Search Results (265)

The sensitivity of classical electrochemical biosensors is constrained by noise processes at the electrode–electrolyte interface: low-frequency 1/f noise, thermal noise and capacitance fluctuations degrade the signal-to-noise ratio in ways that circuit-level mitigation reduces but does not remove. Chirality-Induced Spin Selectivity (CISS) has been proposed as a route past that limit, by shifting transduction from the scalar quantity of charge to the vector property of electron spin. Spin polarizations of up to approximately 60% have been reported at room temperature for double-stranded DNA monolayers in spin-resolved photoemission, while spin-dependent electrochemistry on smaller chiral adsorbates typically yields values in the range of about 5–30%; the reported magnitude is therefore system-, geometry-, technique- and analysis-dependent rather than a universal property of biological helices. Analyte binding modulates this efficiency through changes in helical pitch, dipole and structural integrity. This review unites the physics of CISS with the surface chemistry of spin-selective sensor layers, compares the competing mechanistic accounts of the effect, and then examines a persistent quantitative gap: the polarizations obtained from first-principles transport calculations on isolated chiral molecules remain well below the measured values. Non-relativistic, spin-restricted calculations on closed-shell helices in vacuum yield no polarization by construction, and although spin-polarized and relativistic implementations that treat spin–orbit coupling explicitly are available, they typically still underestimate experiments by orders of magnitude. We argue that a substantial part of this deficit is attributable to the widespread use of static, vacuum-based or implicitly solvated models, and that multiscale quantum mechanics/molecular mechanics (QM/MM) frameworks with explicit solvents are one necessary—though probably not sufficient—correction. Full article

Nighttime driving safety remains a critical challenge in modern transportation systems: insufficient ambient lighting significantly degrades visual perception quality, adversely affecting both human drivers and advanced driver-assistance systems (ADAS) and directly threatening road users’ safety. Traditional image enhancement methods often suffer from color distortion and visual artifacts, whereas existing deep learning approaches typically require paired training data and incur substantial computational overhead. To address these limitations, this paper presents BLEN (bio-inspired low-light enhancement network), a zero-reference deep learning framework that integrates biological vision principles with efficient convolutional architectures. Specifically, BLEN leverages Retinex theory for illumination–reflectance decomposition, is inspired by and functionally approximates lateral inhibition mechanisms for edge enhancement, and incorporates a Large-Kernel Convolution with Attention (LKCA) module that reduces the parameter count of the LKCA encoder block by 76% (0.56 M vs. 2.34 M for a standard 13 × 13 convolution) relative to standard large-kernel operations. Extensive experiments on the SICE and LOL benchmarks demonstrate that BLEN achieves state-of-the-art performance among real-time, edge-deployable zero-reference methods on the SICE benchmark, yielding a peak signal-to-noise ratio (PSNR) of 23.67 ± 0.14 dB and a structural similarity index measure (SSIM) of 0.891 ± 0.004 on SICE while maintaining 2.10 M parameters (2.1 MB in INT8, 8.4 MB in FP32). Furthermore, the proposed method enables real-time inference at 31 frames per second (FPS) on embedded platforms, including the HiSilicon SS928 and Jetson Nano, demonstrating that the proposed method is an efficient and effective front-end for camera-based ADAS perception on automotive-grade edge hardware. Full article

(1) Background: Evoked potentials (EPs) are an elegant, non-invasive, and reliable technique for assessing the functional integrity of neural pathways. They are, however, often limited by the difficulty and time needed to consistently distinguish signal from noise. (2) Methods: We have recently proposed a novel technique based on spectral domain evoked-to-background ratio (EBR) that enables fast data acquisition with online feedback about actual signal quality utilizing state of the art analog-to-digital conversion. Furthermore, we have developed a novel model-based signal-to-noise management concept allowing for suppression of biological and technical interference (swallowing, stimulation artifacts, electropolarization, and powerline potentials) and for online assessment of signal-to-noise ratio (SNR) for EPs. In this work, we experimentally confirmed this concept in ten healthy volunteers by investigating cortical EPs and high-frequency oscillations (HFOs) following median nerve stimulation. (3) Results: Both mathematical model and human data demonstrate that spectral target-band EBR governs the progress in SNR with increasing sweep count. For cortical EPs, SNR exceeded 10 dB beyond 90 averages in all participants. An SNR > 20 dB documented excellent signal quality and reproducibility. For HFOs, the SNR shifted to lower values by 12 dB, displaying pronounced individual variation, however, with smaller variation of HFO-band background activity (1.9 vs. 7.6 dB between the 25% and 75% percentile). Thus, individual HFO responses are more important for actual signal extraction compared to background activity. In subjects displaying a high HFO amplitude, reproducibility was confirmed for less than 1000 sweeps. (4) Conclusions: The present investigations confirm that individual EBR is the major factor defining SNR. Background noise can be reduced to a negligible level. Online assessment of background activity will allow for the most accurate moment-to-moment visualization of raw signal quality. This will facilitate termination of the data acquisition and may be based on quantified signal quality rather than predefined sweep count. Full article

Endpoint detection is a key preprocessing step for passive acoustic monitoring of marine biological sounds, but underwater noise can obscure onset and offset boundaries, especially at low signal-to-noise ratios (SNRs). This study proposes a wavelet-domain adaptive composite thresholding (WD-ACT) method for robust endpoint detection. The method applies Daubechies 4 (Db4) wavelet enhancement with soft-thresholding, then fuses short-time energy, non-zero-lag autocorrelation, and spectral variation into a frame-level composite detection score. An adaptive dual-threshold strategy estimated from background-frame statistics is used to determine the onset and offset of effective acoustic segments. Experiments were conducted on three manually annotated marine mammal vocalization segments, including seal, dolphin, and walrus calls, mixed with white noise, pink noise, ship noise, and wind–wave noise at SNRs from $-5$ to 15 dB. For each combination of acoustic segment, noise type, and SNR level, 100 noise-mixing trials were performed. Compared with short-time energy and zero-crossing rate (STE-ZCR), time-domain autocorrelation detection (TDA), wavelet-domain detection (WD), and spectral-flux detection (SFD), WD-ACT achieved the lowest median missed detection rate (5.56%), the lowest median boundary mean absolute error (222.49 ms), and the highest median frame-level accuracy (87.17%). An additional generalization evaluation using nine independent vocalization segments from five further marine mammal categories provided additional evidence of cross-signal applicability without species-specific parameter tuning. These results indicate that WD-ACT provides a robust and interpretable preprocessing method for marine bioacoustic endpoint detection. Full article

Radiomics features are strongly sensitive to image acquisition, and separating that sensitivity from biological signal usually requires repeated patient scans that cannot be shared. We present an open, fully synthetic framework (radiomics-phantom) that maps and, as a proof of concept, corrects radiomics feature instability without any patient data. Deterministic three-dimensional texture phantoms are generated as anisotropic Gaussian random fields with known ground truth and an optional embedded lesion. An independently implemented feature core aligned with the Image Biomarker Standardization Initiative (IBSI) covers all eleven IBSI-1 feature families and matched all 482 published digital-phantom benchmark values within the applicable tolerances. An image-domain acquisition simulator applies point-spread blur, slice-profile averaging, dose-scaled correlated noise, resampling, and quantisation. Per-feature reproducibility across a sweep of fifteen textures (varying correlation length, anisotropy, and intensity scale) by nine acquisition conditions, with five independent noise realisations per stochastic setting, is summarised by the absolute-agreement intraclass correlation ICC(2,1), with a realisation-aware percentile-bootstrap 95% confidence interval for every estimate; constant features are excluded from estimation. Values span nearly the full range (median 0.13, 95% CI 0.03–0.19), and a hierarchical variance decomposition attributes a median 77% of per-feature variance to the acquisition condition and under 1% to stochastic realisation; the values are interpreted as exploratory rankings within this acquisition envelope. As a proof of concept, intensity variance and grey-level co-occurrence contrast under additive Gaussian noise were normalised using calibrated, invertible response models, returning them to their noiseless values on held-out data (median error below 4% across five textures and repeated noise realisations, and about 11% when the noise level is estimated from the degraded image itself), while features the models cannot describe are refused rather than corrected. All code and a 716-test suite are released openly and archived on Zenodo. The result is a reproducible, patient-data-free testbed for radiomics feature stability. Full article

Raman spectroscopy is a widely used analytical technique; however, the choice of container material can significantly affect repeatability due to noise and unwanted Raman bands. This paper describes the interaction of excitation radiation with the sample’s environment and the materials surrounding or containing it. Polytetrafluoroethylene (PTFE) and Barium Sulphate are analyzed as contact surfaces with the containers. We aim to investigate how different container materials affect the Raman spectrum and identify the optimal material to minimize measurement bias. We compared the Raman spectra of containers made of dissimilar materials, including plastic, glass, quartz, and aluminum, to assess their effects on the spectra, and we propose a container design that enhances Raman spectroscopy signals. We consider Raman bands from biological samples of glucose (in water) and urine (urea and creatinine), which serve as references for comparison in the in vitro analysis of these substances across different containers and backgrounds. Full article

Lower-limb rehabilitation exoskeletons have the potential to improve gait recovery after stroke by providing intensive and repetitive training. However, conventional control strategies often rely on fixed control parameters and exhibit limited adaptability to patient-specific characteristics, sensor noise, and dynamic uncertainties. This paper proposes an adaptive control framework that combines deep reinforcement learning (RL) with model-based impedance control for personalised lower-limb exoskeleton assistance. Patient-specific biological parameters are incorporated into the simulation environment and reward formulation to improve adaptability and robustness. Three state-of-the-art deep RL algorithms, Deep Deterministic Policy Gradient (DDPG), Twin Delayed Deep Deterministic Policy Gradient (TD3), and Soft Actor-Critic (SAC), are evaluated in a continuous control environment under varying signal-to-noise ratio (SNR) conditions ranging from 5 dB to noise-free conditions. Results demonstrate that TD3 achieves the most stable learning performance, obtaining a mean reward of −354.24 under noise-free conditions, while DDPG provides the highest joint-angle tracking accuracy with an RMSE of 0.0369 rad. SAC exhibits superior robustness in noisy environments, achieving the highest learning ratio of 0.51 at 5 dB SNR. Furthermore, the proposed personalised framework reduces tracking errors by up to 27% compared with non-personalised baseline approaches. The findings indicate that integrating patient-specific information with RL-based adaptive control can significantly enhance robustness, tracking performance, and personalisation in exoskeleton-assisted gait rehabilitation, providing a promising direction for future intelligent rehabilitation systems. Full article

Implantable neural probes that can simultaneously possess biocompatibility, electrochemical stability, and high signal fidelity are the core devices in neuroelectrophysiological research. In this article, AlGaN/GaN heterojunction field-effect transistors are used instead of traditional metal microelectrodes to prepare brain nerve probes. By photoelectrochemical etching and optimization of sensing area size, the probes have the maximum transconductance value, i.e., the highest sensitivity, under no gate control. After digital filtering processing, the neural probe achieved a signal-to-noise ratio of 8.04 dB on biological analog signals as low as 50 µV, confirming its ability to detect microvolt-level signals. The ex vivo recording of the bullfrog sciatic nerve further validated its biosensing performance, demonstrating the selective capture of composite action potentials from active neural tissue. Full article

Scaffold architecture shapes tissue regeneration through the mechanical, topographical, and biochemical cues it presents to cells, yet geometrically elaborate scaffolds do not reliably produce more organized tissues, while comparatively simple architectures can exert strong organizational effects. We argue that scaffold performance is better understood by distinguishing geometric complexity from effective informational dimensionality: a relational property of the scaffold–cell system, defined as the number of independently manipulated architectural directions that produce distinguishable, above-noise changes in a jointly measured mechanotransductive response. Unlike structural entropy, fractal dimension, or feature-counting metrics, this construct depends on cellular accessibility, cue persistence, and non-redundancy. Mechanotransduction supplies its biological basis, integrating scaffold-derived cues through focal adhesions, cytoskeletal organization, nuclear deformation, and YAP/TAZ signaling, and we distinguish early resolvability from later organizational stabilization. We outline an operational strategy for estimating both from factorial scaffold libraries, common readout panels, and rank-based analysis of the response mapping, illustrated with selected experimental precedents rather than a systematic evidence sample. Positioned relative to biomimetic, mechanobiology-guided, and morphospace approaches, it yields testable predictions on dimensional compression, redundancy, and the resolvability–stability dissociation. Scaffold design is thus reframed from maximizing complexity or native resemblance toward engineering stable, cell-readable dimensions of organization. Full article

Real-time tissue identification during laser-guided surgery is a critical unmet need for collateral damage avoidance and margin delineation. Laser-Induced Breakdown Spectroscopy (LIBS) is compatible with pulsed laser surgical systems and offers rapid, label-free elemental analysis. This study presents a machine learning pipeline classifying five ex vivo bovine tissue classes, plus one synthetic null-signal control class, from LIBS spectra, designed to control specimen-level data leakage and class imbalance bias. Key contributions are (i) a ‘peak max over baseline’ aggregation strategy suppressing shot noise while preserving emission peaks; (ii) a repeated, group-based cross-validation protocol (GroupShuffleSplit, N = 10) enforcing specimen-level separation; and (iii) a comparison of 30 configurations (10 classifiers × 3 pipelines). Extra Trees with normalization reached the highest weighted F1-score (0.934 ± 0.118); excluding the synthetic control, five-class scores fall to 0.875–0.915 and the ranking changes, so these are the reference figures for biological tissue discrimination. Support Vector Machines were less accurate but more consistent (0.917 ± 0.069). Acquisition takes approximately 3 s per point; inference is sub-millisecond. With five source animals, the best configuration chosen on the same outer splits, and inner tuning that was not group-aware, these estimates are an exploratory step toward real-time guided surgery. Full article

Living nervous systems sense the cardiovascular rhythm with a parsimony that engineered monitors cannot match: sensory receptors encode changes rather than absolute levels, neurons communicate through sparse all-or-none events, retinal circuits sharpen salient features through lateral inhibition, and attention is allocated by surprise. We translate these four principles into BioSpike-Net, a fully event-driven spiking neural network for cardiac-rhythm classification from electrocardiogram (ECG) and photoplethysmogram (PPG) signals. A sensory-adaptive spike encoder (SASE) converts analogue waveforms into ON/OFF spike trains through a mechanoreceptor-inspired gain-control law; adaptive-threshold leaky integrate-and-fire layers integrate these events; a lateral-inhibition spiking convolution emphasises locally salient morphology; and a novelty-gated temporal attention mechanism concentrates computation on the most surprising portions of each beat. Evaluated on the MIT-BIH Arrhythmia Database, PTB-XL, CPSC-2018, and a PhysioNet-derived PPG corpus, BioSpike-Net achieved 97.6 ± 0.3% accuracy and 95.8 ± 0.4% macro-F1 on MIT-BIH five-class arrhythmia classification, and 0.982 ROC-AUC on PPG atrial-fibrillation detection, matching or exceeding strong recurrent, convolutional, and transformer baselines while requiring an estimated 6.4 µJ per inference—approximately 27-fold below the transformer baseline—owing to a mean activation density below 0.10 spikes per neuron per time step. Ablations show that each biological principle contributes a measurable and interpretable accuracy-versus-energy benefit, and the network degrades gracefully under additive noise and motion artefact. By grounding architecture in the economy of biological sensing, this work offers a route to sustainable, always-on cardiac monitoring. Full article

Biosensors have become important analytical platforms that enable rapid, selective, sensitive and portable analysis for early disease diagnosis, biomarker monitoring and point-of-care diagnostic applications. Their analytical performance depends on the coordinated function of molecular recognition elements, surface chemistry, transduction mechanisms and signal-processing strategies. Nevertheless, the analysis of real biological samples remains challenging because of low target concentrations, matrix effects, interfering species, signal noise, sensor drift and device-to-device variability. Therefore, artificial intelligence and machine learning are gaining increasing importance as data-driven tools for signal preprocessing, calibration, feature extraction, pattern recognition, quantitative prediction and diagnostic decision support. These approaches are particularly valuable for interpreting complex datasets generated by electrochemical, optical, wearable and microfluidic biosensors. This review presents an overview of healthcare-oriented biosensor systems beginning with molecular recognition principles, bioreceptor design, and transduction technologies, and extending to applications in clinical diagnosis and health monitoring. It also examines the roles of supervised, unsupervised and deep learning approaches in biosensor data analysis, while critically discussing model validation, generalizability, interpretability and clinical translation. By linking molecular-level recognition with computational signal interpretation, this review highlights the advantages and limitations of artificial intelligence-integrated biosensors for next-generation point-of-care diagnostics, continuous health monitoring, and personalized healthcare applications. Full article

Background: Carotid plaque rupture is a critical event in ischemic stroke, yet the potential involvement of the intraplaque microbiota across disease stages remains unclear. Methods: We performed dual-omics profiling by analyzing host transcriptomes and PathSeq-derived microbiomes from 48 human carotid RNA-seq specimens spanning early lesions (intimal thickening; n = 10), stable plaques (n = 20), and unstable plaques (n = 18). Host transcriptomes were profiled alongside intraplaque microbiomes extracted via the GATK PathSeq pipeline with rigorous in silico decontamination. We integrated differential expression analysis, microbial diversity metrics, and functional inference. Furthermore, an integrated machine learning approach (incorporating Boruta feature selection) was employed to identify exploratory cross-kingdom diagnostic biomarkers. Results: Microbial beta diversity diverged significantly across disease stages, accompanied by the progressive upregulation of 54 host genes critical for extracellular matrix remodeling and immune chemotaxis. Strikingly, despite the inherent noise and artifacts associated with low-biomass sequencing, we computationally detected the distinct enrichment of 21 bacterial taxa in unstable plaques, predominantly oral and gut mucosal pathobionts. Computationally inferred functional profiling revealed that these unstable plaque-associated microbiota were significantly linked to predicted cell death, IL-17, and HIF-1 signaling pathways and exhibited strong positive correlations with host matrix-degrading transcripts. Statistical modeling suggested associative links among specific microbial enrichment, host transcriptomic dysregulation, and plaque instability, highlighting concurrent biological cross-talk. Importantly, our integrated machine learning pipeline established a 14-feature cross-kingdom biomarker panel (10 host genes and 4 bacteria) that discriminated stable from unstable plaques (cross-validated AUC = 0.869). Conclusions: Intraplaque microbiome dynamics computationally associate with host transcriptomic alterations during carotid plaque evolution. This synergistic host–microbiome association provides a hypothesis-generating framework linking microbial dysbiosis to plaque destabilization, offering novel mechanistic insights and highlighting the exploratory cross-kingdom biomarker panel as a highly promising foundation for future experimental validation and stage-tailored clinical diagnostics. Full article

Background: Low sequencing depth causes molecular capture loss and zero inflation in scRNA-seq data, reducing the reliability of downstream analyses. Methods: We propose DepthDiff, a depth-conditional expression enhancement model trained with diffusion-based denoising. Using low-depth expression profiles and sequencing depth ratio as conditions, DepthDiff learns a supervised residual mapping to paired high-depth references. During inference, it requires only a single-step forward prediction. Fixed-UMI downsampling was used to construct benchmarks across three public datasets. Results: DepthDiff outperformed supervised baselines, MAGIC, and unenhanced low-depth data in expression reconstruction and biological signal preservation. Ablation experiments showed that x0 prediction and cosine noise scheduling were important for stability, while full reverse diffusion sampling provided no additional benefit. Cross-dataset transfer and CITE-seq validation supported the generalizability and biological relevance of the recovered signals. Conclusions: DepthDiff is an efficient supervised framework for low-depth scRNA-seq expression enhancement, with gains mainly driven by diffusion-based denoising training rather than generative reverse sampling. Full article

Alzheimer’s disease (AD) is the most prevalent neurodegenerative disorder worldwide, affecting approximately 56.9 million people in 2021 and projected to reach 152 million by 2050. Its defining pathological features, amyloid-beta plaques and neurofibrillary tangles, accumulate for up to two decades before cognitive symptoms emerge, placing the preclinical and mild cognitive impairment (MCI) stages at the centre of the early detection problem. Despite this, current diagnostic practice in routine clinical settings remains unreliable, with post-mortem studies placing the specificity of clinical AD diagnosis between 44.3 and 70.8% even in specialist memory clinics. Machine learning (ML) and deep learning (DL) applied to neuroimaging and electrophysiological data have emerged as candidate tools for closing this diagnostic gap, yet whether the accuracy figures reported in published studies translate into clinically useful performance on independent data remains unresolved. This study presents a structured comparative review of machine learning and deep learning methods reported across four publicly available Alzheimer’s disease datasets, namely the Alzheimer’s Disease Neuroimaging Initiative (ADNI), the Open Access Series of Imaging Studies (OASIS), the OpenNeuro ds004504 electroencephalography (EEG) dataset, and the Kaggle Alzheimer’s magnetic resonance imaging (MRI) dataset. Thirteen model families are examined through the published literature rather than through new experiments, and for each model and dataset combination, the best accuracy reported in the source study is recorded alongside the model’s mathematical formulation. All performance figures reported in this abstract and throughout the paper are taken from the published studies reviewed, not from new experiments conducted by the authors. Across the reviewed studies, deep learning architectures pre-trained on ImageNet and fine-tuned on neuroimaging data are reported to produce the highest accuracy on MRI classification tasks. Residual Network (ResNet)-101 is reported at 98.21 percent on ADNI and 97.45 percent on OASIS, while the IncepRes fusion architecture reaches 98.35% on OASIS by combining multi-scale feature extraction from InceptionV3 with residual connectivity from ResNet152V2. Traditional machine learning classifiers remain competitive on tabular clinical and biomarker data, with Extreme Gradient Boosting (XGBoost) reaching 91% on ADNI multiclass features. Logistic Regression achieves 82 to 85% on binary ADNI classification and is the only classifier in this review that provides explicit per-feature prediction contributions without post hoc tooling. Gaussian Naïve Bayes achieves 80 to 83% on the same task. On the OpenNeuro EEG dataset, K-nearest neighbours (KNN) with singular value decomposition (SVD) entropy features achieves 91% binary accuracy, with feature engineering quality determining performance more reliably than classifier architecture. Eight principal findings emerge from the cross-dataset analysis. Binary classification consistently outperforms multiclass by 10 to 30% across all datasets, reflecting the genuine biological ambiguity of the mild cognitive impairment category. Dataset size and augmentation predict reported accuracy more reliably than model architecture. Ensemble methods outperform individual classifiers by 5 to 8% in nearly every imaging study. Deeper architectures can overfit small clinical cohorts. EEG models trail MRI models by approximately 10 to 15% on comparable binary tasks. Cross-dataset generalisation has not been systematically evaluated in most studies, and the few that have tested it report accuracy drops of 5 to 10% or more when models encounter data from different scanners or cohorts. Eight recurring limitations constrain the clinical utility of these findings. Small sample sizes and limited demographic diversity, severe class imbalance inflating raw accuracy metrics, poor cross-dataset generalisation driven by scanner heterogeneity, limited deep learning interpretability, the dominance of binary over multiclass tasks, the absence of longitudinal modelling despite available datasets, inadequate standardisation of preprocessing and evaluation protocols, and the signal-to-noise ratio constraints specific to EEG recordings of elderly patients collectively define the gap between benchmark performance and clinical readiness. Future work must prioritise multi-centre training cohorts, multimodal fusion architectures, longitudinal progression modelling, and standardised interpretability evaluation as non-optional requirements for any system intended for clinical deployment. Full article