Magnetic tracking has emerged as a promising solution for biomedical applications, as it can estimate the position and orientation of magnetic sources inside the human body without a line of sight. In prosthetic control, we introduced the myokinetic control interface, deriving control signals through real-time localization of small permanent magnets implanted in the residual muscles of amputees. The quality of these signals strongly depends on localization accuracy, which previous studies demonstrated to be substantially affected by geometric constraints. While factors such as magnet-to-sensor distance and sensor arrangement have been extensively investigated, the lateral boundaries of the optimal localization workspace remain poorly understood. To fill this gap, this study systematically investigated how the distance between magnets and the edges of the sensor-area affects localization accuracy. The latter was quantified by computing the localization error of a magnet pair as its position varied within and beyond the sensor-area boundaries. Experimental results showed that localization remained accurate while magnets stayed within the sensor-area, but degraded rapidly once they moved beyond its lateral boundaries. The optimal localization workspace progressively narrowed as the sensor-to-magnet distance increased, defining a pyramidal region ensuring reliable localization accuracy. Similar results emerged in simulations only after modelling sensor nonidealities. To generalize these findings, we introduced two geometry-related parameters describing magnet position relative to the sensor-area, explaining approximately 80% of the variability in localization accuracy. These results provide insights into the complexity of modelling sensor behaviour using simulated data, and practical guidelines for defining the optimal workspace in multi-magnet localization systems.
Background and objective Ultrasound image quality depends on acquisition conditions, while conventional full-reference image quality assessment (FR-IQA) metrics may not fully capture perceptually relevant intensity and texture changes. Radiomics has been explored for image-quality characterization in other medical modalities. This study developed a radiomics-based machine-learning framework for FR-IQA under controlled acquisition-setting variations. Methods Rabbit cadaver ultrasound data included two anatomies and two probe-scanner configurations, defining four reference groups. Varying frequency, gain, power, and dynamic range yielded 192 reference-test pairs. Four sonographers rated pairs. B-mode and wavelet radiomic descriptors formed six full-reference representations evaluated by reference-grouped nested model development. Performance was compared with eight classical FR-IQA methods and two learned perceptual metrics, with robustness and acquisition-setting analyses. Results Average-measure inter-observer agreement was high, with ICC(A,k) = 0.9602. Out-of-reference predictions achieved SROCC = 0.8930, PLCC = 0.8702, RMSE = 0.1462, and R2 = 0.7196. On the primary region of interest (ROI), the framework showed higher observed aggregate performance than all classical FR-IQA methods, while learned perceptual metrics showed higher observed performance. No reference-aware pairwise comparison was significant after Holm correction. Predictions were stable under modest ROI and discretization perturbations. Gain had the largest median radiomic effect and was the dominant factor for subjective quality and predictions across reference groups, enabling feature- and prediction-level assessment of acquisition-setting sensitivity. Conclusions Radiomic full-reference representations provide an interpretable approach for characterizing acquisition-dependent ultrasound image quality, with potential utility for system evaluation and protocol optimization. Larger multi-system, in vivo studies with repeated acquisitions and independently defined references are needed to establish generalizability.
Does a more unequal society emit more CO2? The nexus between carbon emissions and income inequality has been at the core of a vast literature yielding conflicting results. Here, we provide robust evidence of a non-linear relationship that depends on the structural composition of the economy. Specifically, we document a positive association between income inequality—measured using five different indicators, for both pre-tax and post-tax income—and per capita carbon emissions in highly tertiarized countries. By contrast, this relationship turns negative in non-service-intensive economies. We provide evidence for plausible mechanisms mediating this non-linear association. In particular, the investment carbon footprint of the top 1% and the size of the manufacturing sector emerge as key dimensions underlying the observed patterns. Our results point to the structural composition of the economy as a crucial factor shaping the emission-inequality nexus. As such, economic structure helps identify phases of development in which reducing inequality comes with climate-related benefits.
Self Organizing Maps (SOMs) are unsupervised, shallow, artificial neural networks, built on top of the competitive learning principle, and typically employed for clustering, dimensionality reduction and high-dimensional data visualization. This paper presents XPySom, a high-performance and parallel implementation of SOMs for Python, with a focus on acceleration of the computations through the use of multicore processors and GP-GPUs. After describing the major design and implementation choices, we present results from an extensive experimental evaluation of XPySom on the Extended MNIST openly available dataset of handwritten digits. On this dataset, the training time is shown to be up to two orders of magnitude faster than existing open-source SOM implementations, taking as little as 2 s to train the model on the 240000 images in the dataset for 100 epochs. The implementation is released under an open-source GPLv3 license, and made publicly available as the xpysom package for Python.
Regenerative peripheral nerve interfaces (RPNIs) are emerging platforms capable to translate neural activity into controllable myoelectric signals. However, current clinical RPNIs rely on muscle autografts, limiting their scalability and extensive adoption. Here, we report a bioactive, ultrasound-responsive piezoelectric muscle biomaterial designed as a fully tissue-engineered alternative to autologous grafts. The construct consists of fibrinogen-based muscle tissues enriched with barium titanate nanoparticles (BTNPs, diameter∼60 nm) and supported by a biodegradable surgical membrane that promotes the formation of aligned, multinucleated myotubes. The incorporation of BTNPs imparts intrinsic piezoelectric activity to the construct, and the nanoparticles are taken up by developing myotubes, enabling remote mechanoelectrical tissue stimulation under low-intensity pulsed ultrasound (LIPUS).In vivo, piezoelectric constructs implanted around the rat peroneal nerve for two months undergo LIPUS-driven activation of internalized BTNPs, which enhances muscle maturation, and electromechanical responsiveness, yielding myoelectrical signals up to 3.2 mV upon nerve activation. Histological analyses confirm improved structural organization, increased desmin expression, and evidence of neovascularization and axonal regeneration within the engineered interface. These findings suggest that the piezoelectric constructs form stable, functional biointerfaces with peripheral nerves, and that LIPUS-driven activation of embedded BTNPs provides a non-invasive strategy to potentiate muscle development and signal transduction.This study positions LIPUS-responsive, engineered piezoelectric muscle constructs as a donor-free, bioactive platform alternative to traditional autografts for next-generation human-machine interfaces and regenerative bioelectronics.
