Search Results (8,301)

Magnetic resonance imaging (MRI) remains one of the most important tests for diagnosing and monitoring various diseases. In recent years, machine learning methods have been widely applied to automate MRI analysis. It supports decision-making by predicting disease and highlighting relevant regions. However, the proper use of feature extraction methods can improve the performance of the model. This paper proposes a WaveletFusion architecture that combines a two-dimensional Haar wavelet decomposition with a convolutional neural network (CNN) for classification. The approach was demonstrated on the Brain Tumor MRI dataset and further examined on the Br35H :: Brain Tumor Detection 2020 (Br35H). The model decomposes each MRI slice into approximation and directional detail subbands and fuses multi-scale wavelet features within the convolutional pipeline. To evaluate the effect of decomposition depth, WaveletFusion variants from one to eight levels were compared with a Baseline CNN model under the same training protocol. The results showed that performance improved progressively with increasing decomposition depth up to level 7, whereas the 8-level configuration consistently declined, indicating that excessive decomposition introduces information loss and over-compression in the deepest approximation pathway. The best-performing configuration, which outperformed both the Baseline CNN and the WaveletFusion variations in five independent runs, was the 7-level WaveletFusion model, achieving a test accuracy of 0.94 ± 0.01 and test macro-F1 of 0.93 ± 0.02. A similar tendency was observed on the Br35H dataset, where the 7-level model achieved a 0.97 ± 0.01 test accuracy and 0.97 ± 0.01 test macro-F1, while the 8-level configuration remained weaker on both datasets. These results show that multi-scale wavelet fusion can improve Brain Tumor MRI classification while maintaining a compact model size and a fair comparison setting, and that the decomposition depth must be selected carefully. Full article

The Papanicolaou (Pap) smear, a cornerstone of cervical cancer prevention, has emerged as a compelling, though unconventional, biospecimen for the detection of ovarian cancer (OC). This structured narrative review synthesizes the evolving evidence on the utility of cervicovaginal cytology and molecular analysis of Pap test material for OC detection. While conventional cytology provides a proof of concept, its sensitivity is low, ranging from incidental detection of OC in 0.004% of routine screens to 19.3% in patients with known OC. Specific cytologic findings, however, carry significant predictive value: atypical glandular cells (AGC) confer a two-fold increased OC risk, and psammoma bodies (PB) are strongly associated with serous malignancies. Driven by the sensitivity limitations of morphology, the field has undergone a paradigm shift towards molecular detection. Foundational studies confirmed tumor-derived DNA, including hallmark TP53 mutations, is detectable in Pap samples years before diagnosis, though sensitivity is constrained by low DNA abundance and confounded by background clonal mutations. To overcome this, strategies have expanded to target broader genomic signatures, such as somatic copy number alterations (EVA test: 75% sensitivity, 96% specificity), and multi-gene mutation panels (PapSEEK: 33–45% sensitivity). The most promising advances lie in multi-omic approaches, particularly DNA methylation biomarkers, which have demonstrated sensitivities up to 81% with high specificity. Collectively, this evidence argues against repurposing the Pap test as a standalone OC screen but supports its strategic integration into a risk-stratified clinical algorithm. We propose a “reflex-to-molecular” model where high-risk cytology (e.g., AGC, PB) automatically triggers advanced molecular testing on the same sample. This model efficiently leverages existing infrastructure to triage high-risk women for definitive diagnostics. Prospective validation of this integrated approach is the essential next step toward transforming this test into a sentinel for malignancies of the upper female reproductive tract. Full article

Background/Objectives: Oral squamous cell carcinoma (OSCC) exhibits substantial biological heterogeneity, and current clinicopathological risk stratification incompletely reflects tumor metabolic behavior. Stable isotope ratio mass spectrometry enables the quantitative assessment of carbon and nitrogen isotopic composition, potentially capturing cumulative metabolic reprogramming associated with tumor aggressiveness. This study evaluated whether isotopic signatures of tumor tissue and surgical margins are associated with lymph node metastasis and survival outcomes in OSCC. Methods: In this prospective study, 54 consecutive patients undergoing primary surgical treatment for OSCC were enrolled. Paired samples derived from tumor tissue and surgical margins were analyzed using isotope ratio mass spectrometry to determine the relative abundance of nitrogen-15 and carbon-13 isotopes. The primary endpoint was pathological lymph node metastasis. Secondary endpoints included disease-free survival and overall survival. Paired comparisons were performed using Wilcoxon signed-rank tests with false discovery rate correction. Logistic regression models for nodal metastasis were constructed using Firth penalization with bootstrap internal validation, while survival outcomes were evaluated using Cox proportional hazards models with model complexity restricted according to the number of events. Results: Tumor tissues demonstrated significantly lower δ¹³C and δ¹⁵N values and higher nitrogen-to-carbon ratios compared with surgical margins (all adjusted p < 0.05). In multivariable analysis, tumor δ¹⁵N was independently associated with lymph node metastasis and modestly improved model discrimination. However, it was not independently associated with disease-free or overall survival. Exploratory analyses indicated that higher δ¹³C values in surgical margins were independently associated with shorter disease-free survival. Conclusions: These findings suggest that isotope ratio mass spectrometry-based isotopic profiling identifies reproducible metabolic differences between tumor and margin tissues in OSCC. Tumor δ¹⁵N is associated with lymph node metastasis, whereas margin δ¹³C may reflect recurrence risk and potentially capture metabolic field effects. These findings are hypothesis-generating and warrant validation in larger, independent cohorts. Full article

Dynamic knowledge representation in curved manifolds conventionally relies on integer-order Markovian sequence encoders, intrinsically yielding exponential memory decay. This paradigm fails to model the anomalous diffusion and heavy-tailed historical dependencies inherent in complex evolutionary networks and dense physical environments. This manuscript proposes the Fractional–Temporal Lorentz Graph Convolutional Network (FTL-GCN), formalizing temporal evolution as a continuous fractional geometric flow explicitly defined on the tangent bundle of the Lorentz manifold. Analytical derivations demonstrate that the discrete Grünwald–Letnikov memory kernel establishes a non-exponential, power-law lower bound for historical state retention, preventing topological manifold collapse over extended temporal horizons. Empirical evaluations demonstrate that FTL-GCN achieves competitive forecasting accuracy against the latest 2025–2026 state-of-the-art discrete models within specific temporal windows, while uniquely mitigating predictive degradation by up to 52% in long-horizon dependency stress tests and maintaining sub-millisecond latency for physical control. The architecture is subsequently deployed within an in silico biophysical simulation for autonomous micro–nano robotic navigation in the Tumor Microenvironment (TME). By establishing a physical-mathematical structural analogy—mapping the empirical fractional viscoelasticity of the extracellular matrix to the cognitive network’s fractional derivative order—FTL-GCN sustains continuous-space navigation policies in dense anomalous environments where standard integer-order models experience mechanical slip. Full article

Over the past 200 years (1820–2020), global life expectancy has nearly tripled, increasing from 26 to 72.91 years, due to factors such as poverty reduction and public health initiatives. Today, society faces different challenges than it did centuries ago. In patient care and healthcare system priorities, the goal is to develop smart, feasible, long-lasting, cost-effective, readily available, adverse-reaction-free, adaptable, and personalized solutions that minimize patient discomfort, reduce caregiver effort, and decrease hospitalization duration and costs. In this context, biomaterials serve as versatile tools capable of performing a wide range of diagnostic, therapeutic, and theragnostic functions. Thanks to their biocompatibility, biodegradability, surface chemistry, and responsiveness, biomaterials are currently addressing issues such as patient compliance (through controlled drug-delivery systems and smart wound dressings), long transplant waiting lists, transplant rejection, non-adaptable prosthetics (artificial organs), oncology treatment efficacy (nano-formulations for theragnostics and multiple tumor targeting), and inconsistent in vitro drug-testing models (organs-on-a-chip). In this review, we focus on biomaterials’ smartness, then explore databases for efficient product design, and finally highlight their applications in the biomedical field, especially in drug delivery, tissue engineering, and regenerative medicine. Full article

Background: Growth hormone deficiency (GHD) is one of the most common endocrine sequelae in survivors of pediatric medulloblastoma, largely resulting from hypothalamic–pituitary irradiation. Concerns regarding the oncologic safety of growth hormone (GH) replacement have historically limited its use. This study aimed to evaluate growth response to GH therapy and its potential association with tumor relapse in medulloblastoma survivors treated between 2000 and 2023. Methods: We conducted a retrospective single-center cohort study including 74 patients diagnosed with medulloblastoma before 18 years of age. GHD was confirmed by stimulation testing according to standard criteria. Auxological, endocrine, and oncologic data were collected longitudinally. Growth outcomes were compared among patients without GHD (n = 38), patients with untreated GHD (n = 13), and patients with GHD receiving GH treatment (n = 23). Relapse rates were assessed following GH initiation and compared with those of untreated patients. Results: GHD was diagnosed in 48.7% of patients. Baseline height SDS did not differ among groups. Patients with untreated GHD experienced a significant decline in height SDS (−1.93 ± 0.78), whereas GH-treated patients showed a significant increase (+0.39 ± 0.06; p < 0.0001). Final height SDS was significantly lower in untreated GHD patients (−2.45 ± 0.36) compared with GH-treated patients (−1.71 ± 0.68) and patients without GHD (−0.68 ± 0.24; p < 0.0001). No evidence of an increased risk of tumor relapse was observed in association with GH therapy during follow-up. Conclusions: GH replacement significantly improves growth outcomes in medulloblastoma survivors with confirmed GHD without apparent increase in relapse risk when initiated after stable remission. The early identification and multidisciplinary management of GHD are essential components of long-term survivorship care. Full article

Cryptorchidism is the most prevalent congenital anomaly of the male genitourinary tract, with an incidence of approximately 1 to 9 percent in full-term male infants, decreasing with age due to spontaneous descent. It encompasses testes that fail to descend into the scrotum, which may be intra-abdominal, inguinal, or ectopic, and can be associated with syndromic, genetic, or environmental factors. The descent process occurs in two phases: intra-abdominal, driven by gubernacular development and androgen-independent mechanisms, and inguinoscrotal, regulated by hormonal and mechanical factors including androgens and the gubernaculum. Clinically, cryptorchidism manifests as absent or hypoplastic scrotal testes, often with inguinal fullness. Palpation and physical examination are primary diagnostic tools, with imaging such as ultrasound or MRI reserved for specific cases. Surgical exploration remains the definitive diagnostic modality, especially for nonpalpable testes. Early referral, ideally before 12 months of age, is essential for timely orchidopexy, which aims to position the testes within the scrotum to reduce risks of torsion, trauma, subfertility, and malignancy. Hormonal therapy shows limited efficacy and is generally not recommended as a primary treatment modality. Long-term outcomes indicate that early orchidopexy improves spermatogenic potential and fertility. Men with a history of cryptorchidism exhibit elevated risks of subfertility and testicular germ cell tumors, with the risk being higher if surgical correction is delayed or if testes remain intra-abdominal. The increased malignancy risk persists even after orchidopexy, underscoring the importance of vigilant surveillance. Management strategies emphasize a multidisciplinary approach, combining surgical intervention with ongoing monitoring, to optimize functional and oncological outcomes. Early diagnosis, appropriate surgical treatment, and patient education are critical components in minimizing long-term complications associated with cryptorchidism. Full article

Robust brain tumor classification from magnetic resonance imaging (MRI) remains challenging due to complex structural heterogeneity and subtle inter-class variability. Beyond predictive accuracy, conventional convolutional neural networks predominantly rely on texture-dominant features and fixed receptive fields, which may limit the extraction of clinically meaningful structural information. This study proposes a morphology-aware multi-scale deep representation learning framework that embeds morphological inductive bias directly within hierarchical feature extraction. The proposed architecture synergistically integrates trainable morphological operations with multi-scale convolutional feature learning inside a unified residual framework, supported by an in-block morphological refinement mechanism and a morphology-aware downsampling module. Unlike prior approaches that treat morphological operators as preprocessing or auxiliary branches, the proposed design incorporates differentiable dilation and erosion into the core feature hierarchy to guide structure-aware representation formation. The model was evaluated using five-fold cross-validation and an independent test set, achieving an overall test accuracy of 99.31% with consistently high macro-averaged precision, recall, F1-score, and AUC values. Grad-CAM analysis further demonstrates that the learned representations emphasize clinically relevant tumor regions, supporting interpretable structural knowledge extraction. Ablation studies confirm that performance improvements arise from the synergistic integration of multi-scale learning and morphology-aware refinement. Overall, embedding structural inductive bias within multi-scale deep representation learning enhances robustness, stability, and interpretable knowledge extraction for brain tumor MRI analysis. Full article

Wilms tumor (WT), the most common kidney neoplasm in children, is closely associated with hereditary factors. This study included 134 WT patients (62 males, median age of 7 years, age at diagnosis of 24.9 months) with unilateral (n = 90, 67%) or bilateral WT (n = 44, 33%). Genetic testing was performed using targeted sequencing of 415 genes and multiplex ligation–dependent probe amplification (MLPA). Twenty-five mutations in eight genes were found in 17% (n = 23) of patients: WT1 (n = 10), TRIM28 (n = 4), REST (n = 3), CHEK2 (n = 3), BRCA2 (n = 2), NF1 (n = 1), RAD50 (n = 1), and CDC73 (n = 1). Large deletions of the 11p13 region were revealed in 6% (n = 5) of patients. The 11p15 locus methylation was studied in blood, tumor, and healthy kidney tissue of nine patients suspected of Beckwith–Wiedemann syndrome (BWS) using methylation-sensitive MLPA (MS–MLPA). BWS was diagnosed in 3% (n = 4) of cases (one patient had mosaic disease). Thus, genetic and epigenetic aberrations were identified in 32 WT patients (24%). These patients had a higher frequency of bilateral WT and a higher rate of abnormalities compared to patients without aberrations (56% vs. 25%, p = 0.002; and 86% vs. 25%, p < 0.0001, respectively). The detection of WT hereditary predisposing factors is crucial for treatment strategies and long-term patient surveillance. Full article

Dpep is a cell-penetrating peptide that targets transcription factors ATF5, CEBPB and CEBPD to selectively suppress growth and survival of diverse tumor cell types in vitro and in vivo. Due to these actions and its apparent safety, the peptide has potential as a cancer therapeutic. How Dpep might be combined with other anti-cancer agents to achieve synergistic efficacy and to overcome possible peptide resistance has not been assessed in depth. Based on prior work indicating that Dpep promotes apoptotic cancer cell death and up-regulates multiple pro-apoptotic and tumor suppressor genes, we studied combinations of Dpep with ABT-263, a pro-apoptotic BCL2 family inhibitor, and decitabine, a hypomethylating drug. Combining Dpep with each agent alone or together synergistically suppressed the growth of a range of solid and liquid tumor cell types. Moreover, the combinations synergistically inhibited the growth of cells lines that were selected either in vivo or in vitro for Dpep resistance. Finally, we tested the combination of Dpep with ABT-263 in a mouse melanoma xenograft model. The combination more effectively inhibited tumor growth than either agent alone and, in contrast to vehicle or ABT-263, produced a 40% durable survival rate. Taken together, these observations highlight potential drug partners for the therapeutic development of Dpep. Full article

Automated brain tumor classification from MRI scans requires optimized CNN architectures deployable on embedded FPGA platforms. This paper presents an integrated approach combining the Multi-Chaotic Enhanced HEOA (MC-HEOA) for automatic CNN architecture discovery with deployment validation on a Xilinx Zynq-7000 FPGA. A CEC2023 benchmark across 10 test functions evaluates 6 chaotic maps and selects the Tent map as the optimal diversity generator. The NAS search space spans a massive combinatorial space of $1.31 \times {10}^{16}$ configurations encoding architectural choices (layers, convolutions, channels, pooling) under a strict constraint of fewer than one million parameters for FPGA compatibility. The optimal discovered architecture, trained and evaluated using single-channel grayscale input (224 × 224 × 1)—the natural representation for intrinsically monochromatic MRI data— achieves 91.33% test accuracy and 92.44% validation accuracy with 724,200 parameters on the 4-class Brain Tumor MRI dataset (glioma, meningioma, pituitary, no tumor). HLS synthesis on the Zynq-7000 (xc7z020clg484-1) validates embedded deployment feasibility, with DSP utilization of 16%, LUT utilization of 57%, FF utilization of 28%, and an inference latency of 374 ms at 100 MHz. This study demonstrates the effectiveness of MC-HEOA for discovering compact, high-performing CNN architectures compatible with FPGA deployment, opening new perspectives for real-time embedded medical diagnosis. Full article

Background: Carbonic Anhydrases (CAs) represent regulators of cell adaptation to hypoxia, pH regulation, and metabolic fitness. Among cancers, multiple myeloma (MM) is a plasma cell malignancy sustained by hypoxia-driven metabolic adaptation, extracellular acidification, and redox imbalance. Tight regulation of tumor extracellular pH, mediated by Carbonic Anhydrases IX and XII, is crucial for myeloma survival, progression, and stemness, making these isoforms attractive therapeutic targets. Methods: We designed and synthesized a library of terpenoid-based hybrids by derivatizing chlorothymol and 4-isopropyl-3-methylphenol with either the natural coumarin umbelliferon or the 2,2′-dipicolylamine (DPA) scaffold. This chemical strategy aimed to selectively inhibit tumor-associated CAs IX/XII through coumarin- or DPA-mediated recognition, while terpenoid fragments were introduced to enhance lipophilicity, membrane permeability, and potential redox-modulating properties. The compounds were tested by a Stopped-Flow assay for CA inhibition, in cell-based assays for antiproliferative properties and by means of several antioxidant assays. Results: The most active compounds, connecting the coumarin core to a terpenoid tail, inhibited the targeted CAs in the nanomolar range, showing up higher selectivity over off-target isoforms (I and II). In studies performed on MM cell lines, selected derivatives reduced viability (IC₅₀ = 15.8–85.4 µM) and displayed favorable selectivity over normal cells. In silico investigations suggested that the compounds were able to interact selectively with the target enzymes. Conclusions: Collectively, these results support a dual-targeting strategy in which selective inhibition of tumor-associated CAs, combined with redox modulation, interferes with adaptive mechanisms of MM cells, providing a rational framework for the development of multifunctional agents against metabolically resilient hematological malignancies. Full article

Oral diseases remain highly prevalent worldwide and require early diagnosis and continuous monitoring to improve clinical outcomes. Conventional diagnostic methods are often invasive, time-consuming, and limited in their capacity for real-time assessment, which has driven the development of biosensor technologies for point-of-care applications. Among these, cell-based biosensors utilize living cells as sensing elements capable of responding to inflammatory mediators, bacterial toxins, metabolic products, and tumor-associated biomarkers. This narrative review summarizes the principles, cell types, detection mechanisms, and applications of cell-based biosensors in oral health. The literature was identified through a structured search of PubMed, Scopus, Web of Science, and Google Scholar using keywords related to cell-based biosensors, oral diagnostics, salivary biomarkers, periodontal disease, oral cancer, and lab-on-chip technologies. Due to the heterogeneity of biosensor designs and detection methods, the selected studies were analyzed qualitatively. Cell-based biosensors have demonstrated applications in periodontal disease detection, cariogenic biofilm monitoring, oral cancer diagnostics, cytotoxicity testing of dental materials, and salivary biomarker analysis. The integration of microfluidic and lab-on-chip systems enables real-time and multiplex detection, supporting the development of chairside diagnostic platforms in dentistry. However, challenges related to standardization, reproducibility, and clinical validation remain and must be addressed to facilitate broader implementation in routine practice. Full article

Background and Objectives: Sentinel lymph node biopsy (SLNB) is increasingly used for high-risk, clinically node-negative cutaneous squamous cell carcinoma (cSCC), yet pathological reporting remains binary, lacking morphological stratification. The prognostic relevance of nodal tumor burden subtypes—isolated tumor cells (ITC), micrometastases, and macrometastases—is well established in melanoma and breast cancer but remains uncharacterized in cSCC. We aimed to describe the morphological spectrum of sentinel lymph node involvement in a consecutive institutional cohort and determine whether primary tumor characteristics predict the extent of nodal colonization. Materials and Methods: We conducted a retrospective-observational study at Clinical Center Niš (Serbia) including 35 consecutive clinically N0 high-risk cSCC patients who underwent SLNB using a dual-tracer protocol (⁹⁹mTc-labeled albumin and methylene blue). Sentinel nodes were processed by serial sectioning with hematoxylin-eosin and pancytokeratin (AE1/AE3) immunohistochemistry. Deposits were classified as ITC (≤0.2 mm), micrometastases (>0.2–2.0 mm), or macrometastases (>2.0 mm). Clinicopathologic predictors were evaluated using the Mann–Whitney U test, Fisher’s exact test, the Kruskal–Wallis test, and the Spearman rank correlation test. Results: SLN involvement was identified in 12 of 35 patients (34.3%). Among positive cases, ITC accounted for 6 patients (50.0%), micrometastases for 5 (41.7%), and macrometastasis for 1 (8.3%)—minimal nodal disease constituting 91.7% of positive findings. No primary tumor feature—including diameter, thickness, grade, perineural invasion, or lesion multiplicity—significantly distinguished ITC from overt metastatic deposits. Patients with ITC showed numerically higher median tumor thickness (8.0 mm) than those with micrometastases (4.0 mm), though this did not reach significance (Kruskal–Wallis p = 0.065). Conclusions: SLN positivity in high-risk cSCC is morphologically heterogeneous, with minimal nodal disease predominating. Primary tumor features do not reliably stratify the extent of nodal colonization. Structured tumor-burden reporting—distinguishing ITC, micrometastases, and macrometastases—should be adopted as standard practice to enable meaningful prognostic comparisons and inform individualized management. Full article

Background: Chimeric antigen receptor (CAR) T cells achieve cure in the therapy of hematological malignancies. In solid tumors, however, CAR-T cells face an immunosuppressive tumor microenvironment (TME) which crucially impedes their cytotoxic capacities. Citrate accumulating in the TME is a crucial metabolite in mediating immune suppression and is consumed by cancer cells promoting growth of various tumors, including melanoma; blocking the citrate transporter pmCiC with gluconate abrogates citrate-mediated tumor growth. Methods: To bolster treatment of melanoma, we explored gluconate as adjuvant for CAR-T cell therapy. Results: First, gluconate did not impair CAR-T cell functional capacities with regard to cytotoxicity, cytokine secretion, and persistence in a “stress test” based on repetitive antigen stimulation with cognate cancer cells. The addition of gluconate antagonized the citrate-mediated enhanced proliferation of melanoma cells. As a consequence, the elimination of citrate-boosted melanoma cells by CSPG4-specific CAR-T cells was augmented in the presence of gluconate. Conclusions: Taken together, these data suggest that counteracting citrate-mediated enhanced tumor growth with gluconate may improve the cytotoxic activity of CAR-T cells against melanoma. Full article