Search Results (643)

To assess the environmental impact of uranium mining operations in southern Kazakhstan, the elemental and radionuclide composition of soil samples collected from settlements in the Kyzylorda Region was investigated. The analysis was carried out using X-ray fluorescence (XRF) and gamma-ray spectrometry. Mean concentrations and variation ranges were determined for 28 chemical elements, including uranium, lead, antimony, and gamma-emitting radionuclides such as ¹³⁷Cs, ⁴⁰K, ²³²Th, ²³⁸U, ²²⁶Ra, ²¹⁰Pb, and ²⁴¹Am. The analysis of the specific activities of the artificial radionuclides¹³⁷Cs and²⁴¹Am was carried out in order to assess the influence of the Semipalatinsk Test Site on the soils of the study area. Based on the obtained data, heavy metal pollution indices, ecological risk indices, and radiological parameters were calculated to evaluate the potential environmental and human health impacts. Most elements were present at levels below average crustal abundances, suggesting limited anthropogenic influence. Slight exceedances for uranium, lead, and antimony are likely associated with regional geochemical features. Radiological assessment indicated that the radiation environment remains within internationally accepted limits. The lifetime cancer risk values for exposure of humans to natural radionuclides²²⁶Ra, ²³²Th, ⁴⁰K and ¹³⁷Cs from soil at 1 m above ground level ranged from0.19 × 10⁻³ to0.30 × 10⁻³, with an average of0.24 × 10⁻³. Nearly all sampling points remained below the risk threshold of 0.29 × 10⁻³, indicating minimal radiological hazard. The carcinogenic risk remained within the acceptable regulatory range for both adults (2.0 × 10⁻⁵) and children (4.4 × 10⁻⁵), whereas the non-carcinogenic hazard index for children (1.3) slightly exceeded the screening threshold of 1. This finding identifies children as the most sensitive receptor group under the conservative assumptions of the applied screening methodology. The study demonstrates the applicability of combined chemical and radiometric methods for comprehensive environmental assessments in uranium mining regions. The results are of interest both in terms of methodology and in understanding the local geochemical and radiological landscape. Full article

Background and Objectives: Pediatric appendicitis remains a common surgical emergency, but early risk stratification of complicated appendicitis and expected hospital length of stay (LOS) remains challenging. This study developed and internally evaluated an explainable machine learning framework for pediatric appendicitis using routinely available clinical, laboratory, and radiological variables. Materials and Methods: This retrospective single-cohort model-development study included 152 pediatric patients with acute appendicitis treated at King Abdulaziz University Hospital, Jeddah, Saudi Arabia, between January 2019 and December 2024. Two prediction tasks were evaluated: complicated appendicitis classification and expected LOS regression. Prediction was performed after initial clinical assessment, first laboratory testing, and initial diagnostic imaging, but before surgery or definitive conservative treatment. Operative findings, histopathological findings, postoperative variables, treatment-response variables, actual LOS-related variables, and final disease-severity labels were excluded as predictors. Model performance was evaluated using repeated nested cross-validation, bootstrap 95% confidence intervals, calibration analysis, benchmark comparisons, LOS sensitivity analyses, and explainability analysis using feature importance and SHAP values. Results: Complicated appendicitis was present in 57 patients (37.5%), and median LOS was 3 days (IQR: 2–6). For complicated appendicitis prediction, the repeated nested cross-validation framework achieved an AUC of 0.788 (95% CI: 0.710–0.856), accuracy of 0.750 (95% CI: 0.684–0.816), sensitivity of 0.684 (95% CI: 0.559–0.796), specificity of 0.789 (95% CI: 0.705–0.868), and Brier score of 0.183 (95% CI: 0.158–0.210). The framework showed comparable performance to a parsimonious logistic regression model (AUC: 0.807). For expected LOS prediction, the regression framework achieved an MAE of 2.150 days (95% CI: 1.632–2.846), MdAE of 1.305 days (95% CI: 1.004–1.493), RMSE of 4.329 days (95% CI: 2.290–6.297), and R² of 0.292 (95% CI: 0.213–0.481). LOS prediction showed lower MAE, MdAE, and RMSE than median and mean LOS null baseline models. Explainability analysis identified symptom duration, lymphocyte percentage, vomiting, temperature, and age as important predictors of complicated appendicitis, while radiological perforation, symptom duration, radiological collection, CRP, WBC count, NLR, lymphocyte percentage, and appendicolith by radiology contributed to expected LOS prediction. Conclusions: Explainable machine learning methods showed potential for internally validated prediction of complicated appendicitis and expected LOS using post-assessment pre-treatment data. However, this was a retrospective, single-cohort development study with a modest sample size and no external validation cohort. The findings should therefore be interpreted as exploratory and should not be considered evidence of clinical generalizability or readiness for implementation. Larger prospective multicenter studies with external validation, prediction-interval estimation, and clinical utility assessment are required before clinical use. Full article

Background and Objectives: Posterior circulation ischemic stroke (PCIS) is a life-threatening disease with a poor prognosis. It has nonspecific symptoms. It is very difficult to diagnose. The aim of the study is to examine to what degree radiological methods, in addition to standard clinical indicators, can serve as a reliable and independent predictor of death in patients with PCIS. Materials and Methods: It is a retrospective study on 175 patients at the Special Hospital “Saint Sava” (2023–2025) with symptoms of acute PCIS, who did not have proven PCIS on the initial non-contrast computed tomography (NCCT). The radiology protocol included NCCT, CT perfusion (CTP) and CT angiography (CTA). The experimental group (n = 116, CTP+) and the control group (n = 59, CTP-) were monitored clinically and radiologically, with a follow-up NCCT after 24 h when all had confirmed acute PCIS. At discharge, the outcome was monitored on the National Institutes of Health Stroke Scale (NIHSS 2) and the Modified Rankin Scale (mRs). Results: Of 175 patients with acute PCIS, 24% died within 30 days. Those who died were older (p < 0.05), had a lower Posterior Circulation Alberta Stroke Program Early CT Score (pcASPECT), and more often had hyperdensity (26.2% vs. 7.5%), occlusions (69% vs. 31.6%), and symptomatic blood vessels on CTA and CTP+ (81% vs. 61.7%, p < 0.01). Multivariate regression singled out CTP+ (OR = 4.08), complications (OR = 2.96), Israeli Vertebrobasilar Stroke Scale (IVBSS) (OR = 1.22) and Trial of Org 10172 in Acute Stroke Treatment (TOAST) (OR = 1.92) as predictors of mortality, with shorter survival in CTP+ (949 vs. 1449 days, Log-rank p = 0.015), more NIHSS/mRS, lower Glasgow Coma Score, and more frequent obesity in the deceased. Conclusions: Negative CTP represents the protective factor against death in patients with acute PCIS. Complications, higher IVBSS, and TOAST classification (large vessels atherosclerosis) represent an increased risk. The importance of CTP for early detection of acute PCIS and early risk assessment is highlighted. Full article

Background: Recurrent lumbar disc herniation (rLDH) significantly impairs outcomes following percutaneous endoscopic lumbar discectomy (PELD). Accurate individualized risk prediction remains challenging. This study aimed to develop and compare multiple machine learning models for predicting rLDH within two years post-surgery. Methods: A retrospective cohort of 1483 patients undergoing single-level PELD was analyzed. The primary outcome was symptomatic, magnetic resonance imaging-confirmed rLDH requiring reintervention. Candidate predictors included demographic, surgical, and radiographic parameters. The dataset was stratified by outcome and randomly split into training (70%, n = 1038) and validation (30%, n = 445) sets. Feature selection utilized univariate screening (p < 0.2) and least absolute shrinkage and selection operator regression. Six machine learning algorithms were trained and optimized via grid search. Performance was evaluated using the area under the receiver operating characteristic curve (AUC), F1-score, Brier score, calibration and decision curve analysis (DCA). Model interpretability was assessed using SHapley Additive exPlanations (SHAP). Results: The overall recurrence rate was 4.25% (63/1483). Logistic regression achieved the optimal F1-score (0.286), while light Gradient Boosting Machine (LightGBM) demonstrated superior discrimination (AUC = 0.768). DCA indicated clinical utility primarily at low threshold probabilities (<10%). SHAP analysis identified increased sagittal range of motion as the strongest risk factor, followed by reduced facet orientation, advanced age, type II Modic changes, and Michigan State University zone C. Conclusions: This study presents an exploratory, internally validated machine learning framework for rLDH risk stratification. While LightGBM demonstrated moderate discriminative ability, model sensitivity was constrained by the inherent rarity of recurrence events, precluding its use as a definitive standalone screening tool. Notably, clinical utility was restricted to low threshold probabilities (<10%), supporting a focused role in identifying high-risk subgroups for intensified preoperative counseling and postoperative monitoring. Beyond elucidating key radiological and demographic risk factors, our findings underscore that rigorous external prospective validation and probability calibration are indispensable before any future clinical deployment. Full article

Background: Osteoporosis is a major public health concern, and early detection through dual-energy X-ray absorptiometry (DXA) has been pivotal. However, evidence on how insurance coverage relates to the distribution of patients receiving DXA examinations in Saudi Arabia remains limited. This study examines the distribution of patients receiving DXA examinations across different insurance types and healthcare settings among adult patients in Saudi Arabia. Methods: This retrospective observational study used de-identified electronic health record data from a public tertiary academic center and a large private healthcare system in Saudi Arabia between 1 January 2023 and 28 February 2026. Dual-energy X-ray absorptiometry (DXA) was employed as an imaging method used to measure bone mineral density and support osteoporosis diagnosis and fracture-risk assessment. Adult patients aged ≥18 years with completed DXA examinations recorded in radiology service records were included. This study assessed the distribution of patients receiving DXA examinations across insurance types, healthcare setting, ordering department, and year of service. Insurance type was categorized as government coverage, private insurance, or self-pay. Descriptive statistics, chi-square tests, one-way ANOVA, Cramér’s V, and standardized residuals were used to compare distribution patterns across groups. Results: The data of 8930 DXA recipients were analyzed. Clear differences were observed in the distribution of patients by insurance type and healthcare setting (p < 0.001). Privately insured patients were predominantly treated in private facilities, whereas government-insured and self-pay patients were primarily concentrated in public healthcare facilities. Significant variations were also observed across healthcare departments and over time, demonstrating strong system-level and financial stratification in the distribution of patients receiving DXA examinations. Conclusions: The findings of this study demonstrate significant differences in the distribution of patients receiving DXA examinations according to insurance type and healthcare setting. Notably, self-pay patients were predominantly managed in public healthcare facilities, whereas privately insured patients primarily received DXA examinations in private healthcare settings. These observed patterns suggest that differences in healthcare organization and insurance financing may be associated with where patients receive DXA examinations; however, because this study included only individuals who underwent DXA examinations, the findings should not be interpreted as measures of screening uptake, access, or equity among all patients eligible for osteoporosis screening. Future healthcare strategies should focus on strengthening equitable insurance coverage, improving coordination between public and private healthcare sectors, standardizing referral pathways, and supporting integrated preventive care in line with Saudi Arabia’s Vision 2030 healthcare transformation. Full article

Thoracic morphology is increasingly recognized as an important determinant of cardiopulmonary phenotype, influencing cardiovascular mechanics, respiratory physiology, and the interpretation of diagnostic imaging findings. Although the radiological Haller Index (HI) remains the reference standard for quantifying pectus excavatum severity, its dependence on computed tomography and ionizing radiation limits widespread clinical implementation, particularly in settings requiring serial evaluations. To overcome these limitations, the Modified Haller Index (MHI) was developed as a simple, non-invasive, radiation-free alternative that combines external thoracic anthropometry with echocardiographic assessment. Since its introduction, the MHI has undergone clinical validation and has progressively expanded beyond the assessment of chest wall deformities, demonstrating that thoracic conformation is not merely an anatomical characteristic but a clinically relevant determinant of cardiovascular and respiratory physiology. Growing evidence indicates that thoracic morphology influences cardiac chamber geometry, ventricular filling, stroke volume, myocardial deformation, ventricular–arterial coupling, exercise stress echocardiography findings, pulmonary function, and symptom perception across a broad spectrum of cardiovascular and respiratory diseases. Elevated MHI values identify individuals with a reduced antero-posterior thoracic diameter and a distinctive cardiopulmonary phenotype characterized by external cardiac compression, smaller cardiac chambers, restrictive ventilatory physiology, and apparent alterations in myocardial mechanics despite the absence of intrinsic myocardial disease. Building upon the clinical validation of the MHI, a novel patented anatomical measurement device was engineered to standardize thoracic morphometric assessment by enabling direct acquisition of both latero-lateral and antero-posterior thoracic diameters within a single measurement procedure. The device integrates dedicated anatomical reference elements, an innovative adjustable sternal pointer, movable measurement components, and a standardized acquisition workflow into a portable, low-cost, and radiation-free platform, thereby improving measurement reproducibility while simplifying bedside MHI determination. This narrative review summarizes the historical evolution of thoracic morphometry, the development and clinical validation of the MHI, the engineering rationale, structural architecture, and measurement workflow of the patented device, and the growing evidence supporting the clinical significance of thoracic conformation across cardiovascular and respiratory medicine. Together, the MHI and the proposed anatomical measurement device establish a practical platform for standardized, radiation-free thoracic morphometry that may facilitate routine bedside phenotyping. Future integration with digital technologies, artificial intelligence, and advanced imaging systems may further enable next-generation digital thoracic phenotyping for personalized cardiovascular and respiratory characterization, risk stratification, and precision medicine. Full article

Background/Objectives: Angiogenesis in glioblastoma (GBM) is a multifactorial process, and blood–brain barrier disruption enables the detection of circulating mediators. The clinical relevance of circulating placental growth factor (PlGF) in GBM remains unclear. This study aimed to investigate the role of PlGF in GBM and its association with disease characteristics and outcomes. Methods: We conducted a prospective observational study on 54 patients with IDH-wildtype GBM. Plasma samples collected at diagnosis and recurrence were analyzed using a multiplex panel of angiogenesis mediators. Associations with clinical, radiological, molecular, and treatment-related variables were assessed, along with survival outcomes. Statistical analysis was performed with R 4.5.0. Results: At baseline, PlGF correlated with multiple angiogenic mediators, including VEGF, IL-6, angiopoietin-1, EGF, FGF, IL-8, and TNF-α. Higher PlGF levels were associated with radiopathological features of tumor biology, including proliferation markers and the FLAIR/contrast enhancement ratio. In high-risk patients (RPA 3–4; n = 33), low baseline PlGF identified a subgroup with significantly longer overall survival (17.6 vs. 8.5 months; log-rank p = 0.031) and retained a protective association in multivariable models. In the overall cohort, this association was weaker and did not reach statistical significance. Exploratory longitudinal analyses suggested an increase in PlGF at recurrence in selected molecular and treatment-defined subgroups, while no association with bevacizumab exposure was observed. Conclusions: Circulating PlGF may reflect tumor biology in GBM and shows prognostic relevance in high-risk patients, where low baseline levels identify a subgroup with improved survival. These findings support PlGF as a candidate circulating biomarker and warrant validation in larger prospective cohorts. Full article

(1) Objectives: In acute pulmonary embolism (PE), detailed thrombus burden scores are often complex and time-consuming, limiting their integration into urgent radiology reports. We evaluated a simplified modified Ghanima score (GmScore and GmS) designed to provide a structured estimate of thrombus burden and assessed its clinical correlation and reproducibility. (2) Methods: In this retrospective single-center study, 132 consecutive patients with confirmed acute PE were classified according to the modified GmScore: GmS1 (segmental), GmS2 (lobar), and GmS3 (main pulmonary arteries), considering luminal obstruction ≥ 50%. European Society of Cardiology (ESC) risk category, simplified Pulmonary Embolism Severity Index (sPESI), CT right-to-left ventricular (RV/LV) ratio, echocardiographic right ventricular dysfunction, and 30-day mortality were recorded. Inter- and intraobserver agreement were assessed using weighted kappa. (3) Results: In 132 patients (mean age 64.8 ± 16.5 years; 77 men), a significant clinical gradient was observed across GmScore categories. ESC intermediate–high/high risk occurred in 0% of GmS1 and 95.6% of GmS2–3 patients (p < 0.001). The median RV/LV ratio increased progressively (0.76, 1.58, and 1.79 for GmS1–3; p < 0.001), with a strong correlation between the GmScore and RV/LV (Spearman ρ = 0.75). GmS2 and GmS3 showed no significant difference in ventricular repercussion (p = 0.938), whereas GmS1 differed markedly. Using GmS ≥ 2 to identify ESC intermediate–high/high risk yielded 100% sensitivity and negative predictive value. Interobserver agreement was excellent (κ = 0.92). Thirty-day mortality was 0% in GmS1, 2.0% in GmS2, and 14.6% in GmS3 (p = 0.005). (4) Conclusions: The modified GmScore is a simple, reproducible CT-based descriptor that aligns closely with right ventricular repercussion and ESC risk stratification. Full article

Background and Objectives: Synovial chondromatosis is a rare benign disease characterized by chondral metaplasia of the synovial membrane and the formation of intra-articular loose bodies. Although usually benign, delayed diagnosis may lead to progressive joint damage, recurrence, and, rarely, malignant transformation. Materials and Methods: A narrative review of the literature was conducted using the PubMed/MEDLINE database, focusing on studies published between January 2010 and December 2025. Eligible studies addressed the classification, epidemiology, clinical presentation, imaging findings, histopathology, differential diagnosis, treatment, recurrence, and prognosis of synovial chondromatosis. Results: Synovial chondromatosis most commonly affects large joints. Plain radiographs may show calcified loose bodies, whereas MRI is essential for detecting early non-calcified disease and assessing synovial involvement. Treatment depends on symptoms, disease extent, joint damage, and recurrence risk. Surgical removal of loose bodies with synovectomy remains the mainstay of treatment, while arthroplasty may be considered in advanced degenerative disease. Conclusions: This review provides an updated and clinically oriented synthesis of synovial chondromatosis, emphasizing early recognition, multidisciplinary diagnostic assessment, appropriate surgical management, and vigilant follow-up. Particular attention should be paid to recurrent disease and aggressive clinical or radiological features suggestive of malignant transformation. Full article

Objectives: The facial nerve (FN) possesses one of the most intricate anatomical courses in the head and neck, traversing the brainstem, temporal bone, and parotid gland before terminating within the muscles of facial expression. Owing to its complex anatomy, marked anatomical variability, and frequent distortion by adjacent pathology, preservation of FN integrity remains a fundamental challenge in skull base, otologic, and head and neck surgery. This review aims to provide a comprehensive synthesis of the contemporary literature regarding the clinical anatomy of the FN and to examine how anatomical variation, pathology-induced distortion, surgical strategy, and emerging technologies influence nerve preservation and functional outcomes. Methods: A comprehensive narrative review of the literature was conducted using PubMed, Scopus, and Google Scholar. Publications from 1983 through 2026 were searched using combinations of keywords, including “facial nerve,” “facial nerve anatomy,” “anatomical variation,” “vestibular schwannoma,” “hemifacial spasm,” “parotid surgery,” “facial nerve injury,” “facial nerve reconstruction,” “facial reanimation,” “diffusion tensor imaging,” “intraoperative neurophysiological monitoring,” and “artificial intelligence.” Peer-reviewed anatomical, radiological, clinical, and review articles published in English were included, while conference abstracts and studies lacking direct anatomical or surgical relevance were excluded. Particular emphasis was placed on surgically relevant anatomical variations, pathology-related anatomical distortion, advanced imaging modalities, intraoperative neurophysiological monitoring, reconstructive techniques, and predictors of postoperative facial nerve function. Results: Facial nerve preservation was found to depend on the interplay between individual anatomical variability, disease-related anatomical distortion, and operative strategy. In vestibular schwannoma surgery, nerve displacement, capsular adhesion, and cystic tumor degeneration were consistently associated with increased surgical complexity and less favorable postoperative facial function. In hemifacial spasm, successful microvascular decompression relied on precise identification of neurovascular conflict at the root exit zone. Within the parotid gland, substantial variability in branching architecture and surgical landmarks contributed to an increased risk of iatrogenic injury. Advanced imaging techniques, particularly diffusion tensor imaging tractography, improved preoperative prediction of FN location, while intraoperative neurophysiological monitoring enabled real-time assessment of neural integrity and functional preservation. Emerging artificial intelligence-based predictive models demonstrated potential to enhance patient-specific surgical planning and prognostication. Conclusions: Contemporary facial nerve surgery has evolved toward an individualized, anatomy-driven, and function-preserving paradigm supported by advanced imaging, intraoperative monitoring, and reconstructive strategies. Detailed understanding of both normal FN anatomy and pathology-induced anatomical distortion remains essential for optimizing surgical decision-making, maximizing nerve preservation, and improving long-term functional outcomes. Future developments integrating multimodal imaging, predictive analytics, and artificial intelligence may further refine patient-specific management and enhance postoperative facial function. Full article

Pulmonary fibrosis (PF), particularly idiopathic pulmonary fibrosis (IPF), is a progressive and often fatal interstitial lung disease characterised by complex genetic and molecular heterogeneity. Traditional diagnostic approaches, which rely on clinical, radiological and histopathological assessment, are frequently insufficient to capture the underlying biological diversity of the disease. The advent of next-generation sequencing (NGS) has substantially advanced the understanding of PF by enabling comprehensive genomic and transcriptomic profiling. NGS technologies, including whole-exome sequencing (WES), whole-genome sequencing (WGS), RNA sequencing (RNA-seq), and targeted gene panels, have uncovered key genetic determinants. These include mutations in telomere-related genes (TERT, TERC, RTEL1) and surfactant-related genes (SFTPC, SFTPA2), as well as common variants like the MUC5B promoter polymorphism. These discoveries have clarified disease pathogenesis, revealed polygenic risk models, and may improve diagnostic accuracy, particularly in distinguishing overlapping interstitial lung disease (ILD) phenotypes. Beyond genetics, transcriptomic analyses have identified dysregulated pathways, including TGF-β, Wnt/β-catenin, and PI3K/Akt signalling, and have enabled the discovery of novel biomarkers for prognosis and therapeutic response. In selected clinical settings, NGS is beginning to support patient stratification and inform management decisions. Emerging applications, including liquid biopsy and integration with artificial intelligence, further expand the potential clinical utility of NGS. Despite challenges related to cost, data interpretation and standardisation, NGS represents a powerful research tool in PF. Full article

Background: Bone and soft-tissue sarcomas are rare, heterogeneous malignancies whose surveillance remains dominated by imaging despite substantial molecular diversity and variable patterns of relapse. Circulating tumor DNA (ctDNA) and circulating tumor cells (CTCs) offer minimally invasive approaches for monitoring tumor biology, but their performance in sarcoma depends strongly on subtype, disease burden, assay design, and biological shedding. Methods: This narrative review synthesizes evidence published from 2015 to 2026 on ctDNA and CTCs for baseline risk assessment, treatment-response monitoring, minimal residual disease (MRD) detection, molecular relapse, and integration with imaging-based surveillance in bone and soft-tissue sarcomas. Results: Current evidence supports a genomics-guided framework in which liquid-biopsy strategy is selected according to sarcoma subtype, molecular architecture, and clinical purpose. ctDNA is the most mature analyte, with best-supported evidence in osteosarcoma, where tumor-informed assays predict postoperative relapse, and in translocation-associated sarcomas, where breakpoint-guided assays enable highly specific longitudinal monitoring. Copy-number-based approaches are relevant for complex-karyotype tumors, while mutation-, methylation-, fragmentomic-, and RNA-based strategies may be useful in selected contexts. However, detection rates vary, false-negative results occur in low-shedding or low-volume disease, and clinical utility for changing treatment remains incompletely established. CTCs provide complementary cellular and prognostic information, particularly in osteosarcoma, but remain limited by platform heterogeneity and incomplete standardization. Conclusion: Liquid biopsy may refine risk stratification, support treatment-response assessment, clarify indeterminate imaging findings, and identify molecular relapse in selected sarcoma patients. At present, it should be interpreted as an adjunct to imaging and specialist multidisciplinary care rather than as a replacement for standard radiologic surveillance. Full article

Background/Objectives: Machine learning (ML) models have increasingly been applied to prognostic prediction in oral cavity squamous cell carcinoma (OCSCC), though their performance and methodological quality remain variably reported. This systematic review evaluated contemporary ML-based prognostic models for clinically relevant OCSCC outcomes, with emphasis on independently validated studies. Methods: A systematic review was conducted according to PRISMA guidelines using PubMed, Scopus, Cochrane Library, and CINAHL databases through 1 December 2025. Studies evaluating ML or artificial intelligence prognostic models in adult OCSCC patients were included. Outcomes included overall survival, recurrence, disease-free survival, recurrence-free survival, disease-specific survival, cancer-specific survival, progression, and nodal metastasis. Data extraction and risk-of-bias assessment using PROBAST + AI were performed independently by reviewers. Results: Forty studies comprising 105,619 patients met inclusion criteria. ML architectures included random forests, support vector machines, gradient boosting methods, neural networks, and deep learning frameworks. Most models incorporated clinical and pathologic variables, while many integrated radiologic, immunologic, or genomic features. For overall survival prediction, independently validated models generally demonstrated AUCs between 0.80 and 0.90. Recurrence prediction models similarly showed favorable discrimination, with most externally validated studies reporting acceptable predictive performance. Additional prognostic endpoints including disease-free survival, progression, and nodal metastasis demonstrated AUCs ranging from 0.70 to 0.90. Common methodological limitations included retrospective design, small sample size, inadequate external validation, and risk of overfitting. Conclusions: ML-based prognostic models in OCSCC demonstrate generally favorable predictive performance across survival and recurrence outcomes. However, substantial heterogeneity in methodology and limited external validation continue to restrict clinical implementation. Future work should prioritize prospective multicenter validation, standardized reporting, and reproducible modeling frameworks. Full article

Background: Total neoadjuvant therapy is becoming a preferred option for locally advanced rectal cancer, particularly in patients with high-risk baseline features. However, real-world evidence comparing tumor response, MRI-defined high-risk feature clearance, surgical outcomes, and survival after total neoadjuvant therapy versus conventional long-course chemoradiotherapy remains limited. This study aimed to compare outcomes between total neoadjuvant therapy and long-course chemoradiotherapy in patients with locally advanced rectal cancer treated in routine clinical practice. Methods: This is a retrospective, single-centre cohort study focused on patients with stage II–III locally advanced rectal adenocarcinoma treated with curative-intent neoadjuvant therapy using either total neoadjuvant therapy or long-course chemoradiotherapy. Tumor response was assessed using restaging MRI, clinical complete response, and pathological complete response. Surgical outcomes and overall survival were evaluated. Results: A total of 110 patients were included. Patients treated with total neoadjuvant therapy had a higher baseline disease burden reflected by a greater proportion of cT4 tumors (40.6% vs. 19.2%; p = 0.014). Radiologic tumor-length response and clearance of MRI-defined high-risk features were comparable between treatment strategies. Clinical and pathological complete response rates were numerically higher in the total neoadjuvant therapy group, but the differences were not significant (cCR: 15.6% vs. 6.4%, p = 0.151; pCR: 18.5% vs. 9.7%, p = 0.301). Conclusions: In this real-world cohort, TNT was preferentially used in patients with more advanced baseline disease and showed numerically higher complete response rates, although differences were not statistically significant. Radiologic response, surgical outcomes, and short-term survival were comparable between treatment strategies. These findings support the feasibility of TNT in routine clinical practice but should be interpreted as exploratory and hypothesis-generating rather than evidence of treatment superiority. Full article

Recent advances in artificial intelligence have significantly improved performance in medical imaging tasks such as segmentation, quantification, and report generation. However, most existing solutions operate as static pipelines with limited adaptability, quality assurance, and workflow-level reasoning. In this work, we present ARGUS, an agentic framework for multimodal medical image analysis that coordinates specialized agents within a unified architecture. An Orchestrator Agent interprets user requests, identifies the imaging modality, and assembles task-specific execution plans by selectively engaging processing, quantification, verification, knowledge retrieval, and reporting agents. This enables context-aware decision-making and dynamic workflow reconfiguration based on intermediate findings and runtime conditions. A key feature of ARGUS is its ability to supervise and contextualize analytical processes. The Verification Agent performs quality control by assessing intermediate artifacts against task-specific criteria, while the Knowledge Retrieval Agent enriches quantitative findings with evidence from the biomedical literature and established physiological reference ranges. Together, these components promote transparency, support automated error detection, and reduce the risk of propagating unreliable information through downstream stages. The framework was evaluated across three imaging domains: radiology (MRI), pathology (hematopathology), and ophthalmology (OCT). Quantitative evaluation demonstrated strong agreement between ARGUS and reference standards across pathology, OCT, and MRI tasks, achieving a cell-counting bias of $-0.182$ cells (MAE = $0.727$), a full retinal thickness bias of −31.30$\mathrm{\mu }\mathrm{m}$ (MAE = 37.63 $\mathrm{\mu }\mathrm{m}$), and MRI volumetric errors below 3 $\mathrm{mL}$, while also achieving closer agreement with reference measurements than the evaluated general-purpose and domain-specific baseline systems. These results demonstrate the feasibility and potential value of agent-based orchestration for enabling adaptive, validated, and interpretable multimodal imaging workflows while providing a scalable foundation for complex multi-step clinical analysis. Full article