill, providing insight into motor control, coordination, and fatigue. This multilevel physiological approach can inform training strategies and ergonomic interventions to improve surgical performance and reduce fatigue risk.

Synthetic data generation plays a crucial role in medical research by mitigating privacy concerns and enabling large-scale patient data analysis. This study presents a Graph Convolutional Neural Network combined with a Beta-Variational Autoencoder (GCN-β-VAE) framework for generating synthetic Abdominal Aortic Aneurysms (AAA). Using a small real-world dataset, our approach extracts key anatomical features and captures complex statistical relationships within a compact disentangled latent space. To address data limitations, low-impact data augmentation based on Procrustes analysis was employed, preserving anatomical integrity. The generation strategies, both deterministic and stochastic, manage to enhance data diversity while ensuring realism. Compared to PCA-based approaches, our model performs more robustly on unseen data by capturing complex, nonlinear anatomical variations. This enables more comprehensive clinical and statistical analyses than the original dataset alone. The resulting synthetic AAA dataset preserves patient privacy while providing a scalable foundation for medical research, device testing, and computational modeling.

Cerebrovascular regulation, driven by mechanisms such as cerebral autoregulation and the Cushing's reflex, plays a critical role in maintaining cerebral blood flow (CBF) adequate despite changes in arterial pressure (AP), since a dampening of CBF can lead to serious brain pathologies. This study investigates the causal and self-predictable dynamics of cerebrovascular interactions in patients undergoing coronary artery bypass graft surgery, before and after propofol general anaesthesia. The dynamics of the pressure-to-flow and flow-to-pressure links between mean arterial pressure (MAP) and mean cerebral blood velocity (MCBv) is assessed using time-domain and frequency-domain measures of Granger Causality (GC) and Granger Autonomy (GA). The results indicate that while time-domain indices remain stable, frequency-domain measures reveal variations in the very-low-frequency, low-frequency, and high-frequency (HF) bands. The increased spectral GC in the HF band may be related to the effect of mechanical ventilation during anaesthesia. Additionally, a reduction in self-dependency of MCBv in the HF band reflects weakened internal regulatory mechanisms post-anaesthesia. In conclusion, propofol-induced suppression of sympathetic control and the effects of mechanical respiration increase the dependence of cerebral blood flow on arterial pressure in specific bands of cerebrovascular interest. These findings underscore the importance of frequency-domain analysis in detecting subtle cerebrovascular dynamics that time-domain measures may overlook.

Individuals with upper limb dysfunction face significant challenges in performing everyday tasks, often depending on healthcare professionals, caregivers, or family members. Such reliance places a continuous burden on helpers who must remain available for assistance. To address these challenges, this study investigated a virtual hybrid brain-computer interface (BCI) system that integrates gaze tracking with motor imagery (MI) to control a robotic arm, potentially reducing the dependency on human support. Twenty healthy, right-handed participants took part in a virtual game environment where they controlled a robotic arm using both gaze tracking and MI. During an initial training phase, participants' electroencephalography (EEG) signals were recorded with an EEG cap. These signals were then processed and classified using the common spatial pattern (CSP) algorithm and linear discriminant analysis (LDA). In parallel, a webcam was used for real-time gaze calibration to enable accurate target selection. In the subsequent testing phase, MI commands directed the virtual robot toward predetermined targets in a Unity-based game. Training accuracy consistently outperformed online testing accuracy. The MI signal classification achieved a true positive (TP) rate of approximately 75.5%, while a significant negative correlation (r = - 0.45) was observed between MI classification accuracy and game completion times, suggesting that higher MI accuracy led to quicker task execution. These findings demonstrate the potential of combining gaze tracking with MI-based BCI for robotic control as an assistive technology for upper limb impairments. Despite its promise, technical limitations indicate that further improvements are needed to enhance system robustness, practicality, and usability for everyday activities.

Long-term ECG monitoring is crucial for detecting asymptomatic or intermittent myocardial ischemia, as it mitigates irreversible cardiac damage and prevents disease progression. Myocardial ischemia appears on ECG as transient ST-segment level and morphology alterations, known as ischemic ST change events (ISE). However, automatically identifying ISE based on ECG signals is challenging, as its recognition is highly susceptible to interference from non-ischemic ST change events, including heart rate-related ST change events (HRE), axis shift events (ASE), and conduction change events (CCE). To address this challenge, this study proposes ISENet, a lightweight deep learning-based neural network for ISE detection. The model was trained and evaluated using ECG signals and annotations from the PhysioNet long-term ST database, with tenfold cross-validation to ensure robustness and generalizability. Experimental results show that ISENet achieves an average ISE detection accuracy of 83.5%, surpassing benchmark models like VGG19 and ResNet50 while significantly reducing model complexity. This study is the first to apply a deep learning-based neural network for ISE detection using ECG signals from the long-term ST database. Compared to previous feature-engineering and feature-learning approaches, ISENet addresses key limitations in experimental design and methodology, representing a significant advancement in automated myocardial ischemia detection.

Arterial stiffness, a key cardiovascular risk marker, is typically assessed via carotid-femoral pulse wave velocity (cf-PWV), the gold-standard method. In this study, we introduce CAPE (Continuous Automatic PWV Estimation), an innovative framework for near real-time cf-PWV estimation based on beat-to-beat analysis of laser-Doppler vibrometry (LDV) signals. CAPE integrates automatic fiducial point detection, systematic signal quality control, and a cross-channel strategy to provide a highly reliable assessment of cf-PWV. The framework was evaluated using LDV signals acquired from 100 patients with mild to moderate essential hypertension, using a multichannel laser vibrometry system. CAPE calculates cf-PWV as the ratio of carotid-femoral distance to pulse transit time (PTT), which is the delay between carotid and femoral fiducial points. These points are detected using template-matching on the second derivative of LDV displacement signals. Signal quality in CAPE is ensured through an integrated quality assessment based on the number of automatically detected carotid-femoral peaks, which assigns confidence scores (acceptable or excellent) to the PWV measurements. When validated against the gold-standard applanation tonometry, CAPE achieved a mean bias of 0.25 ± 0.77 m/s, demonstrating high reliability and precision. The optimized framework estimates cf-PWV in 3 s, making CAPE ideal for clinical applications requiring real-time cardiovascular assessment.

Antibodies are a key therapeutic class in pharma, enabling precise targeting of disease agents. Traditional methods for their design are slow, costly, and limited. Advances in high-throughput data and artificial intelligence (AI) including machine learning, deep learning, and reinforcement learning have revolutionized antibody sequence design, 3D structure prediction, and optimization of affinity and specificity. Computational approaches enable rapid library generation and efficient screening, reduce experimental sampling, and support rational design with improved immune response. Combining AI with experimental methods allows for de novo, multifunctional antibody development. AI also accelerates the discovery process, target identification, and candidate prioritization by analyzing large datasets, predicting interactions, and guiding modifications to enhance efficacy and safety. Despite challenges, ongoing research continues to expand the potential of AI and transform antibody development and the pharmaceutical industry. Antibodies are a key therapeutic class in pharma, enabling precise targeting of disease agents. Traditional methods for their design are slow, costly, and limited. Advances in high-throughput data and artificial intelligence (AI) including machine learning, deep learning, and reinforcement learning have revolutionized antibody sequence design, 3D structure prediction, and optimization of affinity and specificity. Computational approaches enable rapid library generation and efficient screening, reduce experimental sampling, and support rational design with improved immune response. Combining AI with experimental methods allows for de novo, multifunctional antibody development. AI also accelerates the discovery process, target identification, and candidate prioritization by analyzing large datasets, predicting interactions, and guiding modifications to enhance efficacy and safety. Despite challenges, ongoing research continues to expand the potential of AI and transform antibody development and the pharmaceutical industry.