A Particle Swarm Optimised Independence Estimator for Blind Source Separation of Neurophysiological Time Series.

The decomposition of neurophysiological recordings into their constituent neural sources is of major importance to a diverse range of neuroscientific fields and neuroengineering applications. The advent of high density electrode probes and arrays has driven a major need for novel semi-automated and automated blind source separation methodologies that take advantage of the increased spatial resolution and coverage these new devices offer. Independent component analysis (ICA) offers a principled theoretical framework for such algorithms, but implementation inefficiencies often drive poor performance in practice, particularly for sparse sources. Here we observe that the use of a single non-linear optimization function to identify spiking sources with ICA often has a detrimental effect that precludes the recovery and correct separation of all spiking sources in the signal. We go on to propose a projection-pursuit ICA algorithm designed specifically for spiking sources, which uses a particle swarm methodology to adaptively traverse a polynomial family of non-linearities approximating the asymmetric cumulants of the sources. We robustly prove state-of-the-art decomposition performance on recordings from high density intramuscular probes and demonstrate how the particle swarm quickly finds optimal contrast non-linearities across a range of neurophysiological datasets.

Amplification of autoimmune organ damage by NKp46-activated ILC1.

In systemic lupus erythematosus (SLE) loss of immune tolerance, autoantibody production and immune complex deposition are required but not sufficient for organ damage1. How inflammatory signals are initiated and amplified in the setting of autoimmunity remains elusive. Here, we set out to dissect layers and hierarchies of autoimmune kidney inflammation in order to identify tissue-specific cellular hubs that amplify auto-inflammatory responses. Using high-resolution single-cell profiling of kidney immune and parenchymal cells, in combination with antibody blocking and genetic deficiency, we show that tissue-resident NKp46+ innate lymphoid cells (ILC) are crucial signal amplifiers of disease-associated macrophage expansion and epithelial cell injury in lupus nephritis, downstream of autoantibody production. NKp46 signaling in a distinct subset of ILC1 instructed an unconventional immune-regulatory transcriptional program, which included the expression of the myeloid cell growth factor CSF2. CSF2 production by NKp46+ ILC promoted the population expansion of monocyte-derived macrophages. Blockade of the NKp46 receptor (using the antibody mNCR1.152) or genetic deficiency of NKp46 abrogated epithelial cell injury. The same cellular and molecular patterns were operative in human lupus nephritis. Our data support that NKp46+ ILC1 promote parenchymal cell injury by granting monocyte-derived macrophages access to epithelial cell niches. NKp46 activation in ILC1 thus constitutes a previously unrecognized, critical tissue rheostat that amplifies organ damage in autoimmune hosts, with broad implications for inflammatory pathologies and therapies.

Conditional Generative Models for Simulation of EMG During Naturalistic Movements.

Numerical models of electromyography (EMG) signals have provided a huge contribution to our fundamental understanding of human neurophysiology and remain a central pillar of motor neuroscience and the development of human-machine interfaces. However, while modern biophysical simulations based on finite element methods (FEMs) are highly accurate, they are extremely computationally expensive and thus are generally limited to modeling static systems such as isometrically contracting limbs. As a solution to this problem, we propose to use a conditional generative model to mimic the output of an advanced numerical model. To this end, we present BioMime, a conditional generative neural network trained adversarially to generate motor unit (MU) activation potential waveforms under a wide variety of volume conductor parameters. We demonstrate the ability of such a model to predictively interpolate between a much smaller number of numerical model's outputs with a high accuracy. Consequently, the computational load is dramatically reduced, which allows the rapid simulation of EMG signals during truly dynamic and naturalistic movements.

NeuroMotion: Open-source platform with neuromechanical and deep network modules to generate surface EMG signals during voluntary movement.

Neuromechanical studies investigate how the nervous system interacts with the musculoskeletal (MSK) system to generate volitional movements. Such studies have been supported by simulation models that provide insights into variables that cannot be measured experimentally and allow a large number of conditions to be tested before the experimental analysis. However, current simulation models of electromyography (EMG), a core physiological signal in neuromechanical analyses, remain either limited in accuracy and conditions or are computationally heavy to apply. Here, we provide a computational platform to enable future work to overcome these limitations by presenting NeuroMotion, an open-source simulator that can modularly test a variety of approaches to the full-spectrum synthesis of EMG signals during voluntary movements. We demonstrate NeuroMotion using three sample modules. The first module is an upper-limb MSK model with OpenSim API to estimate the muscle fibre lengths and muscle activations during movements. The second module is BioMime, a deep neural network-based EMG generator that receives nonstationary physiological parameter inputs, like the afore-estimated muscle fibre lengths, and efficiently outputs motor unit action potentials (MUAPs). The third module is a motor unit pool model that transforms the muscle activations into discharge timings of motor units. The discharge timings are convolved with the output of BioMime to simulate EMG signals during the movement. We first show how MUAP waveforms change during different levels of physiological parameter variations and different movements. We then show that the synthetic EMG signals during two-degree-of-freedom hand and wrist movements can be used to augment experimental data for regressing joint angles. Ridge regressors trained on the synthetic dataset were directly used to predict joint angles from experimental data. In this way, NeuroMotion was able to generate full-spectrum EMG for the first use-case of human forearm electrophysiology during voluntary hand, wrist, and forearm movements. All intermediate variables are available, which allows the user to study cause-effect relationships in the complex neuromechanical system, fast iterate algorithms before collecting experimental data, and validate algorithms that estimate non-measurable parameters in experiments. We expect this modular platform will enable validation of generative EMG models, complement experimental approaches and empower neuromechanical research.

Mitochondrial control of lymphocyte homeostasis.

Mitochondria play a multitude of essential roles within mammalian cells, and understanding how they control immunity is an emerging area of study. Lymphocytes, as integral cellular components of the adaptive immune system, rely on mitochondria for their function, and mitochondria can dynamically instruct their differentiation and activation by undergoing rapid and profound remodelling. Energy homeostasis and ATP production are often considered the primary functions of mitochondria in immune cells; however, their importance extends across a spectrum of other molecular processes, including regulation of redox balance, signalling pathways, and biosynthesis. In this review, we explore the dynamic landscape of mitochondrial homeostasis in T and B cells, and discuss how mitochondrial disorders compromise adaptive immunity.

Deep Metric Learning With Locality Sensitive Mining for Self-Correcting Source Separation of Neural Spiking Signals.

Automated source separation algorithms have become a central tool in neuroengineering and neuroscience, where they are used to decompose neurophysiological signal into its constituent spiking sources. However, in noisy or highly multivariate recordings these decomposition techniques often make a large number of errors. Such mistakes degrade online human-machine interfacing methods and require costly post-hoc manual cleaning in the offline setting. In this article we propose an automated error correction methodology using a deep metric learning (DML) framework, generating embedding spaces in which spiking events can be both identified and assigned to their respective sources. Furthermore, we investigate the relative ability of different DML techniques to preserve the intraclass semantic structure needed to identify incorrect class labels in neurophysiological time series. Motivated by this analysis, we propose locality sensitive mining, an easily implemented sampling-based augmentation to typical DML losses which substantially improves the local semantic structure of the embedding space. We demonstrate the utility of this method to generate embedding spaces which can be used to automatically identify incorrectly labeled spiking events with high accuracy.

A cellular overview of immunometabolism in systemic lupus erythematosus.

Systemic lupus erythematosus (SLE) is a complex autoimmune disease, characterized by a breakdown of immune tolerance and the development of autoantibodies against nucleic self-antigens. Immunometabolism is a rapidly expanding scientific field investigating the metabolic programming of cells of the immune system. During the normal immune response, extensive reprogramming of cellular metabolism occurs, both to generate adenosine triphosphate and facilitate protein synthesis, and also to manage cellular stress. Major pathways upregulated include glycolysis, oxidative phosphorylation, the tricarboxylic acid cycle and the pentose phosphate pathway, among others. Metabolic reprogramming also occurs to aid resolution of inflammation. Immune cells of both patients with SLE and lupus-prone mice are characterized by metabolic abnormalities resulting in an altered functional and inflammatory state. Recent studies have described how metabolic reprogramming occurs in many cell populations in SLE, particularly CD4+ T cells, e.g. favouring a glycolytic profile by overactivation of the mechanistic target of rapamycin pathway. These advances have led to an increased understanding of the metabolic changes affecting the inflammatory profile of T and B cells, monocytes, dendritic cells and neutrophils, and how they contribute to autoimmunity and SLE pathogenesis. In the current review, we aim to summarize recent advances in the field of immunometabolism involved in SLE and how these could potentially lead to new therapeutic strategies in the future.

LAT1 enables T cell activation under inflammatory conditions.

The aim of this study was to assess the L-type amino acid transporter-1 (LAT1) as a possible therapeutic target for rheumatoid arthritis (RA). Synovial LAT1 expression in RA was monitored by immunohistochemistry and transcriptomic datasets. The contribution of LAT1 to gene expression and immune synapse formation was assessed by RNA-sequencing and total internal reflection fluorescent (TIRF) microscopy, respectively. Mouse models of RA were used to assess the impact of therapeutic targeting of LAT1. LAT1 was strongly expressed by CD4+ T cells in the synovial membrane of people with active RA and the level of expression correlated with levels of ESR and CRP as well as DAS-28 scores. Deletion of LAT1 in murine CD4+ T cells inhibited the development of experimental arthritis and prevented the differentiation of CD4+ T cells expressing IFN-γ and TNF-α, without affecting regulatory T cells. LAT1 deficient CD4+ T cells demonstrated reduced transcription of genes associated with TCR/CD28 signalling, including Akt1, Akt2, Nfatc2, Nfkb1 and Nfkb2. Functional studies using TIRF microscopy revealed a significant impairment of immune synapse formation with reduced recruitment of CD3ζ and phospho-tyrosine signalling molecules in LAT1 deficient CD4+ T cells from the inflamed joints but not the draining lymph nodes of arthritic mice. Finally, it was shown that a small molecule LAT1 inhibitor, currently undergoing clinical trials in man, was highly effective in treating experimental arthritis in mice. It was concluded that LAT1 plays a critical role in activation of pathogenic T cell subsets under inflammatory conditions and represents a promising new therapeutic target for RA.

Dynamic mitochondrial transcription and translation in B cells control germinal center entry and lymphomagenesis.

Germinal center (GC) B cells undergo proliferation at very high rates in a hypoxic microenvironment but the cellular processes driving this are incompletely understood. Here we show that the mitochondria of GC B cells are highly dynamic, with significantly upregulated transcription and translation rates associated with the activity of transcription factor A, mitochondrial (TFAM). TFAM, while also necessary for normal B cell development, is required for entry of activated GC precursor B cells into the germinal center reaction; deletion of Tfam significantly impairs GC formation, function and output. Loss of TFAM in B cells compromises the actin cytoskeleton and impairs cellular motility of GC B cells in response to chemokine signaling, leading to their spatial disorganization. We show that B cell lymphoma substantially increases mitochondrial translation and that deletion of Tfam in B cells is protective against the development of lymphoma in a c-Myc transgenic mouse model. Finally, we show that pharmacological inhibition of mitochondrial transcription and translation inhibits growth of GC-derived human lymphoma cells and induces similar defects in the actin cytoskeleton.

A myoelectric digital twin for fast and realistic modelling in deep learning.

Muscle electrophysiology has emerged as a powerful tool to drive human machine interfaces, with many new recent applications outside the traditional clinical domains, such as robotics and virtual reality. However, more sophisticated, functional, and robust decoding algorithms are required to meet the fine control requirements of these applications. Deep learning has shown high potential in meeting these demands, but requires a large amount of high-quality annotated data, which is expensive and time-consuming to acquire. Data augmentation using simulations, a strategy applied in other deep learning applications, has never been attempted in electromyography due to the absence of computationally efficient models. We introduce a concept of Myoelectric Digital Twin - highly realistic and fast computational model tailored for the training of deep learning algorithms. It enables simulation of arbitrary large and perfectly annotated datasets of realistic electromyography signals, allowing new approaches to muscular signal decoding, accelerating the development of human-machine interfaces.

Plasma iron controls neutrophil production and function.

Low plasma iron (hypoferremia) induced by hepcidin is a conserved inflammatory response that protects against infections but inhibits erythropoiesis. How hypoferremia influences leukocytogenesis is unclear. Using proteomic data, we predicted that neutrophil production would be profoundly more iron-demanding than generation of other white blood cell types. Accordingly in mice, hepcidin-mediated hypoferremia substantially reduced numbers of granulocytes but not monocytes, lymphocytes, or dendritic cells. Neutrophil rebound after anti-Gr-1-induced neutropenia was blunted during hypoferremia but was rescued by supplemental iron. Similarly, hypoferremia markedly inhibited pharmacologically stimulated granulopoiesis mediated by granulocyte colony-stimulating factor and inflammation-induced accumulation of neutrophils in the spleen and peritoneal cavity. Furthermore, hypoferremia specifically altered neutrophil effector functions, suppressing antibacterial mechanisms but enhancing mitochondrial reactive oxygen species-dependent NETosis associated with chronic inflammation. Notably, antagonizing endogenous hepcidin during acute inflammation enhanced production of neutrophils. We propose plasma iron modulates the profile of innate immunity by controlling monocyte-to-neutrophil ratio and neutrophil activity in a therapeutically targetable system.

A strength inversion origin for non-volcanic tremor.

Non-volcanic tremor is a particularly enigmatic form of seismic activity. In its most studied subduction zone setting, tremor typically occurs within the plate interface at or near the shallow and deep edges of the interseismically locked zone. Detailed seismic observations have shown that tremor is composed of repeating small low-frequency earthquakes, often accompanied by very-low-frequency earthquakes, all involving shear failure and slip. However, low-frequency earthquakes and very-low-frequency earthquakes within each cluster show nearly constant source durations for all observed magnitudes, which implies characteristic tremor sub-event sources of near-constant size. Here we integrate geological observations and geomechanical lab measurements on heterogeneous rock assemblages representative of the shallow tremor region offshore the Middle America Trench with numerical simulations to demonstrate that these tremor events are consistent with the seismic failure of relatively weaker blocks within a stronger matrix. In these subducting rocks, hydrothermalism has led to a strength-inversion from a weak matrix with relatively stronger blocks to a stronger matrix with embedded relatively weaker blocks. Tremor naturally occurs as the now-weaker blocks fail seismically while their surrounding matrix has not yet reached a state of general seismic failure.

Online Tracking of the Phase Difference Between Neural Drives to Antagonist Muscle Pairs in Essential Tremor Patients.

Transcutaneous electrical stimulation has been applied in tremor suppression applications. Out-of-phase stimulation strategies applied above or below motor threshold result in a significant attenuation of pathological tremor. For stimulation to be properly timed, the varying phase relationship between agonist-antagonist muscle activity during tremor needs to be accurately estimated in real-time. Here we propose an online tremor phase and frequency tracking technique for the customized control of electrical stimulation, based on a phase-locked loop (PLL) system applied to the estimated neural drive to muscles. Surface electromyography signals were recorded from the wrist extensor and flexor muscle groups of 13 essential tremor patients during postural tremor. The EMG signals were pre-processed and decomposed online and offline via the convolution kernel compensation algorithm to discriminate motor unit spike trains. The summation of motor unit spike trains detected for each muscle was bandpass filtered between 3 to 10 Hz to isolate the tremor related components of the neural drive to muscles. The estimated tremorogenic neural drive was used as input to a PLL that tracked the phase differences between the two muscle groups. The online estimated phase difference was compared with the phase calculated offline using a Hilbert Transform as a ground truth. The results showed a rate of agreement of 0.88 ± 0.22 between offline and online EMG decomposition. The PLL tracked the phase difference of tremor signals in real-time with an average correlation of 0.86 ± 0.16 with the ground truth (average error of 6.40° ± 3.49°). Finally, the online decomposition and phase estimation components were integrated with an electrical stimulator and applied in closed-loop on one patient, to representatively demonstrate the working principle of the full tremor suppression system. The results of this study support the feasibility of real-time estimation of the phase of tremorogenic neural drive to muscles, providing a methodology for future tremor-suppression neuroprostheses.

A computational fluid dynamics assessment of 3D printed ventilator splitters and restrictors for differential multi-patient ventilation.

BACKGROUND: The global pandemic of novel coronavirus (SARS-CoV-2) has led to global shortages of ventilators and accessories. One solution to this problem is to split ventilators between multiple patients, which poses the difficulty of treating two patients with dissimilar ventilation needs. A proposed solution to this problem is the use of 3D-printed flow splitters and restrictors. There is little data available on the reliability of such devices and how the use of different 3D printing methods might affect their performance. METHODS: We performed flow resistance measurements on 30 different 3D-printed restrictor designs produced using a range of fused deposition modelling and stereolithography printers and materials, from consumer grade printers using polylactic acid filament to professional printers using surgical resin. We compared their performance to novel computational fluid dynamics models driven by empirical ventilator flow rate data. This indicates the ideal performance of a part that matches the computer model. RESULTS: The 3D-printed restrictors varied considerably between printers and materials to a sufficient degree that would make them unsafe for clinical use without individual testing. This occurs because the interior surface of the restrictor is rough and has a reduced nominal average diameter when compared to the computer model. However, we have also shown that with careful calibration it is possible to tune the end-inspiratory (tidal) volume by titrating the inspiratory time on the ventilator. CONCLUSIONS: Computer simulations of differential multi patient ventilation indicate that the use of 3D-printed flow splitters is viable. However, in situ testing indicates that using 3D printers to produce flow restricting orifices is not recommended, as the flow resistance can deviate significantly from expected values depending on the type of printer used. TRIAL REGISTRATION: Not applicable.

B Cell Metabolism and Autophagy in Autoimmunity.

B cells are central to the pathogenesis of multiple autoimmune diseases, through antigen presentation, cytokine secretion, and the production of autoantibodies. During development and differentiation, B cells undergo drastic changes in their physiology. It is emerging that these are accompanied by equally significant shifts in metabolic phenotype, which may themselves also drive and enforce the functional properties of the cell. The dysfunction of B cells during autoimmunity is characterised by the breaching of tolerogenic checkpoints, and there is developing evidence that the metabolic state of B cells may contribute to this. Determining the metabolic phenotype of B cells in autoimmunity is an area of active study, and is important because intervention by metabolism-altering therapeutic approaches may represent an attractive treatment target.

EGFR-HIF1α signaling positively regulates the differentiation of IL-9 producing T helper cells.

Interleukin 9 (IL-9)-producing helper T (Th9) cells are essential for inducing anti-tumor immunity and inflammation in allergic and autoimmune diseases. Although transcription factors that are essential for Th9 cell differentiation have been identified, other signaling pathways that are required for their generation and functions are yet to be explored. Here, we identify that Epidermal Growth Factor Receptor (EGFR) is essential for IL-9 induction in helper T (Th) cells. Moreover, amphiregulin (Areg), an EGFR ligand, is critical for the amplification of Th9 cells induced by TGF-ß1 and IL-4. Furthermore, our data show that Areg-EGFR signaling induces HIF1α, which binds and transactivates IL-9 and NOS2 promoters in Th9 cells. Loss of EGFR or HIF1α abrogates Th9 cell differentiation and suppresses their anti-tumor functions. Moreover, in line with its reliance on HIF1α expression, metabolomics profiling of Th9 cells revealed that Succinate, a TCA cycle metabolite, promotes Th9 cell differentiation and Th9 cell-mediated tumor regression.

Losartan may attenuate altitude-related sleep disturbance.

INTRODUCTION: Sleep disturbance is common at high altitude and likely driven by an exaggerated peripheral chemoreceptor response which leads to apnoeic episodes and arousal. We hypothesised that this heightened response is in part mediated through angiotensin II receptors in the carotid body. To examine this link, we studied the effect of angiotensin II receptor blocker on sleep disturbance. METHODS: Twenty participants paired by age, gender and ACE phenotype ascended to the Whymper Hut (5000 m) on Chimborazo in the Ecuadorean Andes as part of a double-blinded randomised placebo-controlled study of physiological mechanisms. Subjects were randomised to either losartan 100 mg daily or placebo. The primary outcome of sleep efficiency was measured using wrist-mounted actigraphs. One pair was excluded from analysis after descending before the end of the study due to acute mountain sickness. RESULTS: There was a significantly different response to altitude between the two groups (F=3.274, p=0.029), as a decline in sleep efficiency in the placebo group (F=10.259, p<0.001) was not replicated in the angiotensin II receptor blocker group (F=0.459, p=0.713). CONCLUSION: The absence of any significant sleep disturbance in the intervention group suggests that peripheral chemoreceptor hypersensitivity is largely mediated by angiotensin II receptor activation. However, further research is needed to confirm our findings and to study the potential mechanisms of action.

Deep Learning for Robust Decomposition of High-Density Surface EMG Signals.

Blind source separation (BSS) algorithms, such as gradient convolution kernel compensation (gCKC), can efficiently and accurately decompose high-density surface electromyography (HD-sEMG) signals into constituent motor unit (MU) action potential trains. Once the separation matrix is blindly estimated on a signal interval, it is also possible to apply the same matrix to subsequent signal segments. Nonetheless, the trained separation matrices are sub-optimal in noisy conditions and require that incoming data undergo computationally expensive whitening. One unexplored alternative is to instead use the paired HD-sEMG signal and BSS output to train a model to predict MU activations within a supervised learning framework. A gated recurrent unit (GRU) network was trained to decompose both simulated and experimental unwhitened HD-sEMG signal using the output of the gCKC algorithm. The results on the experimental data were validated by comparison with the decomposition of concurrently recorded intramuscular EMG signals. The GRU network outperformed gCKC at low signal-to-noise ratios, proving superior performance in generalising to new data. Using 12 seconds of experimental data per recording, the GRU performed similarly to gCKC, at rates of agreement of 92.5% (84.5%-97.5%) and 94.9% (88.8%-100.0%) respectively for GRU and gCKC against matched intramuscular sources.