Search Results (701)

Background. Accurate prediction of pathologic complete response (pCR) to neoadjuvant chemotherapy (NAC) in muscle-invasive urothelial bladder cancer (MIBC) remains an unmet clinical need. The neutrophil-to-platelet/hemoglobin-to-lymphocyte (NPLH) ratio, a composite hematologic index that reflects both systemic inflammation and nutritional oxygen-carrying capacity, has not been previously evaluated as a predictor of NAC response in this setting. Methods. We retrospectively analyzed 114 consecutive patients with MIBC (cT2–T4, N0–N3) who received accelerated MVAC (aMVAC) NAC followed by radical cystectomy at a single academic center. Pretreatment NPLH (calculated as [neutrophils × platelets]/[hemoglobin × lymphocytes]) was assessed as a predictor of pCR (ypT0N0) and tumor regression grade (TRG). Receiver operating characteristic (ROC) curve analysis, Mann–Whitney U test, and logistic regression were used. NPLH performance was compared to NLR and PLR. Results. pCR was achieved in 35 patients (30.7%). Median NPLH was significantly lower in pCR vs. non-pCR patients (33.9 [IQR 23.1–42.4] vs. 47.6 [IQR 30.7–90.4]; p = 0.0007). NPLH yielded an AUC of 0.700 (bootstrap 95% CI 0.596–0.794) for pCR prediction, numerically superior to NLR (AUC 0.645 [0.542–0.741]) and PLR (AUC 0.643 [0.533–0.747]); DeLong test: NPLH vs. NLR p = 0.079, NPLH vs. PLR p = 0.090. At the optimal cut-off of 44.5, NPLH demonstrated 80.0% sensitivity and 57.0% specificity. pCR rates declined progressively across NPLH quartiles: 48.3% (Q1) to 10.3% (Q4). On multivariate logistic regression, log-transformed NPLH was the only independent predictor of pCR (parsimonious model, OR 0.292, 95% CI 0.131–0.652; p = 0.003; EPV = 17.5). A positive correlation was observed between NPLH and TRG score (Spearman r = 0.284; p = 0.0022), with significant differences between TRG 1 and TRG 3 subgroups (p = 0.0036). Conclusions. Pretreatment NPLH is an independent predictor of pCR to aMVAC in MIBC and is numerically superior to NLR and PLR (DeLong p = 0.079). Consisting exclusively of standard complete blood count parameters, NPLH is readily available and inexpensive. This single-center exploratory study is hypothesis-generating and requires prospective external validation before clinical implementation. Full article

Background: Early and accurate risk stratification of patients suspected of serious infection is essential for improving outcomes, but existing diagnostic and predictive tools have limited accuracy. The objective was to compare the performance of an FDA-authorized AI diagnostic test, the Sepsis ImmunoScore, against widely available biomarkers and clinical tools for diagnosis of sepsis and prediction of in-hospital mortality and intensive care unit (ICU) admission. Methods: This multicenter observational study included 6027 adult patients suspected of infection across 7 U.S. hospital sites. The Sepsis ImmunoScore’s predictive performance was compared to the sequential organ failure assessment (SOFA) score, procalcitonin (PCT), C-reactive protein (CRP), Systemic Inflammatory Response Syndrome (SIRS) score, National Early Warning Score (NEWS), and quick SOFA (qSOFA). Primary outcomes included sepsis as defined by Sepsis-3 criteria, in-hospital mortality, and ICU admission. Predictive accuracy was assessed using area under the receiver operating characteristic curve (AUC), and 95% confidence intervals were generated and hypothesis testing conducted using the bootstrap method. Results: The Sepsis ImmunoScore demonstrated statistically significant superior performance across all outcomes. For sepsis prediction, the Sepsis ImmunoScore achieved an AUC of 0.82, compared to SOFA (0.72), procalcitonin (PCT) (0.70), C-reactive protein (CRP) (0.61), SIRS (0.59), NEWS (0.69), and qSOFA (0.67). For in-hospital mortality prediction, the Sepsis ImmunoScore achieved an AUC of 0.80, outperforming SOFA (0.72), PCT (0.67), CRP (0.58), SIRS (0.60), NEWS (0.72), and qSOFA (0.69). For ICU admission, the Sepsis ImmunoScore reached an AUC of 0.74, superior to SOFA (0.63), PCT (0.64), CRP (0.54), SIRS (0.60), NEWS (0.70), and qSOFA (0.65). All differences between the Sepsis ImmunoScore and comparators were statistically significant. Conclusions: The Sepsis ImmunoScore significantly improved predictive accuracy for sepsis, in-hospital mortality, and ICU admission compared to six conventional clinical scores and biomarkers. This AI-based tool may enhance risk stratification and clinical decision-making, potentially leading to more timely sepsis interventions and improved outcomes. Full article

Anomaly detection models trained exclusively on healthy data are widely used in wind turbine condition monitoring because failure data are scarce, heterogeneous, and often unavailable. However, these models produce anomaly indicators that are sensitive not only to fault-related degradation but also to normal operational variability, transient disturbances, and changes in loading conditions. As a result, the practical behavior of an alarm system depends not only on the anomaly detection model but also on the decision rule used to activate and maintain alarm states. This study presents a decision-oriented evaluation of persistence-based alarm confirmation in wind turbine anomaly detection. Four representative techniques are analyzed within a unified framework: Isolation Forest, One-Class Support Vector Machine, Referenced Moving Window Principal Component Analysis using Q-statistic and percentage component weight indicators, and Autoencoder-based reconstruction error. The evaluation combines controlled OpenFAST simulations of rotor unbalance under different severity and noise conditions with an industrial SCADA case study involving a documented main bearing fault. Results show that temporal persistence strongly shapes alarm outcomes across methods and datasets. Low persistence values favor early detection but promote alarms from isolated threshold exceedances, whereas moderate persistence substantially reduces false positives while preserving detection capability in severe and well-observable faults. Excessive persistence increases detection latency and missed detections, particularly for weak, intermittent, or slowly evolving fault signatures. These findings indicate that persistence-based alarm confirmation should be treated as an explicit decision-level configuration variable, rather than as a fixed post-processing or alarm-state heuristic, when designing anomaly detection systems for wind turbine condition monitoring. Full article

Sweep-field operation in a single-beam spin-exchange relaxation-free (SERF) magnetometer requires stable extraction of the dispersion zero-crossing. A frequency mismatch between the modulation signal and the demodulation references, or an unsuitable low-pass filter, can shift this zero-crossing and affect working-point determination. This paper presents a zero-crossing-stability-oriented FPGA quadrature lock-in module for SERF sweep-field demodulation. The module is designed around two requirements of sweep-field operation: maintaining a common frequency basis between the modulation output and the demodulation references, and preserving the dispersion zero-crossing when the low-pass-filter cutoff frequency is adjusted. A shared direct digital synthesizer generates both the sinusoidal modulation output and the I/Q references, keeping the excitation and demodulation signals on the same frequency basis. After quadrature multiplication, CIC decimation and a reloadable Kaiser-window FIR filter are used for low-pass processing. Board-level tests show a 1000.054 Hz spectral peak for a 1000 Hz setting and a loopback amplitude of 0.496 V, close to the ideal 0.500 V baseband amplitude. On the SERF platform, I/Q rotation reduces the quadrature residual ratio from 32.1% to 0.10%. When the FIR cutoff frequency is changed from 3 to 15 Hz, the maximum zero-crossing difference is about 0.58 ms, corresponding to 0.12% of the 2 Hz sweep period. These results show that the module supports stable zero-crossing extraction and working-point determination during sweep-field operation in a single-beam SERF magnetometer. Full article

We develop a comparative synthesis of quantum-information geometry beyond complex numbers, with emphasis on what different algebraic frameworks contribute to information-processing structure rather than on their formal novelty alone. The organizing idea is a layer-by-layer test of the standard complex Hilbert-space formalism: each non-complex or deformed framework modifies the scalar field, phase group, projective state space, Born-probability semantics, composition rule, measurement geometry, symmetry algebra or representation category. The central thesis is that such frameworks are physically meaningful when they identify which assumptions make complex quantum mechanics operationally stable: positive probabilities, associative multipartite composition, reversible dynamics, experimentally testable phases, locality constraints, informationally complete measurements, error bases and clear operational semantics. Real quantum theory probes the necessity of complex phases and local tomography; quaternionic quantum mechanics probes non-Abelian phase while retaining associativity and admitting complex embeddings; octonionic proposals probe the boundary where exceptional geometry survives but generic circuit composition is obstructed by non-associativity; Jordan algebras test ordered probabilistic state spaces; Clifford algebras and Bott periodicity provide the spinorial and topological grammar connecting gates, Hopf maps and periodic dimensions; and quantum-group or q-deformed constructions probe coproducts, braiding and representation categories rather than scalar amplitudes. We distinguish three roles that are often conflated: genuine hypercomplex kinematics, Hopf-fibration coordinates for ordinary complex multipartite entanglement, and deformed algebraic or categorical structures. The resulting map separates established equivalence and experimental-constraint results from useful representation tools and speculative programs, while identifying concrete open problems for non-complex quantum information. Full article

We examine the association between Environmental, Social, and Governance (ESG) initiatives and corporate performance using a sample of 360 S&P 500 firms from 2010 to 2018. Employing MSCI ESG ratings and controlling for industry and time effects, we find that environmental initiatives positively associate with current profitability (ROA), while gender diversity correlates with long-term growth prospects (Tobin’s Q). This study moves beyond aggregated ESG metrics by providing a disaggregated analysis, revealing that different ESG dimensions affect performance through distinct financial mechanisms. To address common endogeneity concerns, we implement a rigorous empirical identification strategy, including propensity score matching, Heckman selection models, and instrumental variable approaches using industry-average instruments. Our results quantify the economic magnitude of these effects, demonstrating that a one-standard-deviation increase in environmental performance corresponds to a 0.92 percentage point increase in ROA, representing approximately $176 million in additional annual net income for the median firm. These findings provide theoretical advancement for the resource-based view and stakeholder theory by showing that specific ESG capabilities serve as valuable, inimitable resources. Ultimately, the study contributes standardized, high-resolution evidence on how specific ESG dimensions drive superior corporate performance. Through mechanism analysis, we show that environmental effects operate primarily via operational cost reduction and risk mitigation, while gender diversity creates value through enhanced innovation findings, which has direct implications for corporate ESG strategy. Full article

Fourth-generation immunoassays are widely used for HIV and syphilis screening; however, false-reactive results may increase confirmatory testing and operational burden in high-throughput laboratories. This study evaluated the comparative performance of automated chemiluminescent immunoassays (MAGLUMI^® HIV Ab/Ag Combi (Snibe Diagnostics Co. Ltd., Shenzhen, China), VITROS^® ECiQ HIV Combo (Ortho Clinical Diagnostics, Raritan, NJ, USA), MAGLUMI^® Syphilis (Snibe Diagnostics Co. Ltd., Shenzhen, China), and ARCHITECT^® Syphilis TP (Abbott Diagnostics, Abbott Park, IL, USA) within a routine diagnostic algorithm, incorporating antibody differentiation immunoassays (INNO-LIA^® HIV I/II Score (Fujirebio Europe N.V., Ghent, Belgium) and HIV-1 RNA PCR where applicable. A total of 240 archived serum samples for HIV testing and 180 for syphilis testing were analyzed. Agreement-based performance measures including sensitivity, specificity, overall percent agreement (OPA), and Cohen’s kappa (κ) were calculated as comparator-based estimates reflecting concordance within the routine diagnostic algorithm rather than absolute diagnostic accuracy against a universal reference standard. For comparisons with HIV-1 RNA PCR, positive and negative concordance rates are reported to reflect agreement between assays detecting different biological targets. Among samples with definitive (positive or negative) results, the MAGLUMI^® HIV Ab/Ag Combi assay showed complete agreement with INNO-LIA^® HIV I/II Score (κ = 1.00) and high agreement with PCR within the ARCHITECT^® HIV Ag/Ab Combo-reactive subset (κ = 0.90). The VITROS^® ECiQ HIV Combo assay demonstrated high agreement with INNO-LIA^® HIV I/II Score (κ = 0.916) and substantial agreement with PCR (κ = 0.715), with a lower negative concordance rate with PCR observed in the ARCHITECT-reactive subset. A parallel five-modality analysis of 11 discordant samples applying the CDC 2014 algorithm demonstrated that all three immunoassay platforms successfully detected confirmed HIV-seropositive individuals with controlled viremia despite negative PCR, while MAGLUMI^® HIV Ab/Ag Combi produced fewer false-reactive results than both ARCHITECT^® and VITROS^® in this discordant subset. Additionally, two cases showed INNO-LIA^® indeterminate results with positive PCR, consistent with acute HIV infection during the early seroconversion stage; all three immunoassay platforms produced signals above the non-reactive threshold in both cases. For syphilis testing, both MAGLUMI^® Syphilis and ARCHITECT^® Syphilis TP assays showed complete agreement with INNO-LIA^® Syphilis Score among samples with definitive results (κ = 1.00). In contrast, the RPR assay showed reduced positive predictive value (49.4%) and moderate agreement with INNO-LIA^® Syphilis Score (κ = 0.52). Automated chemiluminescent immunoassay (CLIA) platforms demonstrated high agreement within a structured diagnostic algorithm in a high-throughput screening setting. Differences in assay performance were observed across platforms, particularly with respect to discordant results in the ARCHITECT-reactive PCR-evaluated subset for HIV and non-treponemal concordance for syphilis. These platforms may support more efficient laboratory workflows; however, findings should be interpreted within the context of comparator-based classification rather than absolute diagnostic accuracy. Full article

Future 6G networks require higher performance and wider service coverage. Multi-domain Service Function Chain (SFC) deployment enables service provisioning across multiple network domains to meet these demands. However, when collaboration occurs among different network operators, privacy-preserving mechanisms are required to protect sensitive information such as internal topology and resource availability. Existing SIRM-based mechanisms, such as the Privacy-Preserving Deployment Mechanism (PPDM), address this challenge but suffer from structural limitations: PPDM performs whole-chain feasibility evaluation with extensive virtual occupation. This paper proposes a $\mathcal{B}$-Batch Sequential Deployment mechanism for privacy-preserving multi-domain SFC deployment. Instead of evaluating whole-chain feasibility at once, the proposed $\mathcal{B}$-Batch mechanism partitions each incoming SFC into fixed-size VNF batches and constructs a batch-level SIRM. This design confines virtual occupation to the current batch and reduces both its magnitude and duration while remaining fully compatible with the SIRM privacy model and the hierarchical multi-domain control architecture. A Deep Q-Network (DQN) is employed to learn substrate node selection policies based solely on SIRM-based state information, without exposing domain-internal topology or resource details. Simulation results on a three-domain AARNET substrate topology demonstrate that the proposed mechanism consistently improves deployment robustness under varying traffic intensities and SFC lengths, including short (3–6 VNFs), medium (6–9 VNFs), and long (9–12 VNFs) service chains. Compared with PPDM, the proposed $\mathcal{B}$-Batch mechanism achieves higher acceptance ratios under moderate-to-heavy traffic while reducing end-to-end delay and improving average substrate resource utilization. Node selection analysis further shows that smaller batch sizes preserve feasibility through compact node reuse, whereas larger batch sizes encourage broader substrate exploration. Overall, the proposed $\mathcal{B}$-Batch mechanism enhances feasibility preservation and deployment robustness in privacy-preserving multi-domain SFC orchestration. Full article

Simultaneous autologous breast reconstruction (ABR) and lymphatic surgery has emerged as a strategy to address breast cancer-related lymphedema (BCRL) while restoring breast contour within a single operative procedure. In light of the diversity of surgical strategies, we aimed to evaluate the current literature on combined ABR and lymphatic surgery, with particular focus on surgical techniques, clinical outcomes, complications, and patient-reported satisfaction. A systematic review was conducted according to PRISMA guidelines and registered in PROSPERO (CRD420251135446). Medline, Embase, CENTRAL, Web of Science, and PubMed were searched through 9 January 2026 Studies reporting outcomes of simultaneous ABR and lymphatic reconstruction were included. Data on surgical techniques, complications, changes in limb volume and cellulitis incidence, and patient-reported outcomes were extracted. Twenty-seven studies including 499 patients (mean follow-up 23 months) were analyzed. The most common approach was a chimeric deep inferior epigastric perforator (DIEP) flap with inguinal lymph nodes (459 patients), followed by ABR with a separate vascularized lymph node transfer and ABR with lymphovenous anastomosis. Most studies reported postoperative reductions in limb volume and cellulitis, with cellulitis reduction rates up to 100%. Patient-reported outcomes (LYMQOL, ULL-27, LYMPH-Q) showed improved quality of life. Complication rates were low, including 4% seroma and 1% flap loss after chimeric DIEP. Simultaneous ABR and lymphatic reconstruction is feasible and associated with improved clinical and patient-reported outcomes. However, heterogeneity limits comparison between the different surgical techniques, and prospective studies with standardized outcomes measurements are needed. Full article

In this paper, we introduce a new difference operator, called the $\left(p,q\right)$-Hahn difference operator, which extends both the classical Hahn operator and the $\left(p,q\right)$-difference operator by incorporating an additional shift parameter, $\omega$. This extension allows the operator to reflect translation effects together with scaling behavior within a common framework. We investigate the basic algebraic properties of the operator, including linearity, product and quotient rules, and explicit formulas for power functions. A generalized Leibniz rule is established using $\left(p,q\right)$-binomial coefficients. In addition, a corresponding $\left(p,q\right)$-Hahn integral is defined, and a fundamental relation between the operator and the integral is obtained under suitable assumptions. Furthermore, several special and limiting cases are analyzed in order to clarify the connection between the proposed operator and existing difference operators. In particular, it is shown that the operator reduces to the classical derivative, the $q$-difference operator, and the standard $\left(p,q\right)$-difference operator under appropriate parameter choices. Finally, applications to $\left(p,q\right)$-Hahn Sturm–Liouville-type problems and hypergeometric-type difference equations are discussed. These results suggest that the proposed operator provides a consistent extension of existing difference-calculus structures. Full article

The rapid growth of solar photovoltaic (PV) capacity is increasingly reshaping the operation of electricity systems, particularly in countries where renewable energy already represents a large share of generation. In Albania, where electricity production is strongly dominated by hydropower, increasing solar penetration is expected to affect short-term system behaviour, especially in terms of variability, surplus generation, and ramping dynamics. This study investigates PV integration at the system level using hourly electricity demand data for 2024 together with PV generation profiles scaled to different capacity scenarios. PV scenarios representing installed capacities of 150, 300, and 450 MWp, based on real PV deployment data, are analysed under varying levels of hydropower dominance. The analysis combines net-load modeling, ramping assessment, and a simplified flexibility-oriented mitigation approach to evaluate operational impacts under different hydropower conditions. The results indicate that increasing PV capacity significantly modifies the net-load profile. During summer periods, high solar generation substantially reduces midday net load, creating pronounced net-load valleys, whereas winter conditions remain more strongly influenced by electricity demand. As PV penetration increases, ramping intensity also increases. For example, extreme ramp values (Q99) rise from 80.87 MW/h at 300 MWp to 111.45 MW/h at 450 MWp, while the share of hours with ramp events exceeding 100 MW/h increases from 0.05% to 2.55%. The results of a conceptual flexibility approach that limits ramps to 60 MW/h show that extreme ramp events can be effectively mitigated, while moderate variability is largely unaffected. In summary, the results show that increasing solar PV penetration shifts the main operational challenge in Albania from energy balancing toward flexibility and variability management. The findings are particularly relevant for long-term system planning in hydropower-dominated systems and highlight the growing importance of flexibility measures and surplus management under high PV penetration. Full article

Introduction: Standard abdominoplasty inevitably causes scarring in the umbilical region. Given the aesthetic significance of the navel, as the central anatomical landmark of the anterior abdominal wall, techniques that preserve the natural anatomy are warranted. The criteria for patient selection for such operations remain poorly defined, limiting their clinical application. Objective: The aim of this study was to define the indications and contraindications for scarless abdominoplasty in the umbilical region through a comparative analysis of clinical and subjective outcomes. Materials and Methods: A comparative study was conducted involving 115 patients operated on between 2021 and 2026. The experimental group consisted of 32 patients who underwent abdominoplasty without scarring in the umbilical region, while the control group comprised 83 patients who underwent standard abdominoplasty with umbilical transposition. The technique of scarless abdominoplasty in the umbilical region entailed limited mobilization of the cutaneous–subcutaneous flap, energy-based ultrasound-assisted liposuction in the supraumbilical region, suturing of rectus abdominis diastasis when clinically indicated, and fixation of the umbilicus in its physiological position without a skin incision around it. Energy-based ultrasound-assisted liposuction permits skin contraction in the supraumbilical area through a controlled internal thermal effect. The incidence of complications and BODY-Q questionnaire outcomes were evaluated at 3 months, 6 months, and one year. No significant differences in baseline clinical characteristics were observed between the groups. Results: The overall rate of postoperative complications in the experimental group was 18.8% (n = 6), compared to 24.1% (n = 20) in the control group; the difference was not statistically significant (p > 0.05). According to the BODY-Q questionnaire, patients in the experimental group reported significantly higher satisfaction with scar appearance (91.0 ± 3.5 vs. 83.0 ± 5.7; p < 0.001) and the umbilical region. The umbilicus retained its anatomically natural shape and depth. The use of energy-based ultrasound-assisted liposuction in the supraumbilical area contributed to the contraction of the upper cutaneous–subcutaneous flap, which made it possible to prevent extensive mobilization and the formation of a significant scar around the umbilicus in these patients. Conclusions: Scarless abdominoplasty in the umbilical region is a safe and effective technique in appropriately selected patients. Energy-based ultrasound-assisted liposuction produces an internal thermal effect that enhances the contraction of the cutaneous flap. Indications for this method comprise a moderate amount of excess skin tissue in the supraumbilical region, preserved skin elasticity, a normal or high umbilical position, and rectus abdominis diastasis grade I–II. Contraindications include pronounced excess skin tissue in the supraumbilical region, a low umbilical position, and general somatic contraindications to surgery. Full article

The circadian clock is a widespread rhythmic phenomenon across organisms, characterized by distinct gene expression patterns and behaviors at specific times of the day. Extensive genetic studies in the model fungus Neurospora crassa have yielded critical insights into the components and molecular mechanisms of circadian oscillators. However, these understandings remain absent across fungal lineages, especially from edible mushrooms. Morels (Morchella spp.) are well-recognized edible ascomycetes of considerable economic value and are partially artificially cultivated, but their biological characteristics are poorly understood. Investigating the presence of their circadian clock components, as well as the molecular underpinnings of circadian rhythms, holds important biological implications. In this study, we firstly performed a genomic search for homologs of known circadian clock genes in Morchella sextelata. Homologs of seven circadian clock genes, including wc-1, wc-2, fwd-1, frh, frq, and two additional clock-controlled genes, were identified, indicating the components necessary for the operation of a FWC oscillator contained in M. sextelata. Then, using reverse transcription quantitative PCR (RT-qPCR), the expression profiles of these seven circadian clock-related genes and four mating-type genes were examined in RNA samples which were extracted from mycelia of MAT1-1, MAT1-2 and MAT1-1 × MAT1-2 co-culture/crossed condition during conidiation under in vitro cultivation across one day. The expression levels of seven circadian clock genes and four mating-type genes displayed similar time-of-day-specific rhythmic patterns, yet remained consistently distinct across the mating-type strains and their co-culture/crossed condition, indicating a potential correlation between circadian clock and mating-type loci. Collectively, these results suggest that M. sextelata harbors conserved circadian clock-related homologs and displays mating-type-associated temporal expression differences under the tested conidiation conditions, offering a novel perspective for exploring the potential link between clock-related regulation and mating-type background in the future. Full article

As an active defense paradigm, cyber deception technology effectively misleads attackers by constructing deceptive network environments, thereby increasing the cost of attack operations and introducing uncertainty into their decision-making, while providing defenders with critical response time. However, existing deception strategies are mostly based on predefined static rules derived from expert knowledge and lack the ability to adapt to dynamic attack scenarios autonomously and intelligently. This limitation results in poor adaptability and suboptimal performance of the strategy. To solve these issues, this paper proposes an Adaptive Cyber Deception Defense System (ACDDS). Different from off-the-shelf MDP/DQN frameworks in existing adaptive defense, the core innovation of ACDDS is a scenario-customized Trigger-based Adaptive Deception Strategy evolution method using Deep Q-Networks (TADS-DQN). We specifically formulate the dynamic deception strategy optimization as a cyber-deception-tailored Markov Decision Process (MDP). In this model, the state of the system is represented as a state matrix, and the attack behavior defines the environment for agent interaction. The TADS-DQN method employs a trigger-based mechanism: when a threat to real services is detected, a Deep Q-Network agent is activated. This agent takes the current system state as input and outputs the optimal reconfiguration action. The simulation results indicate that, compared to the baseline methods, TADS-DQN provides more stable defense performance, as evidenced by a smaller fluctuation range and a lower standard deviation of the attack success rate. At the same time, it achieves a reduction in the hit rate against real services that is competitive with the baseline methods. Full article

Long-memory degradation processes in rotating machinery often exhibit nonlinear evolution, nonlocal temporal dependence, and hereditary characteristics, which are difficult to fully capture using conventional integer-order models or standard Markovian decision frameworks. To address this issue, this study proposes a hybrid fractional-dynamics and deep reinforcement learning framework for predictive maintenance of memory-dependent degradation systems. First, the Grünwald–Letnikov fractional difference operator is introduced to construct a fractional-memory representation of degradation trajectories, enabling the model to explicitly encode long-range dependence and accumulated historical degradation effects. Then, a bidirectional gated recurrent unit network is employed to learn sequential degradation representations from the fractional-memory state space, while a deep Q-network is designed to optimize maintenance decisions under uncertain degradation evolution. Experimental results on the IEEE PHM 2012 bearing dataset show that the proposed FM-BiGRU-DQN with safety-guided execution achieved a mean maintenance lead time of 33.9 ± 12.6 steps, an in-band rate of 0.85 ± 0.06, a failure rate of 0.00 ± 0.00, and a deployment reliability of 1.00 ± 0.00 over 10 independent random seeds. Compared with NM-BiGRU-DQN, the in-band rate increased from 0.55 ± 0.10 to 0.85 ± 0.06, with a paired-test p-value of 0.013. Cross-dataset validation on the XJTU-SY bearing dataset further achieved an in-band rate of 0.80 and a failure rate of 0.00. These results indicate that embedding fractional-memory dynamics into deep reinforcement learning improves maintenance timing accuracy, policy robustness, and deployment reliability for complex memory-dependent degradation systems. Full article