A team of researchers from Aalborg University and Aalborg University Hospital in Denmark, working with collaborators in Qatar and the United States, has now reported the first controlled in vivo evaluation of a trilayer composite electrospun scaffold for urethral reconstruction. The construct, designated CPC, consists of a core of poly(L-lactide-co-glycolide), or PLGA, sandwiched between two layers of chitosan, a naturally derived polysaccharide. The design rationale is biomimetic: the luminal chitosan layer is intended to mimic the urethral mucosa, providing a hydrophilic, bioactive surface that supports epithelial cell attachment and migration while contributing the antimicrobial properties characteristic of chitosan-based biomaterials. The central electrospun PLGA layer serves as the mechanical backbone, analogous to the submucosal tissue of the native urethral wall, offering tunable tensile strength and gradual hydrolytic degradation over the healing timeframe. The outer chitosan layer interfaces with the surrounding corpus spongiosum, the vascular tissue that envelops the urethra, promoting cellular infiltration and angiogenesis.

The biological logic behind the chitosan layers rests on well-documented immunomodulatory chemistry. Chitosan oligosaccharides act on macrophage pattern-recognition receptors to drive polarization of these immune cells toward the M2 phenotype, the anti-inflammatory state associated with tissue repair rather than with destruction. M2 macrophages secrete pro-angiogenic factors including vascular endothelial growth factor, or VEGF, supporting early vascularization of healing tissue. Chitosan additionally confers hydrophilicity and antimicrobial activity that complement the structural advantages of synthetic polyester cores. The hypothesis was that this layered architecture would attenuate the foreign-body response and support vascularization while preserving the mechanical integrity needed for surgical handling. Prior in vitro work on a related trilayer chitosan/poly(L-lactide)/chitosan construct had demonstrated superior mechanical strength, suture retention, smooth muscle cell growth, and angiogenic potential compared with monolayer PLA scaffolds, motivating the present in vivo evaluation. The Danish team substituted PLGA for PLA in the core layer because PLGA’s degradation rate can be tuned through the lactide-to-glycolide ratio, offering more flexibility than the slower-degrading PLA, and because PLGA carries broad regulatory precedent in absorbable medical devices.

What distinguishes this study from much of the preclinical urethral tissue-engineering literature is its methodological rigor. A 2021 systematic review of 28 rabbit studies published between 2014 and 2020 found a mean adherence of only 53 percent to the ARRIVE guidelines for reporting animal research, with no included study reporting sample-size estimation methodology and the majority lacking blinded outcome assessment. The new study was designed and reported in accordance with ARRIVE 2.0, directly addressing each of these documented gaps. The researchers performed an a priori sample-size calculation, prespecified a continuous quantitative primary endpoint rather than the binary outcomes such as fistula formation that have constrained formal power calculation in earlier work, and implemented blinded outcome assessment throughout. Twenty-four male New Zealand White rabbits were allocated in a one-to-one-to-one ratio to dorsal onlay urethroplasty with the trilayer CPC scaffold, with a monolayer PLGA scaffold, or to a sham surgical control, using weight-stratified allocation. Surgeons were blinded to group allocation during urethral exposure and defect creation, and because both scaffolds were thin, white, semi-flexible electrospun sheets that were not reliably distinguishable on inspection or handling, surgeons remained effectively blinded to scaffold type throughout the procedure. Retrospective misidentification of scaffold type was common when surgeons were informally asked to guess, supporting the integrity of the blinding.

The surgical model itself mirrors adult substitution urethroplasty technique. Following urethral catheterization, the penile urethra was exposed through a ventral midline skin incision, mobilized, and a standardized dorsal urethral defect measuring 10 by 3 millimeters was created. In scaffold-treated groups, a dorsal onlay graft of identical dimensions was placed on the defect and sutured with interrupted 6-0 polydioxanone monofilament sutures. To divert urine and protect the healing site, a perineostomy was created proximal to the urethroplasty, a strategy preferred over indwelling catheterization because rabbits do not reliably tolerate prolonged urethral catheters. The primary endpoint was vessel density on CD31 immunohistochemistry at 12 weeks, with secondary endpoints including tissue thickness, fibrosis, smooth muscle content, re-epithelialization, inflammatory cell scores, and urethral patency assessed by post-mortem retrograde urethrography. All histological assessments were performed by a single pathologist blinded to treatment allocation, and intra-observer reliability of vessel-density scoring was excellent, with an intraclass correlation coefficient of 0.97.

The results told a more nuanced story than the hypothesis predicted. Vessel density differed significantly between treatment groups overall, with sham animals exhibiting the highest density at a median of 78.6 vessels per square millimeter, monolayer PLGA the lowest at 39.1, and the trilayer CPC intermediate at 46.3. The direction of effect matched the in vitro pro-angiogenic activity reported for chitosan-containing constructs, and the observed CPC-PLGA median difference of 7.2 vessels per square millimeter closely matched the prespecified clinically meaningful difference of seven used in the sample-size calculation. However, the within-group variability substantially exceeded the planning assumption, reducing statistical power below the design target and widening the 95 percent confidence interval around the CPC-PLGA difference to span from minus 27.3 to 41.6 vessels per square millimeter. The trilayer did not significantly differ from the monolayer, and the overall between-group variation was driven by higher values in sham animals than in PLGA-treated animals. The authors interpret the sham finding carefully: the rabbit urethra revascularizes robustly after primary closure of a small dorsal defect, so the sham group represents an upper bound for vascularization in this model rather than a vascularization-deficient comparator.

Perhaps the most striking quantitative finding was a coherent host-response signature linking vascularization, tissue thickness, and fibrosis. Vessel density was inversely correlated with tissue thickness and with fibrosis, while tissue thickness was positively correlated with fibrosis. This pattern describes an axis of host response in which scaffold-supported remodeling produced thicker, more fibrotic, and less vascularized tissue, consistent with the behavior of synthetic scaffolds at intermediate timepoints, where a sustained foreign-body response can shift the balance from constructive vascularization toward fibrotic encapsulation. The thickness finding also carries a metric caveat: scaffold-driven regeneration produced thicker tissue at the implantation site over which a comparable absolute vessel population was distributed, biasing the per-area vessel density downward. Fibrosis itself did not differ significantly between groups, with median collagen area near 82 to 84 percent across all arms, a near-equivalence in the absence of any scaffold material suggesting that the dominant contributor to the measured collagen signal was native urethral and adjacent corpus cavernosum collagen captured within the region of interest rather than scaffold-induced scar formation.

On the inflammatory front, scaffold implantation elicited a foreign-body response in both treatment groups, with significantly elevated multinucleated giant-cell scores compared with sham. Yet the macrophage comparison provided the most defensible signal in favor of the trilayer design: median macrophage scores were lower in CPC than in PLGA, a difference that trended toward but did not reach statistical significance. This pattern is consistent with chitosan attenuating the foreign-body response, possibly through its established capacity to favor M2 over M1 macrophage polarization via chitosan-oligosaccharide signaling, while PLGA implants drive a stereotyped macrophage and giant-cell signature partly through acidic by-products of polymer hydrolysis. The authors are careful to note that this chitosan-mediated modulation is inferred from semi-quantitative macrophage counts rather than direct phenotypic markers, and that the elevated giant-cell counts in scaffold-treated groups cannot be classified as uniformly pro-inflammatory without immunohistochemistry for canonical M1 and M2 polarization markers. Meanwhile, re-epithelialization was essentially complete across all groups, with median urothelial coverage of 100 percent in the CPC group, and no animal demonstrated ulceration, unresolved wound bed, or tissue necrosis at the 12-week endpoint. The few focal regions of missing urothelium lacked the edge reaction, granulation tissue, and inflammatory infiltrate characteristic of true ulceration and were interpreted as denudation of mature urothelium, possibly a processing artifact.

Functionally, the news was uniformly good. Retrograde urethrography demonstrated preserved patency in all 23 animals completing the planned 12-week endpoint, with no fistulae or contrast extravasation observed. No animal met the prespecified stricture criterion of a 50 percent or greater reduction in urethral diameter at the implantation site, and lumen-width ratios were numerically wider in scaffold groups than in sham, a directional pattern consistent with the structural role of an onlay graft in providing a template for lumen-supporting tissue regeneration during spongiosal fibrosis. No postoperative complications or infections were observed during follow-up, and no macroscopic fistulae, adhesions, necrosis, or calculi were seen at harvest. One sham animal was euthanized early because of acute-onset bilateral hindlimb paralysis, judged unrelated to the experimental intervention, and no animals met exclusion criteria during acclimatization.

The authors conclude that the trilayer scaffold was feasible to implant and well tolerated, supporting tissue regeneration with preserved urethral patency, but that the findings refine rather than confirm the development hypothesis. They identify several priorities for future work: earlier endpoints at two and four weeks to directly characterize the chitosan-mediated angiogenic and immunomodulatory window, M1/M2 macrophage and giant-cell phenotyping to test the polarization hypothesis, mechanical-modulus tuning to match the elastic modulus of native urethral tissue at approximately 25 kilopascals, and modulation of the chitosan-to-PLGA mass ratio as a further design lever. Crosslinking strategies to slow chitosan resorption could extend the bioactive window of the surface layers, and translational evaluation will ultimately require larger animal models with urethral anatomy closer to humans, longer follow-up, and head-to-head comparison against the autologous buccal mucosa graft that remains the clinical reference standard. The rabbit dorsal onlay model itself demonstrated an excellent safety profile across all groups, supporting its utility as a reproducible platform for controlled comparison of subsequent urethral scaffolds, and the biological criteria validated here, including complete re-epithelialization without ulceration and preserved patency, are of considerable interest for hypospadias-directed scaffold development, where excessive inflammation and fibrosis contribute to fistula formation and repair failure.

Subject of Research: In vivo evaluation of a chitosan/PLGA/chitosan trilayer electrospun scaffold versus monolayer PLGA for urethral reconstruction in rabbits

Article Title: In vivo evaluation of a chitosan/PLGA/chitosan composite versus monolayer PLGA electrospun scaffold for dorsal onlay urethroplasty: a controlled study in rabbits

Article References: Mahdi, M. B., Mohtaram, F., Haase, R. N., Tuckus, G., Fojan, P., Bergman, J., Thomsen, C., Kjærgaard, B., Sørensen, S. R., Alipour, H., Abbas, T. O., Pennisi, C. P., & Nielsen, T. K. (2026). In vivo evaluation of a chitosan/PLGA/chitosan composite versus monolayer PLGA electrospun scaffold for dorsal onlay urethroplasty: a controlled study in rabbits. Frontiers in Bioengineering and Biotechnology, 14, Article 1891449. https://doi.org/10.3389/fbioe.2026.1891449

Image Credits: AI Generated

DOI: 10.3389/fbioe.2026.1891449

Keywords: urethroplasty, urethral stricture, chitosan, PLGA, electrospun scaffold, tissue engineering, vascularization, foreign-body response, rabbit model, regenerative medicine, biomaterials, hypospadias

The review, authored by Miaozhe Huo, Zhengyang Xue, Yifei Sheng, and Shiying Li from City University of Hong Kong and Shenzhen Nanshan People’s Hospital, proposes a resolution-aware analytical framework that treats the intracortical microelectrode–brain interface as a layered system. The first layer is geometric: the implant-centered coordinate that locates every measurement relative to the device surface or reconstructed probe tract. The second layer is biological: the multivariate local context — cell types, molecular states, tissue structure, and cell–cell relationships — that exists within and across those coordinates. The authors argue that combining both layers preserves the defining geometry of the biohybrid interface while exposing biological organization that a distance value alone cannot capture.

The motivation is well established. Intracortical microelectrodes record neural activity from within the cortex, but chronic signal quality varies across channels, devices, animals, and time. Mechanical injury at insertion, persistent micromotion, vascular disruption, inflammatory signaling, glial and immune responses, and neuronal displacement or loss all contribute to an evolving tissue response. Material and device properties influence the magnitude and time course of that response. Around the interface, these interdependent processes produce a heterogeneous, multicomponent reaction rather than a single uniform scar. Recording performance, in turn, varies with this tissue context together with electrode design, electrochemistry, cortical depth, neuronal availability, and temporal history.

What makes the new framework timely is the arrival of spatially resolved molecular technologies at the implant interface. Direct studies now include Visium whole-transcriptome measurements with multicellular capture spots, regional GeoMx RNA or protein profiles, registered histology and immunofluorescence, and targeted Xenium transcript-coordinate maps with marker-based cell-type inference reported in a 2026 preprint. Each assay contributes a distinct native unit: a Visium spot captures a local multicellular mixture at roughly 55 micrometers; a Xenium readout provides transcript coordinates within a targeted 100-gene panel; a GeoMx area of illumination represents a selected tissue aggregate; and an image feature reflects its staining, segmentation, and registration. Together, these measurements make it possible to analyze recurring local contexts within implant-centered coordinates — something earlier bulk approaches could not do.

Central to the framework is a careful separation of three vocabularies that the authors say are often conflated: established implant-referenced coordinates and sampling regions, assay-native measurement units, and analytically derived spatial contexts. A distance ring, a capture spot, or an unsupervised spatial cluster, they argue, is not a niche by itself. A true spatial niche requires four minimum reporting elements: a declared focal unit, a biologically and technically justified neighborhood rule and scale, the input features used to represent context, and a validation strategy stating the criterion used to evaluate the representation. This discipline is designed to make niche-level findings reproducible and comparable across laboratories, platforms, and studies.

The framework organizes niche-level analysis into three complementary directions. The first is niche identification and biological characterization, drawing on method families such as composition-centered approaches like Spatial-LDA, CellCharter, and CytoCommunity; state- and program-centered methods like scNiche and NicheCompass; spatial-domain scaffolds including BASS, BayesSpace, and BANKSY; image-linked tools such as Starfysh, histoCAT, and SPIAT; and serial-section methods like PASTE and STitch3D that extend two-dimensional observations toward three-dimensional reconstruction. The second direction examines implant-referenced spatial organization — patterns, gradients, boundaries, and molecular-state ordering — using tools such as SpatialDE, SPARK-X, GASTON, SpaceFlow, and ONTraC. The third evaluates niche-conditioned transcript and protein relationships, with methods like Niche-DE, MISTy, SpatialDM, COMMOT, and LIANA+ mapping candidate ligand–receptor interactions onto defined niche contexts.

The current evidence base, though still limited, already hints at what niche analysis could reveal. In rat motor cortex, Visium studies established device-referenced whole-transcriptome variation and exploratory molecular associations with multiunit activity, local field potentials, signal-to-noise ratio, neuronal density, and GFAP intensity. In mice, GeoMx studies documented declining immune activation with distance from the probe and persistent neuronal, oligodendrocyte, and autophagy protein deficits across radial regions at eight weeks. The Xenium preprint compared electrically inactive surrogate probes differing in material and cross-sectional size, finding that local transcript and inferred cell-type patterns varied more with probe size than with material, especially at six weeks. Notably, none of the niche-analysis methods reviewed has yet been applied to intracortical microelectrode data — a gap the framework is designed to close.

The authors are careful about statistical practice, distinguishing animal-level replication from within-animal sampling. Cells, spots, sampled regions, and implant sites within the same animal are nested observations; counting them as independent replicates would constitute pseudoreplication. They also flag batch effects, tissue deformation, registration uncertainty, and tract-boundary placement as factors that can shift distance assignments and the features attributed to a local region. The Xenium study, for example, averaged gene-expression estimates for each cell type within specified spatial regions of each animal before analysis of variance, treating the animal as the biological replicate.

Perhaps the most forward-looking part of the framework connects spatial niches to recording performance. Existing studies have posed recording quality as an explicit supervised-learning target, combining temporal, electrochemical, electrode-design, cortical-depth, and prior-recording variables to predict whether a channel will yield sortable units. The review proposes candidate niche-derived predictors at matched functional scales: local niche composition and niche-conditioned state for registered channels or sites; device-relative niche distribution and across-site heterogeneity for shanks and implants; and recurrent niche assemblies or ecotypes for whole specimens. Outcome-linked spatial transcriptomics studies in cancer — from compartment-specific gene signatures to niche-fraction and ecotype models — provide methodological precedents for pooling local features to larger outcome units.

The authors conclude that distance remains an essential coordinate for microelectrode research, but that spatial niches add the cellular, molecular, and structural context present within and across implant-centered coordinates. By reporting the focal unit, neighborhood rule and scale, input features, and validation strategy, researchers can make that context explicit and comparable. The resulting material–niche–function question offers a common language for comparing spatial omics, morphology, interface engineering, and chronic recording studies — and, ultimately, for understanding why some electrodes keep listening to the brain while others fall silent.

Subject of Research: A methodological framework for analyzing spatial niches at the intracortical microelectrode–brain interface using spatial transcriptomics

Article Title: From peri-implant distance ranges to spatial niches: a resolution-aware analytical framework for the intracortical microelectrode–brain interface

Article References: Huo, M., Xue, Z., Sheng, Y., & Li, S. (2026). From peri-implant distance ranges to spatial niches: a resolution-aware analytical framework for the intracortical microelectrode–brain interface. Frontiers in Materials, 13, Article 1969087. https://doi.org/10.3389/fmats.2026.1969087

Image Credits: AI Generated

DOI: 10.3389/fmats.2026.1969087

Keywords: intracortical microelectrodes, spatial transcriptomics, spatial niches, brain–computer interfaces, neural implants, biomaterials, foreign body response, Visium, Xenium, GeoMx, recording performance, neuroinflammation

The engineering logic behind permeability rests on well-established biology. When the body encounters an implanted material, the immune system mounts a foreign body response: proteins adsorb onto the surface, inflammatory cells arrive, and eventually a dense collagenous capsule walls the device off from surrounding tissue. That capsule degrades signal quality for electrodes, isolates sensors from the very biochemistry they are meant to measure, and can cause chronic irritation. Mechanical mismatch compounds the problem. Skin, muscle, brain and blood vessels deform continuously with motion, pulsation and growth, so a stiff slab of silicon or even a thin but airtight polymer film experiences shear stresses at its edges that promote delamination and tissue damage. Porous fibre networks sidestep both failure modes simultaneously, the review explains, because their open structure permits nutrient and waste exchange, supports cell proliferation at the interface in a controlled way, and distributes mechanical strain across millions of interlocking fibre contacts rather than concentrating it at a perimeter.

The authors organise the field around four governing parameters: permeability, conductivity, mechanical adaptability and biocompatibility. These are not independent. Increasing porosity generally improves mass transport and breathability but reduces the conductive cross-section available for electrical signals, so designers must trade one property against another. The review highlights materials strategies that break this trade-off, including liquid-metal fibre mats that remain conductive while stretching to 1,800 percent, silver-gold core-sheath nanowires that combine high conductivity with corrosion resistance, conductive hydrogels whose water-rich networks resemble soft tissue, and MXene-coated fibres that bring high capacitance to microelectrodes. Strain-insensitive conductor designs, such as buckled shells and built-in helical structures, decouple electrical resistance from mechanical deformation, allowing a fibre to stretch without its signal drifting. Zwitterionic coatings and antifouling polymers suppress nonspecific protein adsorption, while biomolecular functionalisation with fibronectin, collagen, laminin or RGD peptides can actively encourage desired cell adhesion and neurite extension.

Manufacturing is where fibre architecture reveals its most practical advantage, because the textile industry has spent centuries perfecting scalable ways to make, weave, knit, braid and coat fibres. The review divides fabrication into top-down and bottom-up routes. Top-down methods include in-textile photolithography, which patterns high-resolution interconnected metal circuits directly inside fabric while preserving stretchability and permeability, as well as screen printing, inkjet printing, dip coating and direct ink writing. Bottom-up approaches centre on electrospinning, which draws nanofibres from polymer solutions under high voltage and can produce mats with controlled hierarchical porosity, and on thermal drawing, coaxial spinning and wet spinning, which build multi-material fibres with core-sheath architectures. A metre-scale heterostructure printing method for high-toughness fibre electrodes, reported in 2025, exemplifies the push toward industrial throughput. The authors emphasise that combining the two families of techniques gives engineers control over structure from the nanometre to the metre, which is precisely the multiscale range over which biological tissues operate.

What elevates these devices from passive conductors to intelligent bioelectronics is the integration of sensing, actuation and therapy into the same fibre platform. On the sensing side, the review catalogues electrophysiological electrodes for electrocardiography, electromyography and electroencephalography, including knit band sensors that let amputees control prosthetic hands through surface EMG; biochemical sensors that track glucose in sweat through electrochemical reactions on functionalised fibres, or monitor amniotic fluid in real time during pregnancy; and biomechanical strain sensors that follow joint bending, respiration and ligament loading. Piezoelectric and triboelectric fibre systems convert body motion into electrical signals and even into power, while magnetoelastic soft fibres allow wireless readout. On the therapeutic side, fibre platforms deliver electrical stimulation, photothermal heating, optogenetic light and drugs. A bioabsorbable mechanoelectric fibre has been demonstrated as an electrical stimulation suture, and multifunctional microelectronic fibres have wirelessly modulated both gut and brain neural circuits in animal studies.

Perhaps the most clinically compelling vision the review describes is the closed-loop therapeutic platform, in which sensing and therapy are fused into a single feedback system. A skin-interfaced electronic wound bandage reported in 2025 combines three-dimensional sensing with targeted treatment, monitoring the wound environment and responding to it. A self-powered permeable electronic dressing manages exudate, applies electrostimulation and releases drugs for chronic wound healing. Liquid-metal stretchable bioelectronic fibres have been designed for electrical stimulation and drug delivery in minimally invasive cardiac therapy, and implantable optical fibres have delivered immunotherapeutics while simultaneously measuring tumour impedance. In such systems, the fibre network acts as both nervous system and hand of the therapist: it perceives a physiological state, computes or transmits the information, and intervenes, all while remaining permeable enough that the healing tissue grows into and around it rather than rejecting it.

Secondary features that sound mundane turn out to be decisive for real-world adoption. Washability, for instance, determines whether textile-based health monitors can survive the laundry, and the review notes that standardised testing protocols for wash fastness, mechanical fatigue and long-term biocompatibility are still missing but essential for regulatory translation. Optical transparency matters for devices that must not obstruct vision or clinical inspection, from gas-permeable contact lens sensors with metal-coated nanofibre meshes to transparent conductors that maintain stable conductivity under large deformation. Biodegradability opens a different design space: polycaprolactone and polylactic acid fibres can be engineered to degrade on predictable timescales, and biodegradable cotton-based piezoresistive textiles point toward devices that simply vanish when their job is done, eliminating a second surgery. The review also confronts the safety shadow hanging over fibrous biomaterials, namely the lessons of asbestos, where biopersistent fibres caused inflammation and carcinogenesis, and stresses that fibre dimensions, persistence and degradation chemistry must be controlled deliberately.

The translational pathway the authors sketch is candid about remaining obstacles. Long-term in vivo stability data for most fibre systems are limited compared with the decades of clinical experience behind conventional implants. Standardisation is immature: without agreed benchmarks for permeability, fatigue life, wash durability and biocompatibility, regulators and clinicians cannot compare devices on equal footing. Power and data remain bottlenecks for fully implantable systems, although fibre-shaped batteries, supercapacitors and biofuel cells, along with embroidered and liquid-metal antennas for wireless powering and communication, are closing the gap. There are also questions of scale-up economics, of sterilisation compatibility, and of the ethical dimensions of continuous physiological monitoring, from data privacy to the psychological burden of being measured around the clock. The review frames these not as reasons for scepticism but as a research agenda, arguing that the field’s rapid material progress has outpaced its testing infrastructure and that catching up is now the priority.

If the authors are right, the significance of fibre-architected permeable bioelectronics extends beyond any single device category. Digital health care depends on continuous, high-fidelity physiological data streams, and the quality of those streams is bounded by how well electronics can live with biology rather than merely sit on it. By making permeability a first-class design requirement alongside conductivity and compliance, the FIPEB framework reframes the biointerface as something to be engineered collaboratively with tissue rather than defended against it. The convergence of advanced soft materials, textile-scale manufacturing and closed-loop intelligence suggests a future in which a cardiac monitor is a comfortable garment, a neural interface is a thread-like probe the brain tolerates for years, and a wound dressing heals as it measures. The remaining distance between laboratory demonstrations and that future, the review concludes, will be covered not by any single breakthrough but by systematic engineering of the four properties, permeability chief among them, that let electronics and life finally share the same space.

Subject of Research: Fibre-architected permeable bioelectronics for long-term biointegrated sensing and therapy

Article Title: Fibre-architected intelligent permeable bioelectronics

Article References: Wu, Z., Zhang, Y., Huang, Q., & Zheng, Z. (2026). Fibre-architected intelligent permeable bioelectronics. Nature Reviews Bioengineering. https://doi.org/10.1038/s44222-026-00500-z

Image Credits: AI Generated

DOI: 10.1038/s44222-026-00500-z

Keywords: bioelectronics, permeable electronics, fibre architecture, wearable sensors, implantable devices, foreign body response, electrospinning, liquid metal conductors, closed-loop therapy, digital health care, biocompatibility, e-textiles