Abstract: Rapid technological progress in the 21st century has transformed neuroradiology from primarily a diagnostic tool into a vital component of modern neuroscience. This abstract summarizes innovations that have marked a paradigm shift in neurological and neurosurgical patient care. From symptom onset through hospitalization and recovery, these advancements have introduced novel techniques and redefined existing models to improve short-term and long-term outcomes.Artificial intelligence (AI) integration has significant implications in stroke triage, brain tumor segmentation, and treatment planning. Wearable AI-enabled devices detect abnormal changes in movement and speech, prompting patients to seek care immediately to ensure they remain within the therapeutic window for reperfusion therapy. AI models can also identify tumors that may be undetectable to the naked eye and provide precise lesion margins, reducing overestimation and underestimation during surgical planning. Image-guided minimally invasive procedures have reduced reliance on open surgery while increasing precision, minimizing perioperative risk, shortening postoperative hospital stays, improving functional outcomes, and lowering recurrence rates. Refinements in stent and catheter technology have further enhanced the safety and efficiency of these procedures in treating neurological disorders. Advanced imaging enables neurosurgeons to target surgically inaccessible lesions while preserving surrounding tissue and to analyze tumor behavior to objectively predict treatment response. Innovations in neuroradiology address disparities in access to care, as minimally invasive approaches offer therapeutic options for patients ineligible for open surgery, and remote interventions aim to expand timely access to specialized treatment in underserved regions. Continued advancements in neuroradiology will further optimize interventions, enhancing patient care.

Abstract: Background: Cryptococcal meningitis (CM) remains a leading cause of mortality among people with advanced HIV disease (AHD) in sub-Saharan Africa. Current guidelines recommend induction therapy with amphotericin B and flucytosine, typically administered in an inpatient setting due to concerns over severe clinical presentation and drug-related toxicities. This requirement poses a significant burden on resource-limited health systems. We evaluated the real-world outcomes of a fully outpatient model for CM therapy in Maputo, Mozambique. Methods: A longitudinal retrospective cohort study was conducted at the Centro de Referência de Alto-Maé (CRAM), a specialized AHD outpatient clinic. We included 83 PLWH with laboratory-confirmed CM treated between October 2020 and December 2024. The primary outcome was hospitalization-free survival (HFS) within the first 10 weeks of treatment. Secondary outcomes included the frequency and severity of adverse drug reactions (ADRs), analysed by tracking haemoglobin (Hgb), potassium (K+), and creatinine (Creat) levels on days 1, 3, and 7 of induction therapy, and retention in care (RIC) at 6, 12, and 24 months. Statistical analyses included Kaplan-Meier survival estimates and paired t-tests. Results: The median age was 37 years (IQR: 27-42), 63.9% were male, and the median CD4 count was 62 cells/µL (IQR: 27-105). Most patients (95.2%) were symptomatic at presentation, and 56.6% had concurrent tuberculosis. For the 52 patients who completed the full induction protocol at CRAM, the HFS rate at 10 weeks was 84.6% (44/52), with an overall survival of 90.4% (47/52). ADR analysis (n=52) showed a predictable pattern of mild, manageable toxicity: a significant decline in Hgb (11.2 ± 1.8 to 10.6 ± 2.0 g/dL, p<0.001) and K+ (4.27 ± 0.66 to 3.86 ± 0.78 mmol/L, p=0.008), and a transient increase in Creat (0.83 ± 0.42 to 1.13 ± 0.64 mg/dL, p=0.001) from day 1 to day 3, with stabilization or trend toward recovery by day 7. No significant differences in ADRs were found between single-dose (47%) and multiple-dose (53%) L-AmB regimens. RIC for the entire cohort (n=83) was high, at 81.9% at 6 months, declining to 74.0% at 12 months and 70.4% at 24 months. Conclusion: An ambulatory model for CM therapy is feasible and effective in a resource-limited setting, demonstrating high hospitalization-free survival, manageable and reversible adverse drug reactions, and excellent medium-term retention in care. These findings provide compelling evidence to reconsider the standard of inpatient care and support the integration of outpatient CM management into AHD care packages to alleviate health system burdens and improve patient outcomesAn integrated care approach is essential to improving survival in resource-limited settings.

Abstract: Background/Objectives: Obesity is frequently associated with dyslipidemia, insulin re-sistance, and increased cardiovascular risk. Bariatric surgery, particularly Roux-en-Y gastric bypass (RYGB), has been shown to improve metabolic outcomes; however, data integrating lipid profile changes with atherogenic and insulin resistance indices over extended follow-up remain limited. This study aimed to evaluate the longitudinal impact of RYGB on lipid parameters, atherogenic indices, and insulin resistance markers up to 18 months postoperatively. Methods: A longitudinal observational study was conducted in 40 adults with obesity (mean age 46.6 years; 82.5% women) undergoing RYGB. Bio-chemical and anthropometric data were collected preoperatively (E1) and at 6 (E3), 12 (E4), and 18 (E5) months after surgery. Lipid profile parameters, atherogenic indices (TG/HDL ratio, atherogenic indices Castelli I and Castelli II), and insulin resistance–related indices (TyG-BMI, MTS-IR) were analyzed descriptively over time. Results: RYGB induced a marked and sustained reduction in triglycerides, total cholesterol, and LDL cholesterol, alongside a progressive increase in HDL cholesterol. All atherogenic indices demon-strated consistent postoperative declines, indicating reduced cardiovascular risk. Insulin resistance indices showed pronounced improvement, particularly within the first 12 months, with stabilization thereafter. These metabolic benefits occurred in parallel with significant weight loss and reductions in central adiposity. Conclusions: Roux-en-Y gastric bypass promotes sustained improvements in lipid metabolism, atherogenic risk, and insulin resistance up to 18 months after surgery. The integration of lipid-derived and insulin resistance indices provides valuable insight into postoperative cardiometabolic risk reduction, supporting the role of bariatric surgery as an effective intervention for long-term cardiovascular risk mitigation.

Abstract: In the framework of (fuzzy) Multi-Criteria Decision-Making, we propose a method that1 allows the decision maker to subjectively approach the problem by suitably modifying the decision matrix. We consider a decision problem related to a random quantity X with set of values {x₁, x₂, . . . , x_n}, and a set of properties {C₁, C₂, . . . , Cm}of X. In this setting, the properties C_j are the criteria of the decision problem, the alternatives represent the events A_i = (X= x_i), for i= 1, . . . , n, and the criteria’s weights w_j, for j= 1, 2, . . . , m, are seen as the probabilities of the events “C_j is relevant with respect to the decision problem”. For each i= 1, . . . , n and j= 1, 2, . . . , m, we interpret the scores aij as membership functions representing “how much alternative A_i satisfies criterion C_j”. By adopting the interpretation of membership functions as suitable conditional probabilities, together with the theory of logical operations among conditional events, we allow logical operations among criteria and consistently apply this interpretation to the corresponding scores. In particular, when considering the complement, conjunction, and disjunction of criteria, the resulting scores are the coherent) previsions of the respective compound conditionals within the framework of conditional random quantities.

Abstract: Infrared traffic object detection faces challenges such as low resolution, weak thermal 2 contrast, and inefficiency in detecting small objects. To address these issues, this paper 3 proposes RES-YOLO, an enhanced YOLOv8n-based architecture. It incorporates Receptive 4 Field Adaptive Convolution for improved multi-scale perception, Efficient Multi-scale 5 Attention for better feature representation, and the Scylla-IoU loss for more accurate 6 and faster bounding box regression. Additionally, a pseudo-color infrared dataset is 7 constructed to enrich texture and contrast information beyond conventional white-hot 8 images. Experiments on both the FLIR public dataset and a self-built dataset show RES- 9 YOLO improves accuracy by 4.9% and 5.5% over the baseline while maintaining real-time 10 performance. These results highlight the method’s effectiveness in integrating lightweight 11 deep learning and dataset enhancement for robust perception in intelligent vehicle systems, 12 supporting AI-driven autonomous driving and driver assistance applications.

Abstract: Potyviruses express their genome as a single large open reading frame, translated into a polyprotein that is post-translationally cleaved by three virus-encoded proteases into 10 functional proteins. Several of these potyviral proteins, including nuclear inclusion protein b (NIb), are multifunctional. Here, using the classic GFP silencing in Nicotiana benthamiana gfp-transgenic plants model, we show that NIb, in addition to its canonical role as the viral RNA-dependent RNA polymerase (RdRP), functions as a suppressor of RNA silencing. Mutational analyses revealed a previously unreported NIb nuclear localization signal (NLS) consisting of a triple-lysine motif, and NIb suppression of RNA silencing activity was lost when this novel NLSs were mutated, suggesting that nuclear localization is required for NIb suppression of RNA silencing activity. Analysis of sequenced GFP siRNAs revealed three prominent siRNA hotspots at ≈nt 175, ≈320–330, and 560–700 nt. These data showed that there were differences in the positional distribution of the siRNAs between samples expressing NIb and those expressing NIbDel3x2, the NIb null mutant that does not suppress RNA silencing. However, there was no increase in the transcript‑wide siRNA burden between the two treatments. Furthermore, NIb was found to interact with four key RNA silencing pathway proteins—AGO4, HSP70, HSP90, and SGS3. Except for HSP90, each of these proteins showed degradation products that were absent in NIb mutants that did not suppress RNA silence. These findings support a role for NIb in countering host defense during virus infection.

Abstract:

Kenya’s recent achievement of eliminating Human African Trypanosomiasis (HAT) as a public health problem, as validated by the World Health Organization, provides a critical model for the systematic defeat of complex zoonotic diseases. This success, marking the end of indigenous transmission since 2018, emerged from a century-long struggle against the disease in historically endemic foci like the Lambwe Valley, where it severely constrained socio-economic development. The elimination architecture rested on three synergistic pillars: a decentralized, community-based surveillance system that transformed health workers into frontline epidemiological sensors; the operationalization of a One Health framework through the Kenya Tsetse and Trypanosomiasis Eradication Council (KENTTEC), enabling integrated vector control and livestock treatment; and a robust post-elimination vigilance system integrated into national surveillance. Key innovations included the deployment of rapid diagnostic tests, targeted use of insecticide-treated targets, and the adoption of oral therapeutics like fexinidazole. Kenya’s experience demonstrates that sustained elimination of a zoonosis requires transitioning from isolated interventions to a coordinated, cross-sectoral system. The model offers technically replicable and economically justifiable lessons for global efforts targeting the 2030 roadmap goals for neglected tropical diseases. This case proves that with strategic integration, political commitment, and sustained vigilance, the eradication of long-standing zoonotic threats is an achievable goal.

Abstract: Mesonephric-like adenocarcinoma is a rare uterine neoplasm recognized in the fifth edition of the World Health Organization Classification of Female Genital Tumors and poses diagnostic challenges due to its morphologic and immunophenotypic overlap with other endometrial carcinomas. We retrospectively reviewed three cases diagnosed at our institution, evaluating clinical presentation, histologic features, immunohistochemical profile, and available molecular findings, and performed a literature review of approximately 200 reported cases with emphasis on comorbidities and potential predisposing factors. All three patients presented with abnormal uterine bleeding and had hypothyroidism, and one patient had Down syndrome, representing the first reported case in this population. Tumors consistently expressed thyroid transcription factor 1, showed variable GATA3 expression, and demonstrated limited or absent hormone receptor expression. Literature review revealed frequent association with Müllerian-type lesions and recurrent KRAS mutations, while thyroid disease was not identified as a consistent comorbidity. These findings support a Müllerian origin for mesonephric-like adenocarcinoma and suggest that hypothyroidism is unlikely to represent a defining association. The occurrence in a patient with Down syndrome raises the possibility of an underlying genetic susceptibility that warrants further investigation.

Abstract: Background: Neuropathic pain is increasingly recognized as a significant con-tributor to chronic pain and reduced quality of life in patients with rheumatoid arthritis (RA). However, its clinical correlates and relationship with disease ac-tivity and patient-reported outcomes in RA remain incompletely understood. Objective: To evaluate the prevalence of likely neuropathic pain features in RA patients and to investigate their associations with disease activity, pain intensity, fatigue, sleep quality, and health-related quality of life. Methods: In this cross-sectional study, 160 RA patients were enrolled. Neuro-pathic pain features were assessed using the PainDETECT questionnaire. Dis-ease activity was evaluated with the Disease Activity Score in 28 joints (DAS28). Pain intensity, fatigue, sleep quality, and health-related quality of life were as-sessed using the visual analog scale (VAS), Functional Assessment of Chronic Illness Therapy–Fatigue (FACIT-F), Pittsburgh Sleep Quality Index (PSQI), and Short Form-36 (SF-36), respectively. Group comparisons, correlation analyses, and multivariate linear regression models were performed to identify factors associated with neuropathic pain. Results: Likely neuropathic pain was identified in 22.5% of patients. Patients with neuropathic pain had significantly higher pain intensity, greater fatigue, poorer sleep quality, and lower SF-36 scores across all domains compared with patients without neuropathic pain (all p < 0.001). Although DAS28 scores were higher in patients with neuropathic pain, PainDETECT scores correlated more strongly with VAS pain in-tensity (ρ = 0.679, p < 0.001) than with DAS28 (ρ = 0.536, p < 0.001). PainDETECT scores were negative-ly correlated with FACIT-F (ρ = −0.512, p < 0.001) and multiple SF-36 domains, indicating substantial quality-of-life impairment. In multivariate regression analysis, only pain intensity independently predicted neuropathic pain features. Conclusions: Neuropathic pain is common in RA and is more closely associated with pain intensity and patient-reported outcomes than with inflammatory activ-ity alone. Routine assessment of neuropathic pain features may facilitate more personalized and effective pain management strategies in RA, highlighting the clinical relevance of targeting neuropathic mechanisms alongside inflammation.

Abstract: Climate change leads to reservoir management challenges especially in areas with high risk of drought and flood. Traditional reservoir rules curves are inappropriate for addressing variation of reservoir inflow. This study presents an integration framework between GCMs from CMIP6 (ACCESS-CM2, MIROC6, and MPI-ESM1-2-LR) under SSP245 and SSP585 scenarios and WEAP, which is validated in accuracy for reservoir inflow and storage capacity. This integration contributes to Hippopotamus Optimization (HO), a technique used to develop Resilience Reservoir Rule Curve (RRRC) for Ubonrat reservoir during 2024-2055 employing dual-objective function that emphasizes the reduction of water shortage and water excess. The results indicate that RRRC developed by HO is more efficient and suitable than Honey Bee Mating Optimization (HBMO) and existing rule curve. After testing the RRRC with historical inflow and future inflow from three GCMs under SSP245 and SSP585, it can reduce average water shortage and demonstrate outstanding efficiency in water excess management. This potential reflects its adaptability under future variation of hydrological condition. This crucial finding illustrates that the integration framework can develop resilient rule curves under uncertainty. HO integrated with various models can be implemented as an optimal framework and have high potential for reservoir operation planning under climate change. The developed methodology can be implemented in other reservoirs to gather additional factors for sustainable promotion of water resource resilience.

Abstract:

This study aimed to analyze the effect of reduced water potential on the imbibition curve and triphasic pattern of seeds in several Solanaceae species. The experiment was conducted at the Seed Physiology and Health Laboratory and the Seed Biology and Biophysics Laboratory, Department of Agronomy and Horticulture, Faculty of Agriculture, Bogor Agricultural University, from April to September 2025. The study used seeds from three Solanaceae crops—chili (Capsicum annuum L., varieties Simpatik and Sempurna), tomato (Solanum lycopersicum L., varieties Niki and Rempai), and eggplant (Solanum melongena L., varieties Tangguh and Provita). The seeds were subjected to various levels of osmotic stress using polyethylene glycol (PEG 6000) to simulate water potentials of 0.00, –0.30, –1.90, and –4.10 MPa. Lower water potential in the growing medium reduced the seed’s ability to absorb the water. The triphasic pattern consistently appeared only in chili seeds, whereas in tomatoes and eggplants, it varied across varieties and water potential conditions. The lower water potential made the later the phase I ended, and the longer the phase II lasted. These findings confirm that the standard imbibition pattern cannot be generalized to all seeds, and therefore, the imbibition response is specific to seed type, variety, and germination environment.

Abstract: Background: Benign prostatic hyperplasia (BPH) is a significant health issue among ageing men, with ongoing research focused on elucidating its underlying mechanisms and improving experimental models. Testosterone Propionate (TP) is the first line of choice for the induction of BPH in experimental rodent models. However, TP's controlled status as a Schedule III drug in the United States and a Class C drug in the UK presents challenges in obtaining TP for experimental use, giving preference to the sulpiride model since it is easily obtained as an alternative for the induction and study of BPH. Method: A comprehensive literature search was conducted across multiple electronic databases, including PubMed/MEDLINE, Embase, and Web of Science. The primary PubMed search strategy included combinations of Medical Subject Headings (MeSH) and free-text terms: (“Benign prostatic hyperplasia induction” OR ‘’ and rodent models’’) AND (“Testosterone Propionate model” ) AND (“sulpiride model”). Studies were included if they induced BPH (using testosterone or sulpiride models). Titles and abstracts were screened for relevance; eligible articles underwent full-text review, with data extracted thematically. No formal risk-of-bias scoring was used due to the narrative approach; instead, studies were appraised by design, rigor, plausibility, and evidence. This study reviewed published and publicly available data, so no ethical approval was required. Results: Although both TP and sulpiride induce BPH via various mechanisms, this review provides a comparative analysis of these two commonly utilised models for studying BPH. In the TP approach, castrated rodents receive daily subcutaneous injections for 4 weeks, resulting in dihydrotestosterone (DHT)-mediated epithelial hyperplasia predominantly affecting the ventral prostate lobes. Conversely, the sulpiride model is non-invasive, employs intact animals treated with sulpiride, and induces hyperprolactinemia-mediated BPH via interactions with androgen and oestrogen receptor pathways that stimulate prostatic stromal and epithelial proliferation, particularly in the lateral and dorsal lobes, representing an alternative method. We also highlight the strengths and limitations of TP and sulpiride in replicating clinical symptoms and examine the toxicological effects of sulpiride on the kidney, testis, liver, and brain. Conclusions: we recommend the sulpiride model for the induction and studying of BPH, as it is readily accessible and closely mimics the pathogenesis of BPH in humans, unlike the TP model, which requires castration.

Abstract: The Next-Generation Security Triad—integrating post-quantum cryptography (PQC), Zero Trust Architecture (ZTA), and AI security—provides comprehensive protection for autonomous sensing systems. However, existing frameworks assume enterprise connectivity is available in tactical environments operating under Disconnected, Intermittent, and Low-bandwidth (DIL) conditions. This paper presents the Tactical Edge Triad Architecture (TETA), adapting enterprise substrate components for disconnected operations through five modules: Edge Cryptographic Module (ECM), Tactical Identity Cache (TIC), Edge Analytics Engine (EAE), Mission Policy Store (MPS), and the Autonomous AI Governance Framework (AAGF). Three mechanisms address DIL-specific challenges: Authority Decay provides a DIL-specific operationalization of continuous verification through progressive privilege reduction with formal attack mitigations; Pre-Mission Consensus Packaging provides cryptographically signed governance envelopes satisfying human oversight requirements; and Triad Integration demonstrates cross-pillar security dependencies. The AAGF systematically adapts established governance mechanisms, behavioral envelopes, watchdog models, autonomy-downgrade, and consensus-backed approval for disconnected operations. Analytical evaluation across two tactical scenarios demonstrates feasibility: PQC overhead estimates derive from published pqm4 benchmarks; governance function estimates (policy evaluation, watchdog inference, audit logging) are engineering projections based on comparable embedded workloads. Combined governance latency is estimated at ~15 ms on Cortex-A53 class processors (±40%), with 0.5% steady-state bandwidth increase for PQC. TETA enables Triad implementation at the tactical edge while preserving security properties and governance accountability.

Abstract: This article examines the development of operator splitting methods in Soviet numerical analysis during 1955–1975, with particular focus on N.N. Yanenko’s formalization of the Method of Fractional Steps at the Siberian Branch of the USSR Academy of Sciences. While similar techniques were independently developed in the West (Peaceman-Rachford 1955, Douglas-Rachford 1956), the Soviet school pursued a distinct trajectory shaped by acute hardware constraints and deep epistemological commitments to operator theory. Through analysis of technical publications, archival materials, and comparative historiography, this study argues that material scarcity catalyzed a systematic research program emphasizing computational economy, while a pre-existing mathematical culture valorizing theoretical elegance reinforced this trajectory. The case illuminates how geopolitical constraints and intellectual traditions jointly shaped algorithmic innovation, contributing to methods that ironically became foundational for modern massively parallel computing. Significant archival gaps limit definitive claims about industrial applications, highlighting the need for further primary source research.

Abstract:

Background: The incidence of patients with nontuberculous mycobacterial pulmonary disease (NTM-PD) complicated by chronic pulmonary aspergillosis (CPA) has been increasing. CPA is known to be associated with complex treatment regimens and a poor prognosis. However, data from mainland China remain scarce. Objective: This single-center retrospective study aimed to evaluate the clinical characteristics, risk factors, and prognoses of patients with NTM-PD who were coinfected with CPA. Methods: We conducted a retrospective review of the medical records of 248 patients diagnosed with NTM-PD. Risk factors for CPA were analyzed via multiple logistic regression, followed by survival analysis. Results: Among the 248 patients with NTM-PD, 66 (26.6%) were diagnosed with CPA. Independent risk factors for NTM-PD and CPA coinfection included male sex(OR 2.13, 95% CI:1.03-4.47), dyspnea(OR 27.9, 95% CI:4.24-570), cavity(OR 5.95, 95% CI:2.76-13.9), use of oral corticosteroids(OR 4.28, 95% CI:1.13-16.6), and interstitial lung disease(OR 15.5, 95% CI:1.89-361). The Kaplan-Meier survival curves indicated a significant divergence between the NTM-PD group and the NTM-PD with CPA group (log-rank test, p = 0.00039; HR 2.01, 95% CI:0.66-6.12). Conclusion: In patients with NTM-PD, the presence of concurrent CPA is associated with a marked increase in mortality. Clinicians should maintain a high index of suspicion for CPA to ensure prompt diagnosis and treatment, particularly in high-risk individuals.

Abstract: Stroke has been a topic of extensive research due to its debilitating consequences and high mortality. New findings offer a deeper understanding of specific factors that affect post-stroke recovery and identify therapies that may facilitate this process. One such factor was microglia, neuronal immune cells that are highly reactive to cytokines in the neuroenvironment and can, in turn, affect the inflammatory cascades that originate after stroke, making them ideal candidates for immunomodulation in the brain. Many FDA-approved immunotherapies have been found to target distinct inflammatory signaling molecules and responders, including IL-6 inhibitors, IL-13 inhibitors, IL-12/IL-23 inhibitors, B-cell modulators, Type I interferon inhibitors, CAR T-cell therapy, Calcineurin inhibitors, Complement inhibitors, and JAK-STAT pathway inhibitors. The FDA-approved immunotherapies discussed in this review demonstrate potential in modulating the immune response after stroke by targeting key inflammatory pathways involved in secondary brain injury. Future research should focus on defining optimal therapeutic windows, identifying suitable patient populations, determining the most appropriate timing of therapy, and targeting specific immune mechanisms to balance the attenuation of harmful inflammation with the preservation of reparative processes.

Abstract: Poly-γ-glutamic acid (γ-PGA) is an important biopolymer produced by various species of Bacillus. Novel γ-PGA producers have shown strain-dependent nutritional and culture requirement that must be characterised and optimised to improve γ-PGA yields. The optimal nutritional and cultural condition for maximum γ-PGA titre in a newly identified γ-PGA producing strain Bacillus licheniformis DPC6338 was determined using one factor at a time (OFAT) and design of experiments (DOE). The optimal nutritional and culture condition for maximum γ–PGA titre in B. licheniformis DPC6338 was 67g/L glutamic acid, 32g/L tryptone, 75g/L glucose, 5g/L citric acid, 2g/L K₂HPO₄, 0.5g/L MgSO₄·7H₂O, 0.02 g/L FeCl₂·4H₂O, 0.1g/L CaCl₂·2H₂O, 0.5 g/L MnSO₄·H₂O, 2g/L ZnSO₄·7H₂O, 42°C, pH 6.5 – 7.0, 1% inoculum, at 250 rpm. Under optimised conditions in shake flask, maximum γ–PGA titre 75.35 ± 0.38 was obtained after 96h while peak productivity of 1.3 g/L/h occurred at 48 h, representing a 27% and 4% improvement in titre and productivity compared to the screening medium. Scale-up to bioreactor conditions significantly enhanced the final titre γ-PGA and early-phase volumetric productivity by ~30% and ~80% respectively. The results obtained in this study highlight the potential of B. licheniformis DPC6338 for industrial γ-PGA producing strain.

Abstract: Financial statement fraud continues to pose a significant challenge to audit effectiveness, investor confidence, and the integrity of financial markets. Fraud detection is particularly complex due to the highly imbalanced nature of financial reporting data, where fraudulent observations constitute only a small fraction of the total sample. In such settings, conventional accuracy-based evaluation often produces misleading conclusions and fails to reflect practical audit value. This study conducts a comparative evaluation of four deep learning models, namely LSTM, GRU, CNN1D, and Transformer, for financial statement fraud detection under class-imbalanced conditions. The analysis is based on a dataset of 805 firm-year observations. It adopts Precision–Recall Area Under the Curve as the primary performance metric, complemented by ROC-AUC, Precision, Recall, F1 score, and Specificity. To assess practical usability, Decision Curve Analysis is employed to evaluate the decision-level net benefit of each model across different threshold probabilities, and bootstrap resampling is used to assess performance stability under random data partitioning. The empirical results show that the Transformer model consistently outperforms the other architectures in terms of discriminative ability, robustness, and decision-level utility. Its attention-based structure enables effective modeling of global relationships among financial indicators, leading to stable performance across varying thresholds and data splits. The CNN1D model demonstrates relatively high specificity and a balanced error structure, suggesting its suitability in audit environments where minimizing false positives and controlling verification costs are critical. In contrast, although the LSTM and GRU models exhibit higher sensitivity to fraudulent cases, their lower precision and stability limit their effectiveness as standalone solutions. Overall, the findings emphasize the importance of imbalance-aware, decision-oriented evaluation frameworks for detecting financial statement fraud. The study offers practical insights for auditors and regulators by identifying deep learning models that combine statistical reliability with operational relevance in real-world auditing contexts.

Abstract: Temporal reasoning is an important part of the field of time geography. Recent advances in qualitative temporal reasoning have developed a set of 74 relations that apply between discretized time intervals. While the identification of specific relations is important, the field of qualitative spatial and temporal reasoning relies on conceptual neighborhood graphs to address relational similarity. This similarity is paramount for generating essential decision support structures, notably reasonable aggregations of concepts into single terms and the determination of nearest neighbor queries. In this paper, conceptual neighborhoods graphs of qualitative topological changes in the form of translation, isotropic scaling, and anisotropic scaling are identified using a simulation protocol. The outputs of this protocol are compared to the extant literature regarding conceptual neighborhood graphs of the Allen interval algebra, demonstrating the theoretical accuracy of the work. This work supports the development of robust spatio-temporal artificial intelligence as well as the future development of spatio-temporal query systems upon the spatio-temporal stack data architecture.

Abstract:

We devised a quantitative scoring function to assess the cumulative effects of nonsynonymous single nucleotide variants (SNVs) on protein-coding genes in patients with ovarian cancer (OvCa) and thyroid cancer (ThCa). The goal is to find novel candidate cancer-related genes for downstream bioinformatics analyses and wet-lab studies. With Genomic Data Commons as primary data resource, SNV information was extracted from whole-exome sequencing data from patients with these cancers. A cumulative variant scoring function, Q(G) was developed to sum up the deleterious effects of the individual SNVs on the gene G. While Q(G) can be computed using any popular functional effect analyzers such as FATHMM-XF, SIFT, PolyPhen, and CADD, we have also established an integrative scoring function iQ(G) that combines the deleterious assessments from different analyzers and demonstrated that iQ(G) is a more effective method for identifying likely cancer-related genes. Based on the iQ(G) rankings, the top three novel genes for OvCa are AHNAK2, UNC13A, and PCDHB4; and those for ThCA are PLEC, HECTD4, and CES1. Furthermore, the top 1% genes with highest iQ(G) scores for each cancer were submitted for KEGG pathway analysis. The results revealed that several genes of the CACNA1 family within the type II diabetes mellitus pathway are likely related to both OvCa and ThCa and suggested other molecular interactions that should be further studied in connection with OvCa prognosis and ThCa treatment.