Search Results (2,126)

Ferulic acid, a natural phenolic phytotherapeutic, which is mainly found in plant cell walls, has attracted the attention of researchers for its multiple pharmacological properties, especially for its anti-tumor potential. As an efficient antioxidant, the compound counteracts oxidative stress and modulates key molecular pathways involved in carcinogenesis. Recent studies demonstrate that ferulic acid exerts antiproliferative, pro-apoptotic, and anti-metastatic effects in various tumor models, including colon, breast, liver, and lung cancers. Mechanistically, ferulic acid affects components of prosurvival-signaling pathways such as PI3K/Akt, MAPK, and NF-κB, and stimulates programmed cell death by diverse mechanisms, including apoptosis, autophagy and ferroptosis. Furthermore, its ability to sensitize cancer cells to chemotherapeutic drugs with low toxicity to normal cells underscores its therapeutic potential. Despite promising preclinical anti-tumor activity, low water solubility and bioavailability limit its clinical use. For this reason, several attempts, ranging from chemical derivatives to nanodevices, have been made to improve ferulicbioavailability and anti-tumor efficacy. This review highlights the most significant and recent strategies to ameliorate the anticancer ability of ferulic acid in the perspective of a clinical application. Full article

Accurate multi-object tracking and metric localization support traffic monitoring and cooperative intelligent transportation at complex intersections. This study presents a fixed-camera vision-map fusion framework that addresses two practical difficulties: axis-aligned boxes poorly represent turning vehicles, and unconstrained image-plane tracking can produce physically implausible trajectories. A map-aided frontend first generates candidate detections using improved You Only Look Once version 8 nano (YOLOv8n) horizontal bounding box (HBB) branch and an improved YOLOv8 oriented bounding box (OBB) branch. A high-definition (HD) map selector then retains the candidate geometry consistent with the straight-driving or turning region and converts it into a unified detection record. The selected reference point is projected to the ground plane through an offline-estimated homography, whereas the appearance feature bypasses the homography and is passed directly to the association stage. The tracking backend uses a 12-dimensional joint image/metric state, symmetric central-difference evaluations of the process and measurement functions, appearance-motion association, and a feasible-road projection derived from HD-map lane polygons. On the evaluated public sequences, the complete configuration achieved a multiple object tracking accuracy (MOTA) of 74.5%, an identification F1 score (IDF1) of 82.6%, 614 identity switches, and a throughput of 26.8 frames per second (FPS) on an RTX 4090 workstation. In a descriptive Vehicle-in-the-Loop case study involving one instrumented vehicle at one intersection, the overall localization mean absolute error (MAE) was 0.180 m, compared with 0.208 m for the baseline end-to-end configuration. These results indicate the feasibility of combining branch-specific vehicle geometry with map-constrained tracking; controlled same-detector comparisons, repeated multi-vehicle trials, and embedded-device latency and power profiling remain necessary for broader claims. Full article

This study presents a multi-tier framework for deploying multi-modal Conversational AI on edge devices, spanning from constrained ultra-low-power systems to high-performance edge workstations. Utilizing an automated model discovery process and a modular benchmarking testbed, the research demonstrates that real-time, fully edge AI execution is feasible through strategic model selection and hardware acceleration. Key outcomes from the performance analysis are as follows. Speech-to-Text: Fun-ASR-Nano achieved the highest transcription accuracy with a Word Error Rate of 0.026, while Moonshine Tiny was the most efficient, recording a Real-Time Factor of 0.036 on the CPU. Scaling up to the high-performance tier, Whisper Large-V3 Turbo demonstrated high speed and robustness on a dedicated GPU, achieving an RTF of 0.093. Language Modeling: The Qwen 2.5 (1.5B Instruct) model, optimized for the Intel edge NPU, delivered robust constrained edge performance with an average generation speed of 20.15 tokens per second and a high semantic accuracy score of 0.86. The non-transformer Liquid LFM-24B model showcased server-level reasoning capabilities on the high-performance edge, reaching an impressive 39.2 tokens per second when fully offloaded to a dedicated GPU, despite its massive VRAM requirements. Text-to-Speech: Piper TTS emerged as the most efficient model for constrained environments (RTF of 0.034). However, Kokoro TTS redefined high-fidelity zero-shot synthesis on the GPU tier, achieving a groundbreaking RTF of 0.024 and far outperforming larger autoregressive audio models like OuteTTS, which remained too slow for real-time use without significant acceleration. Hardware Acceleration and Energy Efficiency: The use of Intel OpenVINO 2026.0 for hardware offloading significantly reduced energy consumption; for example, Whisper Large-V3 Turbo’s energy per audio second dropped from 52.68 Joules on the CPU to just 3.24 Joules on the integrated GPU. Furthermore, dedicated GPU acceleration revealed a critical “race-to-sleep” paradigm, where higher peak wattage is offset by drastically reduced processing times. The study concludes by identifying two optimal cascaded pipelines: a constrained edge tier (Moonshine, Qwen 1.5B, Piper) running on a Khadas NUC (Khadas Technology, Shenzhen, China powered by an Intel processor (Intel Corporation, Santa Clara, CA, USA) maximizing energy efficiency, and a high-performance tier (Whisper V3 Turbo, Liquid LFM-24B, Kokoro) running on an NVIDIA 5060ti, delivering uncompromising accuracy and subsecond latency for privacy-preserving, advanced edge AI. Full article

Conventional oral and injectable drug administration still struggles with unstable plasma levels, weak targeting, and considerable systemic toxicity, problems that become especially acute in chronic disease management. Implantable and wearable biomedical devices offer one path around these limits, yet device-only platforms continue to fall short on drug loading, release control, and protection of fragile therapeutics. Integrating nanoparticle-based biomaterials into such devices has therefore moved from a research curiosity to a serious clinical strategy. As a result, understanding the design principles, translational challenges, and clinical potential of these hybrid platforms has become increasingly important. This review provides a comprehensive assessment of four major nanoparticle families—polymeric carriers (PLGA, chitosan, and micelles), lipid-based vehicles (liposomes, SLNs, and NLCs), inorganic systems (gold, mesoporous silica, iron oxide, and calcium phosphate), and hybrid composites—focusing on how their physicochemical properties govern drug encapsulation, release behavior, and tissue compatibility. These classes are then linked to specific implantable formats such as drug-eluting stents, nano-enabled scaffolds, and reservoir depots, and to wearable formats including transdermal patches, microneedle arrays, biosensor-coupled patches, and patient-actuated devices. A dedicated section addresses stimuli-responsive release driven by pH, enzymes, temperature, and electrical or magnetic fields, alongside closed-loop platforms that pair real-time biosensing with on-demand dosing. Surface engineering strategies, ligand targeting, antifouling coatings, antimicrobial layers, and immune-modulating chemistries are also discussed, together with the central translational hurdles: long-term stability, foreign body response, scale-up, sterilization, and regulatory classification of combination products. Finally, the review outlines near-term directions, including AI-driven dosing, 4D bioprinting, biomimetic nanocarriers, gene therapy delivery, and bioresorbable electronics, that together suggest where these hybrid platforms are likely to mature next. Full article

On-device plant disease perception must reconcile two competing demands: enough spatial detail to localise diseased tissue within a field image and a memory and compute budget small enough for microcontroller-class hardware. Single-network solutions that jointly learn a pixel-wise mask and a fine-grained disease label tend to oversubscribe both the Flash and activation SRAM of such devices. This paper presents Chroma-Sense 2.0, a two-stage lightweight pipeline that decouples the two subproblems and sizes each stage for its own budget. The two stages run sequentially on the same frame: Stage 1 is a per-channel convolutional classifier, derived from Chroma-Sense, that names the disease, and Stage 2 is a compact ESPNet segmenter that produces a binary diseased-versus-healthy mask localising it. Because the stages run one after the other rather than concurrently, the peak working memory of the pipeline is the maximum of the two stages rather than their sum. We evaluate the pipeline on the in-the-wild PlantSeg dataset using a curated 10-species, 34-class subset and a leakage-controlled protocol in which all training crops are derived from PlantSeg’s official training images and all reported metrics are measured on a held-out test set of 5002 crops built from the official test images. The segmentation stage attains a mean foreground recall of 0.97 (mean foreground IoU of 0.49; 0.53 pooled over pixels), a deliberately recall-oriented operating point. Against Fast-SCNN, a small U-Net, LR-ASPP, and DeepLabV3+, ESPNet is the smallest-footprint model (140k parameters, 193 KB Int8 Flash) while retaining the highest foreground recall; the per-channel classifier reaches accuracies comparable to much larger ImageNet-pretrained backbones (MobileNetV3 and EfficientNet) using 10–13× fewer parameters. End to end, the coupled pipeline classifies the disease correctly on 87.8% of the test crops. An on-device profile on the OpenMV H7 and H7 Plus shows that the binding constraint at 256 × 256 is the segmenter’s ≈4 MB contiguous activation arena, rather than parameter Flash: Even on the 32 MB-SDRAM H7 Plus, the usable interpreter heap is only about 4 MB, and the arena cannot be allocated as a single contiguous block from it, so on the tested firmware, the classifier runs on microcontrollers while the segmenter does not; the full pipeline instead fits the gigabyte-scale single-board-computer tier (for example, Raspberry Pi or NVIDIA Jetson Nano), and enabling the segmenter to run on microcontrollers remains the open gap. Full article

Lightweight real-time detectors must balance long-range contextual modeling, cross-scale relational reasoning, and deployment efficiency. We present MambaACE-YOLO, an integration-oriented framework that separates computation into intra-scale encoding, cross-scale relational reasoning, sparse multi-stage feature distribution, and multi-scale detection decoding. Its primary contribution is the coordinated integration of prior frequency-decoupled and state-space ideas with detection-specific interfaces, target-scale leave-one-source aggregation, and physically prunable feature distribution, rather than a new frequency-transform, state-space, or hypergraph mechanism class. Building on prior frequency-decoupled hybrid visual Mamba research, its detection-oriented D-MobileMamba backbone applies bidirectional state-space scanning only to low-frequency components, while directional high-frequency correction and multi-kernel depthwise convolutions preserve boundaries and local shape. Compact Partial-Channel HyperACE (CP-HyperACE) models cross-scale higher-order relations in a semantic subspace and uses target-scale leave-one-source aggregation. Selective Additive FullPAD (SA-FullPAD) projects each cross-scale increment once and selects injection paths through static, physically prunable gates. On MS COCO 2017 val, the unpruned MambaACE-YOLO-N achieves 42.5 AP with 2.6 M parameters, and MambaACE-YOLO-S achieves 48.8 AP with 9.0 M parameters. Under a common documented RTX 5090 TensorRT FP16 setting, the unpruned Nano model records 1.19 ms network-forward latency at 42.5 AP, whereas its physically pruned counterpart retains 42.4 AP and records 1.02 ms. Accuracy and latency values are single-run or single-record point estimates without reported variance, and the 42.5-AP result uses the 600-epoch schedule without a matched 600-epoch YOLOv13-N control. We distinguish published cross-paper results from same-framework, same-device measurements and assess the individual design choices through controlled ablations and physical-pruning experiments. Full article

Multimodal detection models enable flexible object detection through text prompts, but YOLO-World-style models still incur high computational and storage costs on edge devices. To address this problem, this paper develops an edge-oriented adaptation framework for replacing the original visual encoder of YOLOv8l-worldv2 with a compact visual branch. The framework jointly considers visual-branch compression, multi-scale interface consistency, cross-modal feature compatibility, and edge-side inference, rather than optimizing these aspects independently. Specifically, a YOLOv7-based visual branch is reconstructed using depthwise separable convolutions, enhanced by DyHead, and compressed through sensitivity-guided Filter Pruning via Geometric Median (FPGM) under multi-scale interface constraints. An identity-initialized semantic adaptation layer and multi-template text prototypes are then introduced to alleviate the feature distribution mismatch between the compressed visual branch and the original cross-modal fusion space. Finally, offline text prototype generation and TensorRT-based INT8/FP16 mixed-precision inference are used for edge deployment. Experiments on BDD100K show that the lightweight visual encoder achieves 62.45% mAP@0.5 with 22.5 M parameters and 52.8 GFLOPs. After multimodal integration, the proposed model achieves 64.9% mAP@0.5 with 31.5 M parameters and 67.2 GFLOPs, reducing GFLOPs by 67.1% compared with YOLOv8l-worldv2 while causing only a 1.6 percentage-point accuracy drop. On the Jetson Orin Nano Super, the deployed model reaches 32.6 FPS with a model size of 32 MB, demonstrating its feasibility for edge perception scenarios. Full article

The ultra-wide bandgap semiconductor β-Ga₂O₃ is a promising material for next-generation optoelectronic systems and hybrid nanodevices. However, high interface state densities and anomalous trap-assisted leakage severely restrict its performance and signal transduction capabilities. To resolve these fundamental limitations, we investigated a two-dimensional h-BN interlayer to construct a high-quality heterogeneous metal/h-BN/β-Ga₂O₃ structure using experimentally calibrated Sentaurus TCAD simulations. Energy-band analysis and validated I–V simulations reveal that the low-dimensional h-BN interlayer reconstructs the interfacial barrier, suppresses interface-assisted recombination, and shifts the dominant carrier transport from thermionic emission to Fowler–Nordheim tunneling. These effects markedly reduce the interface-state density and effectively suppress the Shockley–Read–Hall recombination current, mechanisms that are critical for minimizing dark current and improving device sensitivity. After systematically examining the effects of key parameters on the electrical characteristics of this hybrid architecture, we quantify the tradeoff between threshold voltage and on-resistance using a comprehensive figure of merit. Specifically, our results indicate that maximum device efficiency is achieved only when an optimal h-BN thickness of 3.56–5.88 nm (10–17 atomic layers) is strategically integrated with the appropriate metal work function and semiconductor doping. Overall, this work suggests the potential advantage of 2D h-BN in mitigating the interfacial bottleneck of traditional β-Ga₂O₃ platforms, providing quantitative design guidelines and theoretical support for the heterogeneous integration of next-generation optoelectronic devices. Full article

Additive manufacturing [AM; three-dimensional (3D) printing] is reshaping heterogeneous photocatalysis by enabling architected materials and devices that span from centimeter-scale monoliths to micro- and lab-on-chip reactors. Heterogeneous photocatalysts have traditionally been deployed as powders or simple coatings, which offer high activity but suffer from mass-transfer limitations, poor light utilization and difficult recovery. Three-dimensional (3D) printing now allows precise control over macroscopic geometry, internal channel networks and micro-/nano-scale surface texturing, creating structured photocatalysts and microreactors that can be tailored for specific photon and flow fields. In contrast to recent reviews that primarily survey materials development or additive-manufacturing routes, this work focuses on photocatalytic microreactors and lab-on-chip devices as multi-scale reactors in which catalyst composition, architected geometry, photon management and hydrodynamics are co-designed across length scales. We summarize three-dimensional 3D-printed photocatalytic systems based on polymer–oxide composites, ceramic scaffolds such as zinc oxide (ZnO)/titanium dioxide (TiO₂) clay monoliths, and laser-written titanium dioxide (TiO₂) nano-architectures, with particular emphasis on microfluidic and lab-on-chip implementations fabricated by fused deposition modeling (FDM), direct ink writing (DIW), stereolithography (SLA), digital light processing (DLP) and laser direct writing (LDW). Based on the literature data and representative case studies, we examine how architected lattices, sponges and microreactor chips affect key performance metrics—apparent rate constants, apparent quantum yield (AQY) and space–time yield (STY)—for the degradation of dyes, antibiotics, detergents and other emerging contaminants in realistic matrices, and we compile reported values to illustrate emerging performance trends and limitations. Representative case studies highlight 3D-printed manganese-doped zinc oxide (Mn:ZnO)-decorated sponges used as modular cartridges for greywater and detergent treatment, as well as laser-written titanium dioxide (TiO₂) nano-photocatalysts integrated into microchannels to couple structured light fields with controlled residence times. Finally, we outline materials and process challenges—including ultraviolet (UV) aging of polymer supports, the energy intensity of ceramic sintering and the lack of standardized testing protocols—and identify future research directions formulti-scalee modeling and techno-economic evaluation of three-dimensional (3D)-printed photocatalytic microreactors and devices. Full article

Touchless interfaces for embedded systems require models that are both compact and evaluated under protocols that reflect realistic generalization. Public low-resolution gesture datasets can obscure this requirement because repeated samples, temporal neighbors, or subject identity may leak across train and test partitions. This study presents a leakage-aware multi-corpus benchmark in which four public corpora are evaluated within their own dataset-specific partitions after conversion to a common 8 × 8 depth representation when needed: a public IEEE DataPort ToF CSV with 8400 frames from four gestures, DS.GestureRecognition_TOF.1.0 with 3109 frames extracted from 40 low-resolution ToF sequences and three gestures, the depth component of the public Kinect+Leap dataset with 1400 samples from 14 subjects and 10 gesture classes, and the public Senz3D depth archive with 1320 samples from 4 subjects and 11 gesture classes. No blurred RGB image dataset, deblurring target, or synthetic blur augmentation is produced in the experiments; the only spatial reduction applied to external depth corpora is deterministic block averaging for a shared low-resolution depth input. The term multi-corpus is used here to clarify scope: the experiments do not train on one dataset and test on another, but instead compare optimistic random splits against structure-aware alternatives matched to each corpus: chronological block-aware, sequence-aware, and subject-wise evaluation. On the original IEEE ToF benchmark, the best macro F1-score decreases from 1.0000 under random splitting to 0.8494 under block-aware evaluation. On DS.GestureRecognition_TOF.1.0, random splitting again reaches 1.0000, whereas the best sequence-aware result falls to 0.6697. On the downsampled Kinect+Leap depth benchmark, random frame classification reaches 0.7639 macro F1, while subject-wise evaluation drops to 0.0566. On Senz3D reduced to 8 × 8, random frame classification reaches 0.9068 macro F1, while subject-wise evaluation drops to 0.0717. Device-side validation on an NVIDIA Jetson Nano Developer Kit preserved a macro F1-score of 0.8667 for the selected GRU on the primary IEEE ToF dataset with 14.48 ms/window steady-state CPU latency. The results show that low-resolution gesture recognition conclusions are strongly protocol- and corpus-dependent: compact temporal models remain viable for genuine ToF streams, but naive random splitting can substantially overstate generalization, and subject-wise recognition after aggressive 8 × 8 conversion remains difficult for the evaluated datasets and architectures. Full article

Piezoelectric flexure micro-positioning stages are essential micromotion actuators for micro-assembly, atomic force microscopy and nano-manufacturing, but intrinsic hysteresis nonlinearity of piezoelectric stacks distorts the linear voltage-to-displacement mapping and induces significant micro-positioning errors. Conventional hysteresis compensation based on offline-calibrated Krasnosel’skii-Pokrovskii (KP) models cannot adapt to time-varying excitation, whereas state-of-the-art adaptive KP control requires auxiliary dynamic equations and imposes high computational overhead on miniature real-time controllers. To address these limitations, this paper develops a single-degree-of-freedom micromotion positioning device equipped with symmetric two-stage displacement amplification mechanisms and straight circular flexure hinges. ANSYS finite element simulations validate the mechanical stiffness, structural safety and linear amplification characteristic of the micro-positioning stage, achieving a maximum output stroke of 95.95 μm. A discretized KP hysteresis model is constructed to accurately capture the asymmetric rate-dependent hysteresis of piezoelectric stacks. On this basis, a lightweight adaptive inverse control framework is proposed, which realizes online tuning of KP weights through gradient descent iteration only relying on real-time position feedback, eliminating static pre-calibration and extra dynamic correction links. Tracking experiments under 0.1–2 Hz sinusoidal waveforms and 3–7 V variable-amplitude sinusoidal waveforms are implemented. Experimental results show that the proposed approach reduces the root-mean-square error (RMSE) by 7.41–85.65% and the mean absolute percentage error (MAPE) by 7.56–87.81% compared with uncompensated open-loop micromotion control. The combined micro-flexure mechanical design and adaptive hysteresis compensation strategy greatly improves positioning accuracy and anti-interference capacity, offering a low-computation technical route for high-performance micro-positioning systems. Full article

The electrocatalytic performance of transition metal dichalcogenides (TMDs) is often hindered by their basal planes. Although defect engineering and heteroatom doping are employed, improving the electrocatalytic activity of the basal plane without disrupting the in-plane lattice structure remains a challenge. The interlayer twist serves as an effective strategy to preserve the in-plane lattice integrity of TMDs and tune their electrocatalytic performance. In this work, high-quality twisted bilayer MoS₂ (TBL-MoS₂) was fabricated through polydimethylsiloxane (PDMS)-assisted mechanical exfoliation and dry transfer stacking. On-chip electrochemical nanodevices were subsequently constructed to investigate hydrogen evolution reaction (HER) performance at different twist angles. This demonstrates that bilayer MoS₂ with a twist angle of 15° (TBL-15°) exhibits superior hydrogen evolution reaction performance with a Tafel slope of 63 mV/dec, significantly lower than the value of 202 mV/dec of the pristine bilayer. This work paves the way for twist angle engineering in designing high-performance 2D electrocatalysts. Full article

Nanoplasmonic waveguides can efficiently manipulate the propagation characteristics of chiral light and hold great promise for integrated nano optics and on-chip optical information processing. Previous studies have demonstrated that gap plasmon structures composed of gold nanorod and silver nanowire can break the propagation symmetry of chiral light, thereby enabling asymmetric directional propagation. However, there remains considerable scope for enhancing the directional selectivity of these structures. In this work, we systematically investigate the mechanism of how the geometrical parameters of nanostructures regulate the directional propagation of chiral light. Based on the finite-difference time-domain method, the propagation behavior and evolution of directionality of chiral light in nanoplasmonic waveguides are analyzed in detail by changing the morphology and length of gold nanorods. The results show that the geometrically optimized nanostructures can significantly enhance the stability of directional chiral light propagation. At the same light-source position, the directionality of the gold nanorod with tips is approximately 56%, whereas that of the gold nanorod without tips remains approximately 89%. Further investigations show that structural modification of the silver nanowire or geometrical optimization of the structure can effectively reshape the local electromagnetic field distribution, enabling precise control over the propagation direction of chiral light. This work elucidates the physical mechanism underlying geometry-controlled chiral plasmonic propagation and provides a new design strategy for the structural design and performance optimization of high-performance chiral nanophotonic devices, with promising potential applications in chiral optical recognition, on-chip optical information processing, and nanoscale optical communication. Full article

Micro-Newton cold-gas thrusters are promising actuators for precision space missions, but their performance is strongly influenced by the integrated head architecture. This study presents the design, fabrication, and experimental validation of a piezoelectrically controlled cold-gas microthruster head for space-based gravitational-wave detection missions. The proposed head integrates a cone-needle throttle, a micro-orifice interface, and a downstream micro-nozzle, thereby converting actuator displacement into a regulated mass flow and ultimately into thrust. One-dimensional theory was first used for preliminary sizing, and Direct Simulation Monte Carlo (DSMC) analysis of the complete throttle-nozzle geometry was then applied to determine the final design parameters under rarefied-flow conditions. The selected design uses a throat radius of 29 μm and a needle half-angle of 10 degrees. Following fabrication and structural characterization, the integrated device was validated through mass-flow calibration and vacuum thrust testing. The experimental results show that the pressure-decay-based calibration provides a consistent mapping between actuation command, calibrated flow rate, and thrust output. The measured flow–thrust relation preserves the high linearity predicted by simulation, while the experimentally evaluated specific impulse meets the specified design target over the tested range. In addition, thrust-resolution testing at a baseline thrust of approximately 98.4 micro-Newton demonstrates a minimum resolvable step of 50 nano-Newton, and the measured thrust-noise amplitude spectral density remains below 0.07 micro-Newton/sqrt(Hz) over the 10 mHz–1 Hz band for the tested thrust levels. These results support the feasibility of the proposed integrated cold-gas microthruster head and its device-level validation approach for future space-based gravitational-wave detection applications. Full article

Nose-to-brain (N2B) delivery is a practical, non-invasive strategy for CNS targeting that can increase brain exposure while limiting systemic exposure. This review integrates three milestones in N2B delivery, formulations, devices, and quantitative evaluation strategies, to define design rules for effective olfactory/trigeminal deposition and enhance translational relevance. Formulations emphasize mucoadhesive systems, nanoparticle carriers (polymeric, lipid-based, and hybrid), nano-emulsions, and stimuli-responsive “smart” gels that prolong nasal residence. Regarding device advancements, the review covers conventional nasal sprays optimized for plume geometry and droplet size. Furthermore, it examines breath-actuated metered sprays, which promote soft palate closure to route aerosols to superior regions, and vibrating mesh nebulizers capable of low-velocity mists for improved upper cavity deposition. Quantitative evaluation is discussed, including 3D-printed, anatomy-accurate nasal casts, high-speed spray diagnostics, and computational fluid dynamics (CFD). This review further links formulation and device parameters to regional deposition. Available clinical and animal data illustrate the feasibility of these approaches, safety considerations, and user-technique dependencies, while highlighting the need for standardized, anatomy-aware testing protocols. Together, these developments suggest that co-designed formulation device platforms, validated by cast/CFD metrics and supported by clinical imaging or pharmacokinetic data, can support N2B product development toward consistent, patient-relevant outcomes. Full article