Medical & Biological Engineering & Computing最新文献

Three-dimensional models have been widely used to study knee joint biomechanics in both healthy and pathological conditions. However, the lack of data on pediatric knee models affected by a discoid lateral meniscus necessitates further investigation. This study analyzed the biomechanical behavior of a pediatric knee joint with a discoid lateral meniscus malformation and the effects of partial meniscectomy on restoring its normal configuration. The three-dimensional geometry was reconstructed from computed tomography and magnetic resonance imaging data to develop a finite element model of the pediatric knee. The finite element method was used to simulate the joint in an upright position, and contact, compressive, and shear stresses were analyzed across seven lateral meniscus configurations with varying residual tissue widths to simulate progressive degrees of partial meniscectomy. A discoid lateral meniscus altered knee biomechanics, increasing medial-compartment stress, associated with femoral cartilage damage. Under body weight loading, the pediatric model showed a significant rise in stress when the meniscal width fell below 12 mm. A residual meniscal width of 12 mm provided a more favorable biomechanical response in this pediatric knee model, potentially reducing cartilage damage and the risk of early degeneration after partial meniscectomy.

Rapid serial visual presentation (RSVP) is a promising paradigm for visual brain-computer interfaces (BCIs) based on event-related potentials (ERPs) for patients with limited muscle and eye movement. This study explores the impact of video background and stimulus transparency on BCI control, factors that have not been previously examined together under RSVP. Two experimental sessions were conducted with 12 participants each. Four BCI conditions were tested: opaque pictograms, and white background (A255W); opaque pictograms, and video background (A255V); intermediate transparent pictograms, and video background (A085); and highly transparent pictograms, and video background (A028V). The results indicated that the video background had a negative impact on BCI performance. In addition, the intermediate transparent pictograms (A085V) proved to be balanced, as it did not show significant performance differences compared to opaque pictograms (A255V) but was rated significantly better by users on subjective measures related to attending to the video background. Therefore, in applications where users must shift attention between BCI control and their surroundings, balancing stimulus transparency is a suitable option for enhancing system usability. These findings are particularly relevant for designing asynchronous ERP-BCIs using RSVP for patients with impaired oculomotor control.

Exoskeletons have exhibited increasingly diverse designs and broader applications in rehabilitation and medical fields. To achieve optimal assistance performance, it is essential that the assistive force be synchronized with human biomechanics. However, current soft exoskeletons still face challenges in achieving natural gait synchronization and providing stable, comfortable assistance. This study proposes a biomimetic assistance method for a hip soft exoskeleton that better matches natural human gait. It explores how integrating a dual‑pretension mechanism with biomechanics‑based force modeling can enhance assistive performance, improve user comfort, and reduce metabolic energy consumption during walking. By mimicking the muscle force characteristics of hip joint extension and flexion, two sets of assistive torque curves were developed to correspond with human biomechanical motion. Additionally, to compensate for the hysteresis inherent in the exoskeleton system, a pre‑tension force was applied before and after the assistive curves to improve response time. To enhance the accuracy of gait cycle prediction and achieve better synchronization with natural gait patterns, a Gaussian-weighted moving average algorithm was employed to adaptively assign higher weights to recent gait data, thereby improving the responsiveness and adaptability of the exoskeleton. In the experiments, six subjects participated, and their net metabolic rates were compared under assisted and unassisted conditions. The results showed that the subjects' average metabolic cost decreased by 16.4% at a walking speed of 3 km/h and by 14.1% on a 4° slope. Compared with previous approaches, the proposed algorithm achieved more accurate gait‑phase adaptation and reduced metabolic expenditure, highlighting its potential for human-exoskeleton co‑adaptation.

Quantification of left ventricular function is essential for diagnosing cardiovascular diseases. Current clinical practice requires interactive segmentation of ultrasound images to delineate the left ventricular region and identify keypoints such as the apex and mitral annulus, a process that is both time-consuming and inefficient. To address these limitations, we introduce MultiEchoNet, a multi-task network employing a weakly supervised learning strategy to automatically calculate the left ventricular ejection fraction (LVEF) and mitral annulus diameter (MAD). Our approach integrates a novel task propagation module designed to improve the network's ability to capture global semantic information for each task at reduced computational cost, thereby minimizing task interference and enhancing task-specific feature extraction. Furthermore, we developed a multi-task Transformer module to facilitate the extraction of complementary modality information across tasks, promoting mutual guidance and optimization. This enables concurrent left ventricular segmentation and keypoint localization. In addition, peak detection is utilized to identify the end-systolic frame and end-diastolic frame in the echocardiographic sequence generated by the network, allowing for the precise calculation of related parameters. Experimental evaluations on public datasets EchoNet-Dynamic and CAMUS demonstrate that our algorithm achieves Dice similarity coefficients of 93.51% and 93.18%, respectively, and the highest keypoint similarity scores were 0.958 and 0.940, respectively. Additionally, the correlation coefficients between the predicted and true LVEF values were 0.845 and 0.82, respectively, while those for MAD were 0.971 and 0.963, respectively. These results suggest that MultiEchoNet offers robust support for the auxiliary diagnosis of cardiovascular diseases. Code is available at https://github.com/zzzmmmlll965/MultiEchoNet .

Effective Electroencephalogram (EEG) signal processing necessitates the mitigation of physiological artifacts. While deep learning frameworks have demonstrated superior performance over traditional methods for this task, their high complexity and computational demands hinder deployment on resource-constrained platforms. In this work, denoising network called EEGPARnet is proposed to address this limitation. The proposed architecture integrates transformer encoders equipped with temporal and spectral attention modules and a Gated Recurrent Unit (GRU)-based decoder. This fusion enables the model to learn time-frequency long-range similarities, facilitating efficient feature extraction and a reduced number of trainable parameters. Experimental validation of the proposed model on the EEGDenoiseNet dataset revealed an average temporal relative root mean square error ([Formula: see text]) of 0.289, spectral relative root mean square error ([Formula: see text]) of 0.312, and a correlation coefficient (CC) of 0.942 for ocular artifact removal. For muscular artifact removal, the proposed method achieved competitive results against state-of-the-art techniques, with mean [Formula: see text], [Formula: see text], and CC values of 0.458, 0.428, and 0.855, respectively. Compared to state-of-the-art model, the proposed EEGPARnet demonstrated a significant reductions in computational complexity with [Formula: see text] fewer trainable parameters, [Formula: see text] less FLOPS, and [Formula: see text] smaller storage, making it a step closer towards deployment on resource-constrained devices for real-time EEG denoising without compromising performance.

Accurate and accessible imaging techniques are essential for evaluating muscle morphology in both clinical and research settings. This study examined the validity and reliability of freehand three-dimensional ultrasound (3DUS) employing a multiple-sweep technique for measuring the volume of complexly shaped muscles such as the gluteus medius (GMed), with magnetic resonance imaging (MRI) as the reference standard. Twelve healthy participants (21.1 ± 1.5 years) underwent both 3DUS and MRI scans. Each GMed was scanned using three overlapping 3DUS sweeps, and two processors independently segmented all 3DUS and MRI images to calculate muscle volume. Agreement between 3DUS and MRI was evaluated using Bland-Altman plots, while intra- and inter-processor reliability for 3DUS were assessed using intraclass correlation coefficients (ICCs), coefficients of variation (CV%), typical error (TE), and minimal detectable change (MDC). The mean difference between 3DUS and MRI was minimal, with Bland-Altman plots demonstrating good agreement and no systematic bias. Inter- and intra-processor reliability were excellent (ICC = 0.972 and 0.999, respectively). A subgroup analysis (n = 10) comparing prone and side-lying positions using 3DUS also demonstrated good between-position reliability (ICC = 0.94). Freehand 3DUS with the multiple-sweep technique provides a valid, reliable, and practical alternative to MRI for measuring GMed muscle volume in both clinical and research applications.