Radiocapitellar joint pressures following transradial amputation increase during elbow motion

Is implant choice associated with fixation strength for displaced radial neck fracture: a network meta-analysis of biomechanical studies

The multitude of fixation options for radial neck fractures, such as pins, screws, biodegradable pins and screws, locking plates, and blade plates, has led to a lack of consensus on the optimal implant choice and associated biomechanical properties. This study aims to evaluate the biomechanical strength of various fixation constructs in axial, sagittal, and torsional loading directions. We included biomechanical studies comparing different interventions, such as cross/parallel screws, nonlocking plates with or without augmented screws, fixed angle devices (T or anatomic locking plates or blade plates), and cross pins. A systematic search of MEDLINE (Ovid), Embase, Scopus, and CINAHL EBSCO databases was conducted on September 26th, 2022. Data extraction was carried out by one author and verified by another. A network meta-analysis (NMA) was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines. Primary outcomes encompassed axial, bending, and torsional stiffness, while the secondary outcome was bending load to failure. Effect sizes were calculated for continuous outcomes, and relative treatment ranking was measured using the surface under the cumulative ranking curve (SUCRA). Our analysis encompassed eight studies, incorporating 172 specimens. The findings indicated that fixed angle constructs, specifically the anatomic locking plate, demonstrated superior axial stiffness (mean difference [MD]: 23.59 N/mm; 95% CI 8.12-39.06) in comparison to the cross screw. Additionally, the blade plate construct excelled in bending stiffness (MD: 32.37 N/mm; 95% CI - 47.37 to 112.11) relative to the cross screw construct, while the cross-screw construct proved to be the most robust in terms of bending load failure. The parallel screw construct performed optimally in torsional stiffness (MD: 139.39 Nm/degree; 95% CI 0.79-277.98) when compared to the cross screw construct. Lastly, the nonlocking plate, locking T plate, and cross-pin constructs were found to be inferior in most respects to alternative interventions. The NMA indicated that fixed angle devices (blade plate and anatomic locking plate) and screw fixations may exhibit enhanced biomechanical strength in axial and bending directions, whereas cross screws demonstrated reduced torsional stability in comparison to parallel screws. It is imperative for clinicians to consider the application of these findings in constraining forces across various directions during early range of motion exercises, taking into account the distinct biomechanical properties of the respective implants.

Biosignal-based transferable attention Bi-ConvGRU deep network for hand-gesture recognition towards online upper-limb prosthesis control

BACKGROUND AND OBJECTIVE

Upper-limb amputation can significantly affect a person's capabilities with a dramatic impact on their quality of life. As a biological signal, surface electromyogram (sEMG) provides a non-invasive means to measure underlying muscle activation patterns, corresponding to specific hand gestures. This project aims to develop a real-time deep learning based recognition model to automatically and reliably recognise these complex signals of a wide range of daily hand gestures from amputees and non-amputees.

METHODS

This paper proposes an attention bidirectional Convolutional Gated Recurrent Unit (Bi-ConvGRU) deep neural network for hand-gesture recognition. By training on sEMG data from both amputees and non-amputees, the model can learn to recognise a group of fine-grained hand movements. This is a significantly more challenging and underexplored area, compared to existing studies on coarse-control in lower limbs. One dimensional CNNs are initially used to extract intra-channel features. The novel use of a bidirectional sequential GRU (Bi-GRU) deep neural network allows the exploration of correlation of muscle activation among multi-channel sEMG signals from both prior and posterior time sequences. Importantly, the attention mechanism is employed following Bi-GRU layers. This enables the model to learn vital parts and feature weights, increasing robustness to bio-data noise and irregularity. Finally, we introduce the first of its kind transfer learning, demonstrating that a baseline model pre-trained with non-amputee data can be effectively refined with amputee data to build a personalised model for amputees.

RESULTS

The attention Bi-ConvGRU was evaluated on the benchmark database Ninapro, and achieved an average accuracy of 88.7%, outperforming the state-of-the-art on 18 gesture recognition by 6.7%.

CONCLUSIONS

To our knowledge, the developed end-to-end deep learning model is the first of its kind that enables reliable predictive decision making in short time windows (160ms). This reduced latency limits physiological awareness, enabling the potential for real-time, online and thus more intuitive bio-control of prosthetic devices for amputees.