Search Results (401)

Tool wear degrades machining quality and, if unchecked, can progress to breakage, causing workpiece defects, downtime, and spindle damage; automatic tool condition monitoring is therefore essential. In real production, new and reground tools are used interchangeably and differ slightly in geometry and material, so the “normal” baseline shifts from one tool and mounting to the next. This makes both global-baseline diagnostics and deep-learning methods that require large labeled fault datasets difficult to apply. This study proposes a lightweight, self-referencing framework—whose per-tool baseline is built from acceptable-state data alone—for diagnosing milling tool wear under tool-to-tool variability. Its novelty lies not in the individual techniques—self-referencing, order tracking, and the Mahalanobis distance, which are established—but in their integration into a single framework, designed for fault-label-free operation, that rebuilds a dedicated baseline for every newly mounted tool. Whenever a tool is mounted, its own acceptable data, a short initial segment of machining taken as healthy immediately after mounting, form the baseline; kinematics-based order-tracked features from a single spindle-bearing accelerometer are used to compute the Mahalanobis distance from the acceptable state, which serves as a continuous health index. A warning limit set statistically from the acceptable data alone, together with a defect limit set as a pragmatic engineering multiple of it, separates the acceptable, warning, and defect grades. On four end mills of identical specification, the primary full-baseline analysis yielded a warning-detection AUC of 0.986 and a defect-detection AUC of 0.936; with a persistence rule, defect-grade wear was detected in all four tools with no false alarms in the acceptable state. Because the baseline is built from only a short acceptable segment and the computation is inexpensive, the framework is, in principle, suited to shop floors with frequent tool changes and to on-machine or edge deployment (not yet benchmarked on an edge device); the present validation, however, is limited to four tools and a single workpiece material under fixed cutting conditions with accelerated wear, so broader verification remains necessary. Full article

As a noninvasive liquid biopsy approach, extracellular vesicle (EV)-based detection offers significant advantages in reflecting real-time tumor dynamis and overcoming the limitations of conventional tissue biopsy. EVs, nanoscale vesicles secreted by cells, carry diverse biomolecules such as proteins and nucleic acids, playing key roles in tumor progression, metastasis, and immune evasion, and have emerged as promising biomarkers for cancer liquid biopsy. Surface-enhanced Raman spectroscopy (SERS), characterized by high sensitivity, resistance to photobleaching, minimal sample consumption, and multiplexing capability, has shown great potential in EV analysis. This review systematically summarizes current methods for EV isolation, characterization, and storage, with a focus on label-free and label-based SERS detection strategies for early cancer diagnosis, treatment response monitoring, and prognosis evaluation. Furthermore, the integration of SERS with machine learning and deep learning algorithms has substantially improved diagnostic accuracy and cancer subtyping. Despite remaining challenges, such as optimization of SERS substrate performance, intelligent processing of Raman spectral fingerprints, and clinical translation, EV-based SERS technology holds great promise for precision oncology. Full article

Cacao swollen shoot disease (CSSD) remains a major viral threat to cacao production in West Africa and is associated with a genetically diverse complex of badnaviruses. We combined alignment-free whole-genome distances with circular sliding-window mosaic barcodes to characterize global divergence and local compositional modularity across 48 full-length CSSD-associated badnavirus genomes, hereafter termed the cacao swollen shoot virus (CSSV) complex. Whole-genome tetranucleotide cosine distance clusters matched published species assignments exactly (adjusted Rand index = 1.00; normalized mutual information = 1.00; silhouette = 0.703) and were strongly concordant with an alignment-based distance derived from open reading frame 3 (ORF3) proteins (Spearman ρ ≈ 0.951; permutation p ≈ 0.0002). A three-component mosaic-complexity score summarized low dominant-label purity, barcode entropy normalized by log₂(K), and circular switch rate; ORF–mosaic agreement was retained as a separate diagnostic and did not contribute to the ranking. Threshold sensitivity showed that barcode switchpoints were enriched within ±100 and ±200 bp of predicted ORF boundaries, but not within ±400 bp. The ten highest-scoring genomes showed slightly higher mean local nucleotide entropy but lower entropy variance and dispersion than the remaining genomes, indicating more uniformly distributed compositional complexity rather than isolated local spikes. These analyses provide a transparent framework for post-sequencing classification, comparison, and prioritization of complete viral genomes. The barcode and score outputs are exploratory summaries and are not direct field diagnostic assays or nucleotide-resolution recombination tests. Full article

Alzheimer’s disease (AD) classification from structural magnetic resonance imaging (MRI) may benefit from weak supervision that uses clinically meaningful but imperfect supervisory signals. We evaluated a weakly supervised framework in which a multilayer perceptron (MLP) trained on age, sex, and Mini-Mental State Examination (MMSE) scores generated clinical pseudo-labels to initialize a patch-based fully convolutional network (FCN). For 260 Alzheimer’s Disease Neuroimaging Initiative (ADNI) training participants, subsequent refinement combined 80% of the preceding MRI-model probability with 20% of the participant’s ground-truth diagnostic label. This design preserves a dominant pseudo-label/self-training component while using partial diagnostic guidance to stabilize refinement. The FCN generated whole-brain probability maps, and selected voxel probabilities were classified by a second MLP. The framework was developed using ADNI (n = 417). Using ADNI validation data only, iteration 3 and a classification threshold of 0.5 were selected and then applied unchanged to the held-out ADNI test set and the external AIBL (n = 182), FHS (n = 102), and NACC (n = 265) cohorts. The selected model achieved F1 scores of 0.853 in ADNI, 0.707 in AIBL, 0.765 in FHS, and 0.807 in NACC. These results support the feasibility and cross-cohort transferability of clinical pseudo-label-based weak supervision for MRI classification. The framework is not intended to be label-free; rather, it provides a transparent strategy for integrating imperfect clinical pseudo-labels with partially weighted diagnostic guidance during training. Full article

We set out to fix a “contamination” problem in reconstruction-based Temporal Convolutional Network VAEs (TCN-VAEs) for cross-dataset IoT flow anomaly detection: when attack flows share an encoder window with benign flows, the shared latent code is allegedly distorted, inflating benign reconstruction error and producing false positive rates (FPRs) of 22–65% despite an ROC-AUC above 0.93. Our proposed fix, TCN-Pred, excludes the target flow from the encoder and scores it by next-flow prediction error, reducing FPR to 0.65–13%. We subjected this causal explanation to a battery of controlled ablations, holding architecture, decoder, loss, and thresholding fixed while varying one factor at a time. Each one falsified the original hypothesis: target inclusion/masking changes FPR by at most 0.001; context shuffling/reversing/zeroing changes it by at most 0.003; a context-blind constant-output predictor matches TCN-Pred’s FPR and F1 to three decimal places on all three datasets. The actual cause, confirmed on the original trained models with no retraining, is a scoring-granularity mismatch: the TCN-VAE threshold is calibrated from per-window errors averaged over 20 flows but applied to per-flow errors at evaluation (standard deviation $\approx \sqrt{20}\times$ higher, measured ratio 4.46 against a predicted 4.47). Recalibrating the identical model at matching granularity drops FPR from 22.7/47.6/64.6% to 0.65/5.0/12.5% on BoT-IoT, IoT-23 and ToN-IoT, closing 89–97% of the reported FPR gap without changing a single model weight. We report this diagnostic chain, together with an attack-prevalence sensitivity analysis, sample-disjoint calibration, normality diagnostics, and label-free and redundancy-aware (mRMR) feature-selection benchmarks, as a methodology other work should apply before attributing fixed-threshold performance to architecture. The pipeline is supervised source-domain feature selection followed by benign-only detector training, not fully unsupervised, a distinction we quantify later in the paper. Investigating dataset representativeness, we found that all three provided files reduce to only ≈6000 genuinely distinct flows via an undocumented row-duplication procedure, causing 97.8% BoT-IoT train/test near-duplicate overlap; a leakage-free re-evaluation changes FPR by only 0.23 percentage points. We also found that the TLS-metadata columns are already transformed upstream of every available artefact, so the proportion of genuinely TLS-encrypted flows cannot be recovered, and we soften the paper’s encrypted-traffic framing accordingly. Full article

Background/Objectives: OrganoHIPEC is a phase-II, two-stage, open-label clinical trial that investigates if cytoreductive surgery (CRS) and patient-tailored HIPEC, based on a preclinical platform using patient-derived organoids, can improve disease control in peritoneal metastases from colorectal cancer (CRC-PM). Methods: Adults with limited CRC-PM and no distant metastases were included. CRC-PM were sampled for organoid development during diagnostic laparoscopy. These organoids were used in an in vitro HIPEC model to test various drugs suitable for intraperitoneal administration. After 3–6 months of systemic chemotherapy, patients without progression underwent CRS/HIPEC with personalized regimens based on organoid drug response. To detect an increase in 12-month peritoneal disease-free survival from 40% to 60%, 24 patients are needed. According to the two-stage design, if <7 of 10 patients in Stage-1 remain PM-free at 12 months, the trial is terminated. Results: Forty-seven patients were enrolled. Among 31 patients with available organoid data, the most active drugs were mitomycin-C (n = 14), cisplatin/mitomycin-C (n = 12), and low-dose (120 min) oxaliplatin (n = 4). No patient was sensitive to high-dose oxaliplatin (30 min) and cisplatin/doxorubicin. Ten patients had a potential follow-up >12 months. Peritoneal relapse occurred at 8 months in two patients, and one died of liver metastases at 7 months. Seven patients remained PM-free for >12 months (median 16.4, range 12.6–28.4). Conclusions: A comprehensive precision approach using patient-derived organoids to guide personalized HIPEC is feasible and shows promising early results. High-dose oxaliplatin is poorly active. As 7/10 patients achieved the endpoint of 12-month PM-free survival, Stage-1 was successfully completed. The trial is proceeding to Stage-2. Full article

Macrophages are highly plastic innate immune cells that integrate diverse microenvironmental cues to adopt pro-inflammatory (M1) or anti-inflammatory (M2) functional states, which critically influence the pathogenesis of infectious, autoimmune, inflammatory, and malignant diseases. This review provides an overview of current concepts of macrophage ontogeny, functional heterogeneity, and disease-associated phenotypes, with a specific focus on experimental approaches used as M1/M2 test systems for macrophage phenotyping. The scope of the review encompasses commonly used experimental models, induction protocols for M1- and M2-like polarization, and key readouts, including gene-expression signatures and metabolic parameters. Particular emphasis is placed on reporter-based platforms (luciferase, BRET, fluorescent nanoparticle probes), label-free biophysical methods such as electrical impedance monitoring and metabolic profiling, and their application to dynamic, real-time assessment of macrophage phenotype in the context of tumor microenvironments and chemotherapeutic exposure. The potential of integrating reporter systems with single-cell omics, spatial transcriptomics, and patient-derived ex vivo platforms is considered, with a view to transforming M1/M2 test systems into clinically oriented assays capable of tracking macrophage programs during therapy and supporting the development of macrophage-targeted diagnostics and treatments. Full article

Conventional histopathological diagnostics rely on morphological assessment of stained tissue sections, requiring extensive sample preparation and subjective expert interpretation. Mid-infrared (MIR) imaging offers a complementary approach by providing spatially resolved, label-free access to the intrinsic biochemical composition of tissue without exogenous contrast agents. This work introduces a preprocessing and analysis pipeline for multiphotometric MIR data, applied to formalin-fixed, paraffin-embedded tissue sections from two patients with histopathologically confirmed head and neck squamous cell carcinoma. Combining differential scattering correction, sub-pixel channel registration, and automated tissue segmentation with unsupervised K-Means clustering, the pipeline achieves label-free discrimination of biochemically distinct tissue compartments. K-Means clustering identified four distinct clusters, of which three corresponded to tissue compartments with protein-to-lipid ratios tentatively consistent with epithelial, tumor-associated, and stromal compartments. The resulting cluster maps showed partial spatial correspondence with mIF reference stainings targeting epithelial and stromal markers, supporting the potential of this approach for label-free tissue characterization in digital pathology. Full article

Body mass index (BMI) does not capture fat distribution or muscle–fat heterogeneity, so adverse patterns go undetected. We present a methodological framework—not a validated prediction tool—combining three laboratory-free constructs: an Office-Based Framingham cardiovascular risk score, a modified proxy-based FINDRISC, and a direct segmental multi-frequency bioelectrical impedance analysis (DSM-BIA)-derived Metabolically Unhealthy Obesity (MUO) index, with a weighted patient similarity network. We tested six predefined hypotheses in 1684 young adults (mean age 22.5 ± 8.3 years; 50.2% female). Framingham was applied off-label below 30 years, so its outputs give only relative within-cohort ordering; unavailable FINDRISC items were scored zero, so standard FINDRISC categories do not apply. BMI-defined obesity occurred in 6.9%, high visceral fat in 21.7%, high MUO in 21.8%, a BIA-defined TOFI (thin outside, fat inside)-like phenotype in 5.2%, and a sarcopenic-obesity-like phenotype in 9.9%. Visceral fat correlated with percent body fat (r = 0.855). The network resolved nine interpretable communities (modularity Q = 0.63; permutation p = 0.005), including a BIA-defined TOFI-like community (cross-validated AUC = 0.93); k-means, hierarchical, PCA and UMAP clustering recovered convergent axes. All hypotheses were supported, indicating internal construct consistency, not external validation. Laboratory, imaging and longitudinal validation is required before any diagnostic or prognostic claim. Full article

Antimicrobial resistance (AMR), largely driven by the inappropriate and excessive use of antibiotics, requires rapid and reliable bacterial detection and antibiotic susceptibility testing (AST), as conventional culture-based methods remain time-intensive. Here, we report a real-time, label-free interdigitated wave electrode array (IWEA) that combines dielectrophoretic (DEP) enrichment with impedance spectroscopy for rapid bacterial detection and AST. The proposed IWEA was designed and optimized via COMSOL Multiphysics to enhance DEP-relevant electric-field strength, thereby improving DEP-assisted bacterial enrichment compared to conventional planar interdigitated electrodes. The platform enabled sensitive detection of both Gram-positive (Staphylococcus aureus) and Gram-negative (Escherichia coli) in 0.1× PBS across 10–10⁵ CFU/mL within 30 min using DEP-assisted preconcentration (100 kHz, 10 V_pp), outperforming passive (non-DEP) operation (10²–10⁵ CFU/mL). For AST, bacterial responses to vancomycin, gentamicin, and ampicillin were monitored through impedance variations following DEP-based enrichment. Susceptible bacteria produced concentration-dependent reductions in ΔZ/Z₀, whereas resistant bacteria showed similar responses to untreated controls. Quantitative assessment yielded CC₅₀ values of 2.86 ± 0.22 µg/mL and 3.06 ± 0.20 µg/mL for vancomycin and gentamicin against S. aureus, and 4.59 ± 0.30 µg/mL for gentamicin against E. coli. Resistance profiles of S. aureus to ampicillin and E. coli to vancomycin and ampicillin were clearly distinguished. SEM imaging and disk diffusion assays independently validated the impedance-derived susceptibility results. Collectively, this IWEA platform offers a rapid, label-free, and quantitative approach for bacterial detection and AST, with strong potential for antimicrobial screening and point-of-care diagnostics. Full article

We present a microfluidic platform for real-time, label-free cytometry and sorting of single cells based on intrinsic biophysical properties. The system integrates impedance-based electrokinetic sensing with a constriction microchannel architecture to induce controlled deformation during cell transit. Electrical signals captured via lock-in amplification are processed through a parallel software pipeline incorporating deep learning algorithms for event detection and feature extraction. A biomechanical model enables conversion of raw signal features into cell size and whole-cell elasticity, validated against imaging measurements. Experimental results demonstrate accurate phenotyping and sorting of live and dead MCF-7 cells, achieving sorting accuracy of 84%. This approach offers a cost-effective and scalable solution for biophysical analysis, with potential applications in liquid biopsy, disease diagnostics, and personalized medicine. Full article

Plasmonic biosensing technologies have emerged as powerful analytical tools for sensitive and label-free characterisation of biomolecular interactions and complex samples. The increasing demand for comprehensive molecular profiling has accelerated the development of multiplexing strategies that enable simultaneous analysis of multiple analytes and molecular interactions. This Feature Paper examines multiplexing through the complementary spatial, spectral, and temporal dimensions of multiplexing, together with their hybrid combinations and associated analytical trade-offs. Compared with other optical biosensing approaches, including interferometric, photonic, and fluorescence-based sensing platforms, plasmonic biosensors remain attractive owing to their combination of label-free detection, real-time interaction monitoring, sensitive interfacial analysis, and compatibility with multiplexed assay formats. This Feature Paper critically discusses current multiplexing strategies, focusing primarily on surface plasmon resonance (SPR), imaging SPR (SPRi), localised SPR (LSPR), surface-enhanced Raman scattering (SERS), and related nanoplasmonic biosensing approaches, together with recent advances in surface biofunctionalisation, antifouling interfaces, and molecular recognition strategies. Representative applications in biomedical diagnostics and non-clinical settings are highlighted, with examples such as liquid biopsy, glycoprofiling, extracellular vesicle profiling, and food and environmental analysis, alongside key challenges in reproducibility, standardisation, data interpretation, and clinical translation. In addition, selected non-plasmonic optical biosensing technologies are briefly discussed to position plasmonic biosensing within the broader landscape of multiplexed optical biosensing. This Feature Paper argues that the future of multiplexed plasmonic biosensing will depend less on further improvements in sensor performance than on robust, standardised analytical systems. Full article

Raman spectroscopy has emerged as a powerful tool for gas detection due to its label-free operation, molecular specificity, and multi-component analysis capabilities. However, its widespread application is hindered by limited sensitivity, particularly for trace gas analysis. To overcome this challenge, this study introduced an effective Raman spectroscopy detection system that synergistically combines a 532 nm high-power laser with a near-concentric multipass cell (MPC), enabling dual enhancement of both Raman excitation and signal collection. We simultaneously determined three critical performance metrics, including gas Raman signal intensity, signal-to-noise ratio (SNR), and limit of detection (LOD). Under the optimized experimental conditions, the CO₂ Raman signal reached an SNR of approximately 58 for laboratory air containing 916 ppm CO₂, corresponding to a concentration-equivalent detection limit of 48 ppm according to the 3σ criterion. Notably, the intensity of the generated Raman scattering signal is proportional to the average power. The intensity of Raman signals at different wave numbers increases at different rates with the increase in the average excitation power. The system achieves a remarkable LOD of 48 ppm for CO₂, representing an advancement over conventional Raman gas sensors. This work validates high-power near-concentric cavity-enhanced Raman spectroscopy as a reliable method for trace gas detection, with potential implications for multi-component gas analyzers in environmental monitoring, industrial safety, and medical diagnostics. Full article

Rapid, point-of-care detection of Escherichia coli O157:H7 remains an unmet clinical need, as culture and molecular methods are slow and poorly suited to decentralized or emergency settings. A label-free, monolithic aptasensor biochip was fabricated in a standard 65 nm CMOS process, featuring three aptamer-functionalized gold sensing pads with matched reference pads for differential readout. A 37-mer DNA aptamer targeting the E. coli O157:H7 lipopolysaccharide was immobilized via thiol–gold self-assembled monolayer chemistry. Binding events were transduced into surface-potential shifts, amplified by an on-chip analog front-end (~100 V/V gain, 101.5 µW), and evaluated using calibration standards, patient specimens, and hospital environmental samples, with fluorescence microscopy for validation. The sensor achieved 47.42 mV/decade sensitivity across 1–10,000 CFU/mL, an IUPAC detection limit near 3.74 CFU/mL, and an empirical LOD of about 11 CFU/mL, with outputs tracking bacterial load and ~5.7% matrix-related deviation. Hospital samples were detectable to 28 CFU/mL. Because the patient-derived and hospital-acquired cohorts (n = 10 and n = 6, respectively) were assembled for pilot analytical and matrix-tolerance characterization rather than for diagnostic-accuracy determination, these results establish detectability and matrix robustness in real clinical and environmental specimens rather than clinical diagnostic sensitivity or specificity, which will require a larger, prospectively enrolled cohort in future work. Sensor kinetics followed Langmuir-type adsorption, saturating within 16–25 min for target pathogens versus slower responses for non-target strains. Selectivity tests against six bacterial species showed discrimination, with cross-reactivity decreasing from related E. coli pathotypes to Enterobacteriaceae to Gram-positive species. Inter-pad variability stayed below 1.5 mV, supporting this compact, low-power platform for scalable, enrichment-free point-of-care pathogen detection. Full article

In recent years, nanopore technology has enhanced analyte detection, enabled higher resolution and achieved single-molecule sensing capability. Aptamer-conjugated nanopore sensing technology combines the high specificity of aptamers with the single-molecule resolution of nanopores. By anchoring aptamers to biological, solid-state or hybrid nanopores, target binding events produce distinct electrical signatures that allow sensitive and label-free detection. This approach enables real-time monitoring of small molecules, proteins, and even pathogens, with promising applications in diagnostics, drug screening, environmental monitoring, etc. Hybrid biological/solid state devices produce robust signals and are suitable for PoC applications. The aptamers “magic bullets” have also been exploited to develop single-molecule antigen detection using nanopores, which offers a promising alternative for accurate virus testing to contain their transmission. Chemical conjugation of aptamers to nanopore interfaces improves selectivity for peptides/amino acids and expands robustness for practical samples. Aptamer-based nanopipettes offer high analytical precision by enabling label-free, real-time detection of target molecules in ultra-small sample volumes. This review maps aptamer–nanopore integration across biological, solid-state, and hybrid platforms. It also explores various types of aptamers integrated into nanopore platforms that cater to precise, single-molecule recognition, paving the way for highly sensitive, portable diagnostics and next-generation therapeutic monitoring tools. Full article