搜索结果：Diagnostics (Basel, Switzerland)

In the original publication [...].

Repeated evidence demonstrates limited reproducibility and accuracy of the visual quantification (VQ) of the tumor cell content (TCC) by clinical pathologists for downstream molecular testing. Artificial intelligence (AI)-based digital quantification (DQ) of TCC represents a promising alternative, yet real-world evidence from routine molecular diagnostic workflows remain limited. In this study, we evaluated the analytical performance and practical aspects of analytical validation process of an AI-based DQ tool in routine molecular diagnostics. The clinical-grade AIM-TumorCellularity (AIM-TC; PathAI©) workflow was tested in molecular diagnostics for samples analyzed by comprehensive genomic profiling (FoundationOne®CDx (F1CDx), Foundation medicine Inc.). The cohort included 300 non-paired resection, biopsy, and cytology/cell block) specimens from primary and metastatic breast (n = 66), lung (n = 117), colorectal (n = 40), pancreatic (n = 38), and prostate (n = 39) cancers, reflecting real-world diagnostic sample heterogeneity of a tertiary care center. We compared TCC estimates generated by pathologists' VQ, AI-based DQ, and molecular quantification (MQ) by bioinformatic deconvolution. Agreement was lowest between VQ and MQ (Spearman Rs = 0.38) and between VQ and DQ (Rs = 0.44), while DQ showed stronger concordance with MQ (Rs = 0.63). Single-cell validation against expert ground truth demonstrated high performance of DQ in tumor cell detection, with sensitivity of 0.98.5, specificity of 0.99, and accuracy of 0.99, based on 27,958 annotated cells across 60 regions of interest comparable to microscopic high-power fields. Analysis of pre-analytical and analytical factors identified specimen type and cautery/crush artifacts as the main pre-analytical contributors to DQ-VQ discrepancies, while overall variations in specimen cellularity was the dominant analytical factor. In summary, this study provides the first comprehensive real-world evaluation of AI-based TCC quantification in routine molecular pathology workflow, highlighting its robustness, accuracy, and the critical role of pre-analytical standardization, as well as pathologists` oversight for successful clinical implementation.

Multiple endocrine neoplasia (MEN) syndromes are rare hereditary disorders characterized by the development of multiple endocrine and non-endocrine tumours with variable penetrance and age-dependent expression. Although uncommon, these syndromes are highly relevant from both biological and clinical perspectives, as they exemplify the direct link between germline genetic alterations and tumorigenesis. Early tumour detection is critical in MEN syndromes because many associated neoplasms-such as medullary thyroid carcinoma (MTC), pancreatic neuroendocrine tumours (NETs), pheochromocytomas, and parathyroid disease-may remain clinically silent for prolonged periods while retaining malignant potential. Delayed diagnosis is associated with advanced disease and worse outcomes, whereas early identification enables curative or organ-preserving interventions. This clinical challenge has driven the development of integrated diagnostic strategies combining genetic testing, biochemical markers, and imaging. Among these, genetic testing plays a pivotal role, providing definitive diagnosis, enabling family screening, and guiding risk-adapted surveillance. The aim of this review is to provide a comprehensive synthesis of genetically driven diagnostics in MEN syndromes, outlining the current state of the art and future directions in precision medicine.

Training and body composition requirements in elite sports may elevate eating disorder (ED) risk. Current estimates suggest a range of 1-28% prevalence of ED in elite athletes, reflecting methodological heterogeneity, underreporting, and limited research focus. EDs adversely affect physical health (e.g., osteoporosis, fatigue, injury), psychological well-being, and athletic performance, often persisting beyond athletic careers. Sport-specific demands may obscure ED symptoms, making underdiagnoses likely. This longitudinal study's objective is to furnish proof for the theoretical model of disordered eating in elite athletes by examining the role of sport-specific indicators, to improve early identification and psychological diagnostics. Three-hundred elite athletes from weight-sensitive (ballet, bodybuilding) and less weight-sensitive (soccer, racket sports, basketball) sports will complete assessments at four time points. A subgroup of 90 athletes will participate in clinical interviews. Logistic and hierarchical regressions will identify ED risk indicators and estimate the prevalence of EDs in athletes. Receiver Operating Characteristic (ROC) analyses will assess diagnostic accuracy of single and combined indicators with reporting of sensitivity, specificity, and the Area Under the Curve (AUC). Group differences and measurement invariance of ED instruments will be tested between the athlete group and 300 non-athlete controls. Identifying reliable ED indicators in elite sports may support early intervention. This theory-based approach aims to enhance diagnostic accuracy and athletes' care. Prospective registration of the study in the German Clinical Trials Register (DRKS00035100) on 03 February 2025.

Background: Artificial intelligence (AI) is increasingly being integrated into modern dental practice, particularly in diagnostics, radiographic analysis, treatment planning, and practice management. Despite the rapid advancement of AI-based technologies, evidence regarding dentists' attitudes toward AI in Central and Eastern Europe remains limited. This study aimed to evaluate Polish dentists' attitudes toward artificial intelligence in contemporary dental practice and to investigate differences in AI acceptance according to sex, age group, and dental specialty. Materials and Methods: A cross-sectional online questionnaire-based study was conducted among licensed dentists practicing in Poland. An anonymous questionnaire comprising 15 attitude statements rated on a five-point Likert scale was distributed through professional social media groups. The survey assessed attitudes toward the use of AI in clinical, diagnostic, and administrative aspects of dentistry. Statistical analyses were performed using Statistica 16.0. Group comparisons were conducted using the Mann-Whitney U test and Kruskal-Wallis test with Benjamini-Hochberg false discovery rate correction. Internal consistency of the questionnaire was assessed using Cronbach's alpha coefficient. Results: A total of 183 completed questionnaires were included in the analysis. The internal consistency of the questionnaire was high (Cronbach's &#x3b1; = 0.900). The overall acceptance of AI was moderate (mean score: 3.18 &#xb1; 0.73). The highest levels of agreement were observed for the perceived potential of AI to improve dental practice management (mean = 4.05) and for general openness toward AI implementation in dentistry (mean = 4.05). Respondents also expressed a high willingness to use AI for generating clinical documentation (mean = 3.69). In contrast, the lowest acceptance was observed for statements suggesting that AI could replace dentists (mean: 1.36). Within the study sample, men demonstrated significantly higher overall AI acceptance than women (p < 0.001). Significant differences were also observed between age groups and dental specialties. Orthodontists demonstrated the highest AI acceptance among the surveyed specialties; however, these findings should be interpreted as exploratory because of unequal subgroup sizes. After FDR correction, significant differences between age groups remained only for selected questionnaire items. Conclusions: Within the limitations of this convenience sample, the findings suggest that Polish dentists generally perceive artificial intelligence as a supportive tool rather than a replacement for clinicians. Acceptance was greatest for administrative and organizational applications of AI, whereas autonomous clinical decision-making received substantially lower support. These findings suggest that the successful implementation of AI in dentistry should prioritize assistive technologies that enhance clinical workflows while preserving the central role of the dentist in patient care.

Background/Objectives: To develop a fast and interpretable multimodal framework for the automatic detection of cardiac abnormalities using electrocardiogram (ECG) and phonocardiogram (PCG) signals. Methods: A multimodal classification scheme was designed by combining ECG and PCG recordings. For each modality, tailored preprocessing and temporal and spectral feature extraction were applied. The resulting information was fused through a quality-aware strategy that prioritized more reliable signal segments. The explainability results showed that the model focused on physiologically meaningful regions, providing supportive interpretability for its predictions. Experiments were conducted on the PhysioNet/CinC 2016 heart sound dataset, including normal and pathological recordings, using 10-fold cross-validation. Results: The proposed method achieved a mean F1 score of 97.2%, an accuracy of 95.6%, a specificity of 88.6%, and a sensitivity of 97.7%. In addition, the lightweight preprocessing and fast feature extraction pipeline allowed the full 10-fold cross-validation procedure to be completed in only 66 s. Conclusions: The proposed ECG-PCG framework provides a fast, accurate, and interpretable solution for automated cardiac abnormality detection and appears well suited for real-time cardiac screening applications.

Mitral stenosis (MS) remains a clinically relevant condition worldwide, with rheumatic and degenerative aetiologies contributing to a broad spectrum of disease. Accurate assessment of MS severity is essential for clinical decision-making but is often challenged by technical limitations, complex hemodynamic interactions, and the heterogeneous anatomical characteristics of different MS aetiologies. This review aims to provide a comprehensive overview of the contemporary multimodality imaging assessment of MS, with particular emphasis on the pitfalls of conventional transthoracic echocardiography (TTE), the incremental value of advanced imaging modalities, and their integration into diagnostic and therapeutic decision-making. A targeted narrative review of the literature was conducted focusing on multimodality imaging approaches and their integration into clinical practice. Particular emphasis was placed on TTE parameters, three-dimensional (3D) TTE, stress echocardiography, transoesophageal echocardiography (TOE), cardiac computed tomography (cCT), and cardiac magnetic resonance (CMR), highlighting their complementary roles in anatomical characterization, hemodynamic assessment, procedural planning, and clinical decision-making. Conventional 2D TTE remains the cornerstone for MS evaluation; however, widely used parameters such as mean transmitral gradient (TMG) and pressure half-time (PHT) are highly load-dependent and may lead to misclassification of disease severity in the presence of altered hemodynamic conditions (e.g., tachycardia, atrial fibrillation (AF), or reduced cardiac output). Although direct planimetry remains the anatomical reference standard, its accuracy may be limited by operator dependency, calcification, and complex valve geometry. In this context, three-dimensional TTE (3D TTE) provides incremental value by enabling more accurate visualization of the mitral valve (MV) orifice and improving measurement reproducibility. Stress TTE plays a key role in patients with discordant symptoms or borderline resting findings by unmasking clinically significant disease during exercise. In selected patients, TOE, cCT, and CMR provide complementary information for the evaluation of complex valve morphology, mitral annular calcification (MAC), ventricular remodelling, and procedural planning, particularly in degenerative mitral stenosis (DMS) and candidates for transcatheter interventions. The evaluation of MS requires an integrated, multiparametric, multimodality imaging approach. Combining conventional TTE with stress imaging and complementary advanced modalities improves diagnostic accuracy, facilitates patient selection for intervention, and supports individualized clinical decision-making, particularly in patients with complex anatomy or discordant imaging findings.

Background: Psychiatric inpatients may be vulnerable to head trauma; however, associations among psychiatric diagnosis, trauma mechanisms, and acute trauma-related neuroimaging findings among patients requiring neurosurgical consultation remain unclear. Methods: This retrospective observational study included psychiatric inpatients referred for neurosurgical consultation after head trauma at a tertiary neuropsychiatric specialty hospital between July 2022 and December 2025. Psychiatric diagnoses were grouped into psychotic, mood, substance use, organic/symptomatic, and other disorders. Trauma mechanisms were classified as falls, self-harm, assault/fight-related injuries, and other trauma types. Radiological findings were classified as trauma-related abnormalities or secondary incidental findings, including age-related changes. Associations were evaluated using chi-square tests, with false discovery rate correction applied to global and binary comparisons. Exploratory multivariable regression models were used as sensitivity analyses. Results: The cohort comprised 461 psychiatric inpatients who sustained head trauma, underwent radiological imaging, and were referred for neurosurgical consultation. Trauma mechanisms differed significantly across psychiatric diagnostic groups (&#x3c7;2 = 30.89, p = 0.002). Falls were the most common mechanism (189/461, 41.0%). Assault/fight-related trauma showed the clearest diagnosis-associated pattern (&#x3c7;2 = 19.72, p < 0.001, Cram&#xe9;r's V = 0.207) and remained the only significant binary outcome after false discovery rate correction. Acute trauma-related imaging findings were observed in 35 patients (7.6%) and were not associated with psychiatric diagnosis (&#x3c7;2 = 6.16, p = 0.187). Calcifications were the most frequent imaging finding (191/461, 41.4%). Conclusions: In psychiatric inpatients, while head trauma patterns and neuroimaging findings have the potential to be explained by psychiatric diagnoses, they may also reflect unmeasured behavioral and clinical factors.

Background/Objectives: Brain tumor segmentation from magnetic resonance imaging (MRI) plays an important role in clinical assessment and treatment planning. However, accurate segmentation remains challenging because of the complex anatomical structure of the brain, variations in tumor size and shape, and the imbalance between tumor and non-tumor regions in MRI datasets. These challenges highlight the need for reliable automated segmentation methods. Methods: This study proposes a multilevel deep learning model for automated brain tumor segmentation using MRI images. The BraTS dataset was used for model development and evaluation. To address class imbalance, a modified Synthetic Minority Oversampling Technique (SMOTE) was incorporated during preprocessing. A Multilevel Architecture-Based Modified U-Net was then employed to learn multiscale spatial features and generate pixel-wise tumor segmentation. The proposed framework was evaluated using the Dice coefficient, Jaccard coefficient, Matthews Correlation Coefficient (MCC), and accuracy. Results: The experimental results demonstrate that the proposed model consistently outperformed the Berkeley Wavelet Transform (BWT)-based method and the conventional U-Net across different tumor grades. Higher Dice, Jaccard, and MCC values indicate improved agreement between the predicted segmentation and the expert-annotated ground truth masks, demonstrating more accurate and consistent tumor delineation. Conclusions: The proposed multilevel deep learning model provides an effective framework for automated brain tumor segmentation from MRI images. By combining imbalance-aware preprocessing with a lightweight Modified U-Net architecture, the proposed method improves segmentation performance while maintaining a relatively simple network design. Future work will focus on validating the proposed framework using external clinical datasets and comparing it with recent state-of-the-art segmentation models.

Large-bore mechanical thrombectomy (LBMT) is a catheter-directed therapy for acute pulmonary embolism (PE). The relationship between aspirated thrombus weight and volume with outcomes remains unclear. The aim was to evaluate the impact of aspirated thrombus weight and volume on outcomes after LBMT. This prospective, open-label, single-arm, single-center registry study included 48 patients undergoing LBMT using the FlowTriever system (Inari Medical/Stryker, Irvine, CA, USA). Thrombus weight and volume were quantified, and clot composition assessed. Associations between thrombus characteristics and clinical, invasive hemodynamics, echocardiographic parameters, and biomarkers were evaluated immediately after the procedure, at hospital discharge, and at 3-month follow-up. Thrombus material was available in 48 patients (31% women), of which 41 presented with intermediate-risk and 7 with high-risk PE. LBMT resulted in significant reductions in systolic pulmonary artery pressures (sPAP) intraprocedurally (-&#x2009;12.8&#x2009;&#xb1;&#x2009;8.3 mmHg, p&#x2009;<&#x2009;0.001), with a further invasively measured decrease through 3 months (-&#x2009;9.6&#x2009;&#xb1;&#x2009;10.6 mmHg, p&#x2009;<&#x2009;0.001). From baseline to discharge, right ventricular coupling improved (+&#x2009;0.24, p&#x2009;<&#x2009;0.001) and right ventricle (RV)/left ventricle (LV) ratio decreased (-&#x2009;0.23, p&#x2009;<&#x2009;0.001). NT-proBNP and high-sensitivity cardiac troponin T declined significantly. Neither thrombus weight nor volume correlated with acute changes in sPAP (&#x3c1;volume&#x2009;=&#x2009;0.06; &#x3c1;weight&#x2009;=&#x2009;0.10), RV-uncoupling (&#x3c1;volume&#x2009;=&#x2009;0.47; &#x3c1;weight&#x2009;=&#x2009;0.46), and RV/LV ratio (&#x3c1;volume&#x2009;= -&#x2009;0.02; &#x3c1;weight&#x2009;= -&#x2009;0.005) (p for all&#x2009;>&#x2009;0.05), nor with outcomes at 3 months. Aspirated thrombus weight and volume were not associated with improvements after LBMT, suggesting that, beyond mechanical obstruction, additional mechanisms potentially including paracrine and endocrine effects of thrombus material may contribute to acute and chronic PE-related cardiopulmonary dysfunction.

A 56-year-old patient was initially diagnosed with aseptic loosening of a left acetabular prosthesis based on 68Ga-FAPI. Half a year after revision, this patient developed right hip pain and a subsequent 68Ga-FAPI was performed revealing increased tracer uptake in the right hip, leading to a diagnosis of synovitis-induced osteoarthritis. This case highlights the utility of 68Ga-FAPI-PET/CT in visualizing synovial activity and underscores the potential role of synovitis in the pathogenesis of osteoarthritis.

Background/Objectives: Modifiable intrinsic factors such as foot posture, hallux mobility, dynamic balance, and body mass index (BMI) are widely assessed in athletic screening, but their reliability and interrelationships remain unclear. The aim of the study was to evaluate the reliability of foot posture, hallux mobility, dynamic balance, Q-angle, and navicular drop measures in collegiate athletes and to examine the relationships among these measures and BMI. The concurrent, discriminant, and predictive validity of the Foot Posture Index (FPI6) relative to the navicular drop test (NDT) was also assessed. Methods: Fifty-nine athletes participated: 10 in a reliability subsample and 49 in a separate validity subsample. Assessments included BMI, FPI-6, hallux valgus angle, hallux dorsiflexion, Y Balance Test (YBT), Q-angle, and NDT. Reliability was assessed using intraclass correlation coefficients (ICCs). Associations among measures and FPI6-NDT concurrent validity were examined using Spearman correlations and non-parametric group comparisons. Classification performance indices relative to NDT were calculated for FPI-6. Results: All measures demonstrated good to excellent reliability (ICC = 0.81-0.99). FPI-6 showed a strong concurrent association with NDT (rho = 0.735). More pronated foot posture was moderately associated with lower YBT performance and reduced hallux dorsiflexion (rho = -0.301 to -0.635). Higher BMI was associated with greater navicular drop and lower dynamic balance (rho = 0.301 to -0.433). FPI-6 showed high classification agreement with NDT-defined foot posture groups (specificity 96.9%; overall agreement 91.8%). Conclusions: Lower extremity alignment, hallux mobility, dynamic balance, and BMI showed modest interrelationships. FPI-6 demonstrated strong concurrent validity and high classification agreement relative to NDT, supporting practical screening use. Given the correlational nature of the findings, interpretations should remain cautious, and larger multimodal studies are warranted.

Obesity is a chronic, relapsing disease and a significant oncological risk factor. The correlation is most pronounced and consistent for endometrial cancer. Conversely, evidence linking obesity to ovarian cancer is less robust and varies by histotype, while the association with cervical cancer is influenced by factors related to screening, diagnosis, treatment, and survival. This review examines obesity, particularly class III (morbid) obesity, in relation to the risk of gynaecological cancer, diagnostic approaches, and management strategies. A structured narrative review of PubMed/MEDLINE, Cochrane Library, Scopus and Web of Science Core Collection was conducted for literature published between January 2000 and December 2025. Eligible evidence included systematic reviews, meta-analyses, cohort and case-control studies, mechanistic studies and clinical guidance relevant to obesity and endometrial, ovarian or cervical cancer. Title/abstract screening and full-text selection were conducted using predefined criteria for conceptual relevance and clinical applicability. Excess adiposity contributes to endometrial carcinogenesis through hormonal dysregulation, insulin resistance and hyperinsulinaemia, adipokine imbalance, chronic inflammation, and oxidative stress. In ovarian cancer, associations are generally weaker but appear more relevant for selected histological subtypes and cumulative adiposity exposure. In cervical cancer, obesity should not be interpreted as replacing HPV-driven pathogenesis; rather, it may affect screening adequacy, treatment selection, perioperative risk, and disease-specific survival in morbidly obese patients. Current evidence does not support morbid obesity as an independent driver of all gynaecological cancers. It supports obesity as a major modifiable risk factor and clinical modifier, particularly for endometrial cancer, and highlights the need for pragmatic risk stratification based on BMI class, adiposity distribution, metabolic comorbidity, functional status and cancer-site-specific pathways. Biomarker evidence remains hypothesis-generating, and obesity-integrated oncological pathways require prospective validation in patients with a BMI &#x2265; 40 kg/m2.

Background/Objectives: This study aimed to evaluate the usability and comprehension of at-home respiratory diagnostic tests for COVID-19 and influenza among rural and urban populations in Georgia. It sought to identify disparities in test interpretation and outcomes that could inform public health interventions tailored to underserved communities. Methods: Participants with respiratory symptoms (N = 592) were enrolled through community events in three rural Georgia counties (N = 188) and urban clinics in Atlanta (N = 404). Structured usability assessments and RT-PCR-confirmed diagnostic testing were conducted. Demographic, clinical, and survey data were analyzed using t-tests, z-tests, and Fisher's exact tests to compare rural and urban participants. Results: Rural participants were older and more likely to be White. Urban participants had significantly higher infection positivity rates for COVID-19 (18.5% vs. 10.2%) and Flu A (29.2% vs. 1.8%). While overall usability ratings were high across both groups (>94%), rural participants had significantly lower odds of identifying invalid test results and were more likely to continue using tests after procedural errors. Conclusions: Despite strong usability ratings, diagnostic test comprehension differed by geography. Rural participants demonstrated notable gaps in recognizing invalid results, underscoring the need for targeted educational strategies and simplified instructions to ensure equitable test interpretation and public health response.

Background/Objectives: Early identification of adult major trauma patients at high risk of death remains challenging in the emergency department. Lactate reflects tissue hypoperfusion, whereas ionized calcium contributes to coagulation and cardiovascular function. This study evaluated the prognostic performance of the lactate-to-ionized calcium ratio (LiCa) for in-hospital mortality and compared it with established trauma scores. Methods: This single-center retrospective cohort study included adults with major trauma, defined as an Injury Severity Score (ISS) &#x2265; 16, who presented to a tertiary emergency department between May 2021 and April 2023. Demographic, clinical, laboratory, and outcome data were obtained from electronic records. LiCa was calculated as lactate divided by ionized calcium using the initial arterial blood gas sample obtained within 30 min of emergency department arrival. Factors associated with mortality were examined using multivariable logistic regression. Discrimination was assessed using receiver operating characteristic analysis and compared across LiCa, ISS, and Trauma and Injury Severity Score (TRISS) models. Results: A total of 236 patients were analyzed. In-hospital mortality occurred in 49 patients (20.8%). LiCa remained associated with in-hospital mortality after adjustment (adjusted odds ratio, 1.42; 95% confidence interval, 1.23-1.62; p < 0.001). LiCa showed good discrimination in this cohort, with an area under the curve of 0.93 (95% confidence interval, 0.88-0.97), compared with 0.94 for both TRISS 1995 and TRISS 2010 and 0.85 for ISS. At an exploratory threshold of &#x2265;12.59, LiCa yielded 82% sensitivity, 94% specificity, a positive likelihood ratio of 12.72, and a negative likelihood ratio of 0.20. Conclusions: LiCa was independently associated with in-hospital mortality and showed good discrimination within this single-center retrospective cohort. As a rapidly available blood gas-derived measure, LiCa may provide exploratory adjunctive prognostic information at presentation, but it should not be interpreted as a replacement for validated trauma scoring systems. Prospective multicenter validation is needed before routine clinical use.

Background: Parkinson's disease (PD) diagnosis is often delayed until signature motor symptoms manifest, at which point profound dopaminergic neuron loss has already occurred, necessitating advanced motor diagnostic biomarkers. Quantitative gait analysis is a promising tool, but phase-specific kinematic parameters remain underexplored. This study aims to identify novel, stage-divided Timed Up and Go (TUG) biomarkers not only to differentiate healthy controls (HCs) from patients with PD but also to objectively monitor and track disease progression across advancing severity stages, which are further validated against conventional clinical motor scales. Methods: A total of 81 participants (48 PD, 33 HCs) performed a 3 m TUG test using MotionCore (JEIOS Inc., Busan, Republic of Korea). The test was subdivided into three movement phases (Stage 1, sit-to-walk; Stage 2, turning; Stage 3, walk-to-sit). Results: PD patients exhibited significantly prolonged durations and altered turning metrics compared to HCs. Turning parameters including turning radius (ETR), area (EMA), and turning stability (FN) demonstrated strong correlations with disease severity and clinical scales. Notably, stage-specific analyses revealed that step counts, time, and speed metrics across Stages 1, 2, and 3 effectively differentiated disease severity, with transitional decelerating and seating metrics in Stage 3 showing the most pronounced clinical correlations. Discussion: This study confirms that the TUG test systematically deteriorates with increasing PD severity. The robust correlations with clinical scales (UPDRS, FOG-Q, BBS) validate TUG metrics as objective measures of motor and balance impairments. Utilizing novel, staging-specific indices significantly enhances the TUG test's clinical utility for supporting diagnosis, accurate staging, and monitoring disease progression. Although the overall group comparisons demonstrated statistical significance, a data overlap remains between mild PD and HCs, underscoring the need for large-scale longitudinal studies to validate these metrics for early detection.

Peripheral facial palsy is often attributed to idiopathic Bell's palsy, but secondary structural causes should be considered when local red flags are present. A 61-year-old man with a 40-pack-year smoking history and former betel quid chewing presented with a verrucous-appearing mass involving the left oral commissure and buccal mucosa, intermittent purulent discharge from the lesion, progressive left facial swelling, and ipsilateral lower-motor-neuron facial palsy with lagophthalmos. Magnetic resonance imaging demonstrated a left buccal/oral-cavity lesion with ipsilateral parotid duct obstruction. He underwent tracheostomy, wide excision, left supraomohyoid neck dissection, and radial forearm free-flap reconstruction. Pathology confirmed squamous cell carcinoma, pT2N0, cM0 (stage II), with perineural invasion; surgical margins were negative, and lymphovascular invasion was not identified. At follow-up, wound healing was satisfactory and purulent discharge had resolved, but lower-motor-neuron facial palsy persisted; adjuvant radiotherapy was recommended. This case emphasizes that lower-motor-neuron facial palsy with an oral mass, purulent discharge, facial swelling, or salivary-duct obstruction should prompt careful oral examination and head-and-neck imaging.

Artificial intelligence (AI) is transforming gastrointestinal (GI) endoscopy into a more standardized, data-driven, and workflow-integrated field. Advances in computer-assisted detection (CADe), diagnosis (CADx), quality assessment (CAQ), natural language processing (NLP), and multimodal deep learning have expanded AI applications across colonoscopy, upper endoscopy, endoscopic ultrasound (EUS), ERCP, cholangioscopy, and capsule endoscopy. These systems have demonstrated improvements in lesion detection, procedural quality assessment, workflow efficiency, and diagnostic support. However, current evidence remains largely focused on surrogate outcomes rather than patient-centered clinical benefits, while challenges related to generalizability, explainability, regulatory oversight, automation bias, and workflow integration continue to limit widespread adoption. Future progress will depend on prospective real-world validation, diverse datasets, explainable AI frameworks, and careful integration of human-AI interaction into clinical practice. Overall, AI is evolving from a supportive adjunct into an increasingly integrated component of gastrointestinal endoscopy with the potential to improve procedural quality, diagnostic consistency, and clinical efficiency.

Background: Pathogenic variants in genes that cause skeletal dysplasias may, instead of producing classic findings, present in children with a phenotype whose hip radiographs resemble bilateral Legg-Calv&#xe9;-Perthes disease (LCPD). Objectives: This study aims to investigate the efficacy of genetic diagnosis in children with waddling gait or joint pain and radiological evidence of hip dysplasia mimicking bilateral LCPD. Methods: Forty children with bilateral femoral head dysplasia from 36 families were included in the study. Exome sequencing was performed, and all identified variants were confirmed within the families by Sanger sequencing. Results: Twelve pathogenic or likely pathogenic variants were identified: six in COL2A1, two in COL9A1, and one each in RPL13, EIF2AK3, DNAJC21, and ARSK; six are novel. The diagnostic yield was 33.3% (12/36) in 12 families. Additionally, variants of uncertain significance (VUS), proposed as causative, were detected in five families (5/36:13.9%): two in COL11A1 and one each in COL9A3, COL11A2, and ARSK. Based on bilateral epiphyseal dysplasia of the femoral head, it was observed that seven families may be compatible with mild spondyloepiphyseal dysplasia and six families may have Stickler syndrome. Notably, among these, three children carrying closely localized pathogenic/likely pathogenic variants in COL2A1 shared a novel phenotype characterized by short stature and bilateral irregular femoral heads. In four families, EIF2AK3, DNAJC21, and ARSK were also responsible for the ultra-rare disorders Wolcott-Rallison syndrome, bone marrow failure syndrome 3, and mucopolysaccharidosis 10, respectively. Conclusions: This study, for the first time, investigated the frequency of associated genes in a pediatric cohort with bilateral hip dysplasia resembling LCPD, providing important information for pathogenesis and differential diagnosis.