Search Results (206)

Background: Treatment monitoring in pulmonary tuberculosis increasingly requires assessment of residual inflammatory burden and structural lung damage beyond microbiologic response alone. High-resolution computed tomography (HRCT) can provide this information, but interpretation of serial examinations is time-consuming and partly subjective. This study did not aim to evaluate AI for the diagnosis of pulmonary tuberculosis. Instead, it explored whether artificial intelligence (AI)-assisted quantitative HRCT analysis could support longitudinal assessment of treatment-related imaging changes in patients with microbiologically confirmed pulmonary tuberculosis. Methods: We conducted a retrospective, single-center, exploratory longitudinal study of patients receiving treatment for pulmonary tuberculosis. HRCT examinations acquired at diagnosis and during follow-up were anonymized, reviewed by an expert thoracic radiologist, and processed using AVIEW Lung Texture (Coreline Soft v2.0). The software quantified total lung volume and six predefined parenchymal categories: normal lung, ground-glass opacity, consolidation, reticulation, honeycombing, and emphysema. Results: Ninety-six patients contributed 256 HRCT examinations. The most frequent software-detected abnormalities were ground-glass opacity, consolidation, and emphysema-labeled low-attenuation areas. Ground-glass opacity and consolidation showed the clearest decline across serial examinations, consistent with regression of active inflammatory disease during treatment. Reticulation showed a heterogeneous course, likely reflecting both inflammatory resolution and residual structural remodeling. Honeycombing was infrequent and quantitatively limited. Lung volume changed variably and did not consistently parallel visual improvement. A key methodological limitation was the absence of a dedicated cavity class. As a result, emphysema-labeled low-attenuation areas should not be interpreted as conventional emphysema alone, because tuberculous cavities and post-destructive abnormalities were frequently included in this category. Conclusions: AI-assisted HRCT quantification may support longitudinal assessment of pulmonary tuberculosis by providing structured and reproducible measures of interval change. However, tuberculosis-specific interpretation remains dependent on expert radiologic oversight, particularly in cavitary disease. Full article

Objective: To estimate the pooled incidence of cancer therapy-related cardiac dysfunction (CTRCD), assess longitudinal changes in global longitudinal strain (GLS) and left ventricular ejection fraction (LVEF), and summarise the available evidence comparing echocardiography and cardiac magnetic resonance (CMR) for serial functional assessment. Methods: We performed a systematic review and meta-analysis of observational studies reporting CTRCD in adults receiving cancer therapy. Pooled CTRCD incidence was estimated, study-level meta-regression explored associations between baseline mean GLS and LVEF and CTRCD event rates, and longitudinal changes in GLS and LVEF were compared between CTRCD and non-CTRCD cohorts across follow-up visits. Exploratory subgroup analyses compared findings derived from echocardiography and CMR. Results: Thirty-three studies were included (total enrolled n = 2083; analysed at follow-up n = 1973), including 27 echocardiography-only studies, 4 CMR-only studies, and 2 studies reporting both modalities at baseline. The pooled incidence of CTRCD was 27% (95% CI 0.20–0.35), with substantial heterogeneity across studies. In study-level meta-regression, baseline mean GLS (p = 0.195) and baseline mean LVEF (p = 0.668) were not significantly associated with CTRCD event rates. Compared with non-CTRCD cohorts, CTRCD cohorts showed greater worsening in GLS across all analysed follow-up intervals. Within CTRCD cohorts, both GLS and LVEF deteriorated over time, whereas non-CTRCD cohorts showed smaller changes overall. Exploratory subgroup analyses did not detect statistically significant differences between echocardiography and CMR in the available datasets. Conclusions: CTRCD is reported in approximately one-quarter of patients across published studies, although estimates vary substantially by population, therapy, follow-up schedule, and CTRCD definition. Longitudinal deterioration in GLS appears earlier and more consistently than decline in LVEF, supporting the role of serial deformation imaging in surveillance. Baseline study-level mean GLS and LVEF were not significantly associated with CTRCD event rates, and direct comparative evidence between echocardiography and CMR remains limited. Full article

Background: Arthropathies are a substantial source of global morbidity and healthcare costs, and there is a clinical need for accessible tools capable of detecting inflammatory and metabolic changes beyond conventional structural imaging. This review consolidates the recent evidence on infrared thermography (IRT) as a diagnostic and monitoring adjunct in the major arthropathies. Methods: A structured narrative review was conducted. A literature search of PubMed, Web of Science Core Collection, and Scopus was performed to identify relevant studies published between January 2016 and December 2025 using thermography- and arthropathy-related keywords and controlled-vocabulary terms combined with Boolean operators; only original full-text studies in English published within the previous decade were eligible. The structured search yielded 53 primary studies. Additional sources, including narrative and systematic reviews, methodological references, and book chapters, were drawn upon to inform the Introduction, Discussion, and interpretation but were not included in the primary evidence synthesis. Results: Across the included studies, IRT detected clinically meaningful thermal changes in most cases of osteoarthritis, rheumatoid arthritis, juvenile idiopathic arthritis, Charcot neuroarthropathy, and post-arthroplasty states, with thermal signals correlating moderately with ultrasound-detected synovitis, inflammatory biomarkers, and symptom distribution. Discussion: The evidence base is heterogeneous, however: temperature distributions overlap substantially between patients and controls, well-conducted negative results exist for hand thermography in low-activity rheumatoid arthritis, and reported effect sizes vary widely across devices and protocols. Quantitative thermographic metrics and machine-learning approaches may further refine diagnostic performance and enable remote monitoring. Conclusions: IRT is a promising rapid, non-invasive, radiation-free adjunctive imaging modality, but its clinical adoption is constrained by methodological variability, environmental and vascular confounders, and the absence of prospective validation. Standardised acquisition protocols and prospective multi-site validation are required before routine clinical use. Full article

Background: Prostate cancer (PCa) is biologically heterogeneous, requiring management strategies that balance oncologic benefit with preservation of quality of life. Prostate-specific membrane antigen (PSMA) has emerged as a key theranostic biomarker enabling highly sensitive molecular imaging and targeted therapy. Purpose: The present manuscript aims to summarize the clinical role of PSMA-PET/CT in PCa across staging, treatment selection, and response assessment, with a special focus on its contribution to personalized management. Key Findings: PSMA-PET/CT demonstrates superior accuracy compared with conventional imaging, frequently leading to stage migration and changes in therapeutic strategy. It improves detection of metastatic and recurrent disease, guides selection for systemic and PSMA-targeted therapies, and supports metastasis-directed treatment in oligometastatic settings, potentially delaying androgen-deprivation therapy and preserving quality of life. Additionally, PSMA-PET enhances intra-prostatic lesion delineation for focal therapies and radiotherapy planning. Limitations include reduced sensitivity for very small lesions, possible false positives, variability among tracers, and issues related to access and standardization. Conclusions: PSMA-PET/CT is a cornerstone of precision imaging in PCa, enabling more individualized treatment decisions across the disease continuum. Ongoing studies will further define its long-term clinical impact and integration into routine care. Full article

Background/Objectives: Photon-counting computed tomography (PCCT) combined with radiomics enables advanced myocardial tissue characterization beyond conventional imaging. This study investigated whether myocardial radiomic features derived from PCCT are associated with nicotine status in patients without coronary artery disease. Methods: In this retrospective, single-center study, 104 patients (38 men, 66 women; median age 54 years) without coronary calcification (Agatston score = 0) underwent cardiac PCCT. Myocardial septal thickness was measured at three points during the 65–70% cardiac phase. Myocardial tissue was manually segmented, and 105 radiomic features were extracted. After correlation-based feature reduction, 45 independent features were used for analysis. Patients were categorized based on nicotine status. Machine learning models, including logistic regression, random forest, and gradient boosting, were trained and evaluated using stratified five-fold cross-validation. Model performance was assessed using the area under the receiver operating characteristic curve (ROC-AUC) and additional classification metrics. Results: No significant differences in myocardial septal thickness were observed between smokers and non-smokers (p > 0.05). However, radiomic features enabled moderate discrimination between smokers and non-smokers. Logistic regression with L2 regularization achieved the best performance (ROC-AUC 0.66, balanced accuracy 0.67), outperforming random forest and gradient boosting models. The most relevant radiomic features primarily comprised higher-order texture and shape-based parameters associated with spatial gray-level heterogeneity and subtle variations in myocardial tissue architecture. Conclusions: PCCT-based radiomics may capture subtle myocardial imaging signatures associated with smoking status, even in the absence of structural changes detectable by conventional metrics. These findings highlight the potential of cardiac radiomics as a non-invasive imaging biomarker for early cardiovascular risk assessment and support its integration into advanced cardiac imaging workflows. Future multicenter studies with larger cohorts, external validation, and multimodal correlation are warranted to improve robustness and facilitate clinical translation. Full article

Background: Substance Use Disorder (SUD) is characterized by recurrent craving episodes frequently associated with emotional dysregulation and altered reward processing. This study aimed to evaluate whether emotional states associated with craving episodes can be detected through automated facial emotion recognition during controlled emotional induction. Methods: Forty-one participants completed a 14-day ecological momentary assessment (EMA) monitoring anxiety and craving levels, followed by an emotional induction task using standardized stimuli from the EmoMadrid database and addiction-related images. Facial expressions were recorded and analyzed in real time using a computational facial emotion recognition model trained on the FER-2013 dataset. Results: Participants with SUD exhibited significantly reduced positive emotional valence and emotional activation in response to positive stimuli compared with healthy controls (HC), with large effect sizes observed for emotional valence (Hedges’ g = 1.76) and emotional activation (Hedges’ g = 1.33). Item-level analyses revealed that most between-group differences occurred in stimuli depicting social interactions. Individuals with SUD also showed higher frequencies of fear-related facial expressions and lower frequencies of disgust-related expressions compared with HC, with moderate effect sizes observed for both emotional dimensions (Hedges’ g = 0.72; p = 0.02). Conclusions: These results suggest that people with SUD have changes in how they process emotions, showing less response to positive things and unique facial expressions related to craving. However, given the relatively modest and clinically heterogeneous sample, the findings should be interpreted cautiously and require replication in larger and more homogeneous populations. Full article

Edge computing on unmanned aerial vehicles (UAVs) enables low-latency wildfire monitoring by performing visual inference onboard; however, practical deployment is constrained by limited labeled data and resource budgets that often preclude reliance on large GPU servers. This work investigates transfer learning (TL) for UAV-based wildfire smoke and flame detection and evaluates its impact on both detection accuracy and edge deployment performance. We introduce the Aerial Fire and Smoke Essential (AFSE) dataset (282 aerial-view images; classes—smoke and fire), compiled from publicly available wildfire footage and FLAME2. Lightweight YOLO models are fine-tuned using heterogeneous (MS COCO) and homogeneous (FASDD) source pretraining and are assessed using mAP@0.5 together with frames per second (FPS), average inference power, energy consumption, and the normalized energy–delay product (EDP) on an edge computing platform. Results show that TL substantially improves detection accuracy on AFSE, achieving up to 79.2% mAP@0.5, while reducing training time, and improving cross-validation stability. On the tested edge platform, TL does not materially change inference speed or energy use, indicating that accuracy gains from TL do not automatically translate to improved efficiency without additional optimization. Among the evaluated lightweight detectors, YOLOv5n achieves the best mAP@0.5 while maintaining the highest edge device throughput, processing images nearly twice as fast as YOLO11n without hardware acceleration. More broadly, the measured throughput and energy differences among lightweight YOLO variants show that edge model selection should be guided by application-specific accuracy, latency, and energy constraints. Full article

Background: Therapies for metastatic colorectal cancer (CRC) are largely chosen without considering inter-patient heterogeneity, leading to unnecessary side effects without clinical benefit for some patients. Patient-derived cancer organoids (PDCOs) faithfully recapitulate the morphology and molecular profiles of the primary tumors from which they are derived, making them attractive preclinical models for predicting response to standard therapies, and thus, for use in drug development assays. Methods: Here, we investigate the hidden subclonal driver mutations in CRC PDCOs and the importance of individual PDCO-level heterogeneity when addressing PDCO treatment responses using changes in diameter and optical redox imaging. PDCO response is then compared to clinical response across a cohort of subjects with CRC receiving chemotherapy and/or chemoradiation. Results: Change in diameter and optical redox imaging are more sensitive to detecting therapeutic response than endpoint and well-level measurements. These measurements accurately reflect clinical responses to chemotherapy and chemoradiation. Conclusions: Overall, these studies demonstrate the importance of PDCO-level methods of PDCO assessment and further establish PDCOs as powerful tools for drug response assessment and developmental therapeutic studies. Full article

Binary change detection (BCD) using heterogeneous optical and SAR imagery faces challenges due to modality-specific noise and the lack of adaptive fusion strategies. Existing methods often fail to suppress SAR speckle noise and accurately localize fine boundaries. This study proposes a novel deep architecture, termed Dual-Modal Mixture-of-Experts Attention U-Net (DMoE-AttU-Net), featuring (i) dual-stream encoders for modality-specific feature extraction, (ii) a mixture-of-experts (MoE) module in the SAR stream with a gating network for dynamic fusion, (iii) Squeeze-and-Excitation (SE) and spatial attention mechanisms in the decoder, and (iv) hierarchical skip connections for multi-scale fusion. Unlike existing multimodal change detection frameworks that apply uniform feature fusion, the proposed architecture introduces a modality-aware design in which the MoE mechanism is selectively applied to the SAR stream, enabling adaptive suppression of speckle noise while preserving complementary optical information. These components collectively enhance change localization and reduce noise-induced artifacts. The proposed model achieved a mean IoU of 0.855 and a kappa coefficient of 0.836 on three optical–SAR datasets, outperforming state-of-the-art methods in both accuracy and spatial consistency. Full article

Background/Objectives: Hematopoietic stem cell transplantation (HSCT) exposes patients to cardiovascular stress through inflammation, metabolic disturbances, and prior cardiotoxic therapies. Although overt dysfunction is uncommon early after transplantation, subclinical cardiac adaptation remains poorly defined. Methods: We evaluated early electrical and mechanical cardiac responses after HSCT using integrated electrocardiographic (ECG) and echocardiographic assessment. In this prospective cohort study, patients underwent paired pre-transplant and early post-transplant (3–6 months) ECG and comprehensive echocardiography, including tissue Doppler and speckle-tracking analyses of atrial, ventricular, and right heart function. Results: Patients were stratified into multiple myeloma (MM) and non-MM subgroups. ECG voltage increased after HSCT, particularly in MM patients, without changes in left ventricular (LV) mass, geometry, or global systolic indices, suggesting electrical–structural dissociation. Left atrial (LA) reservoir strain decreased despite stable atrial volumes. Diastolic indices showed selective modulation, including a group-time interaction in the lateral e′/a′ ratio without elevated filling pressures. Subgroup analyses suggested divergent myocardial responses, with numerical global longitudinal strain (GLS) improvement in MM patients and reduced LV deformation and right ventricular (RV) fractional area change in non-MM patients. Conclusions: Early cardiac responses after HSCT were heterogeneous and compartment-specific, supporting multiparametric imaging for detection of subclinical cardiac adaptation. Full article

Background: The clinical benefit of percutaneous coronary intervention (PCI) for chronic total occlusion (CTO) remains controversial, particularly regarding left ventricular (LV) functional recovery. Global longitudinal strain (GLS) has emerged as a more sensitive marker of myocardial function than left ventricular ejection fraction (LVEF). This study aimed to evaluate the effect of CTO revascularization on LV function using GLS. Methods: This systematic review and meta-analysis were conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) guidelines. A comprehensive literature search was performed in the PubMed/MEDLINE database from inception through March 2026 using predefined search terms and Boolean operators. Reference lists of relevant articles were also screened to ensure completeness. Studies evaluating GLS before and after PCI for CTO and reporting quantitative strain data were included. Pooled effect estimates were calculated as mean differences (MDs) with 95% confidence intervals (CIs) using a random-effects model. Subgroup and sensitivity analyses were performed to explore heterogeneity and assess the robustness of the findings. Results: Six studies involving 376 patients were included. Successful CTO-PCI may be associated with an improvement in GLS (MD = 1.69; 95% CI: 1.09–2.29; p < 0.001), with substantial heterogeneity (I² = 81%). Subgroup analysis demonstrated greater GLS improvement in studies with longer follow-up durations. Sensitivity analyses confirmed the robustness of the results. Conclusions: CTO revascularization may be associated with an improvement in LV myocardial function as assessed by GLS, even in the absence of marked changes in conventional parameters such as LVEF. These findings support the clinical utility of GLS as a sensitive imaging biomarker for detecting early myocardial recovery and for guiding risk stratification in patients undergoing CTO-PCI. Full article

Vascular intervention has advanced technically faster than it has matured methodologically. Across neurovascular, carotid, peripheral, and aortic practice, complications and outcomes are often reported using different definitions, thresholds, surveillance strategies, adjudication methods, follow-up schedules, and units of analysis. As a result, studies that appear to assess the same treatment may in fact be measuring different outcome constructs. This problem is particularly visible in neurovascular intervention, where technical, radiographic, and clinical outcomes are often combined within the same evaluative framework. In acute ischemic stroke thrombectomy, changes in reperfusion thresholds can alter the meaning of procedural success. In intracranial aneurysm treatment, angiographic occlusion, retreatment, delayed stenosis, and neurological morbidity are often reported together despite representing different dimensions of efficacy and safety, while the interpretation of surrogate angiographic outcomes may vary across device classes. Similar issues arise in carotid intervention, peripheral endovascular therapy, and endovascular aneurysm repair, where composite outcomes, imaging-detected complications, and inconsistent surveillance protocols further complicate interpretation. These variations limit cross-study comparability, weaken meta-analytic synthesis, and may distort judgments about treatment effectiveness and safety. Endpoint heterogeneity persists partly through disciplinary silos, device-driven evaluation frameworks, and regulatory pathways that favor surrogate over clinical endpoints; addressing it will require not only better reporting but standardized outcome constructs, coordinated international registries, and broader adoption of core outcome set methodology. Greater discipline in endpoint definition and reporting, together with broader adoption of standardized outcome frameworks and core outcome set methodology, is needed if evidence in vascular intervention is to accumulate coherently. Full article

For irregularly shaped 820CrMnTi carburizing and nitriding parts, the challenges of high reflectivity-induced overexposure, low surface contrast, and interference from minute burrs in industrial online inspection are addressed in this paper. An innovative precision detection method integrating adaptive imaging and a dual-drive heterogeneous coupling model (RGFCN) is proposed. Such parts, due to surface photovoltaic characteristic changes caused by carburizing and nitriding heat treatment and the complex on-site lighting environment, are prone to local overexposure and “false out-of-tolerance” measurements caused by outlier sensitivity in traditional inspections. First, an innovative programmatic adaptive exposure control algorithm based on grayscale histogram feedback is introduced, which dynamically adjusts imaging parameters in real time to effectively suppress high-brightness overexposure under specific working conditions. Second, a novel adaptive main-axis scanning strategy is designed to construct a dynamic follow-up coordinate system, eliminating projection errors introduced by random positioning from a geometric perspective. Additionally, Gaussian gradient energy fields are combined with the Huber M-estimation robust fitting mechanism to suppress thermal noise while automatically reducing the weight of burrs and oil stains, achieving “immunity” to non-functional defects. Meanwhile, a data-driven innovative compensation approach is introduced. Based on sample training, gradient boosting decision trees (GBDTs) are integrated to explore the nonlinear mapping relationship between multidimensional feature spaces and system residuals, achieving implicit calibration of lens distortion and environmental coupling errors. By simulating factory conditions with drastic 24 h day–night lighting fluctuations and strong oil stain interference, statistical analysis of over 1000 mass-produced parts shows that this method exhibits excellent robustness in complex environments. It reduces the false out-of-tolerance rate caused by burrs by over 90%, and the standard deviation of repeated measurements converges to the micrometer level. This effectively addresses the visual inspection challenges of irregular, highly reflective parts on dynamic production lines. Full article

Introduction and Aim: Assessment of antitumor activity in preclinical models remains challenging when relying solely on conventional size-based imaging, particularly for complex agents such as cannabinoids, whose biological effects may not translate into early volumetric tumor changes. Cannabinoid formulations, including the synthetic cannabinoid JWH-182, Cannabixir^® Medium dried flowers, and Cannabixir^® THC full extract, exhibit diverse and potentially subtle effects on tumor biology. Radiomics enables high-throughput extraction of quantitative imaging features that capture intratumoral heterogeneity beyond gross tumor volume. The primary aim of this study was to evaluate the utility of MRI-based radiomics as a sensitive tool for detecting cannabinoid-associated tumor phenotypic modulation in a preclinical breast cancer model. Methods: Orthotopic breast tumors were induced in mice using the 4T1 cell line. Animals received cannabinoid formulations in combination with chemotherapy according to a predefined protocol. Tumor burden was assessed at baseline and post-treatment using ultrasonography and whole-body MRI to calculate tumor doubling time. T1- and T2-weighted MRI datasets were segmented and analyzed using radiomics to extract morphometric and signal-based features. Results: Conventional imaging revealed no significant differences in tumor doubling time between most cannabinoid-treated groups and controls, except for accelerated growth in animals treated with Cannabixir^® THC full extract. In contrast, radiomics identified distinct, compound-specific tumor phenotypes, including structural features consistent with reduced aggressiveness, in JWH-182-treated tumors, despite similar volumetric growth patterns. Conclusion: MRI-based radiomics sensitively captures cannabinoid-associated tumor phenotype alterations beyond volumetric assessment, supporting its value as a pharmaco-imaging tool for characterizing treatment-related tumor biology in preclinical oncology. Full article

Unsupervised change detection (CD) in heterogeneous remote sensing images is intrinsically difficult due to severe sensor-specific discrepancies. In the absence of ground truth, these discrepancies result in ambiguous optimization objectives that make it difficult for models to distinguish true land-cover changes from modality-driven pseudo-changes. To address these challenges, we propose MaskUCD, a novel unsupervised framework that reformulates heterogeneous CD as a dynamic mask-driven constraint scheduling problem. Fundamentally distinct from conventional strategies that enforce selective feature alignment, MaskUCD employs a spatially adaptive optimization mechanism. Specifically, the iteratively refined mask serves as a geometric reference to guide optimization. It enforces strict feature alignment in mask-unchanged regions to suppress modality-induced discrepancies, while simultaneously promoting feature divergence in mask-changed regions to emphasize semantic inconsistencies. In this way, explicit optimization objectives are established, together with an intrinsic interpretability constraint that guides the CD process. This strategy treats the mask as a structural guide for representation learning rather than a ground-truth reference, thereby avoiding error accumulation caused by directly using inaccurate masks as supervisory signals. To facilitate this optimization, we design a specialized asymmetric autoencoder with a hybrid encoder architecture, utilizing multi-scale frequency analysis and global context modeling to enhance feature representation capabilities. Consequently, this design enables the generation of refined and semantically consistent masks, which provide increasingly precise structural guidance, yielding converged and discriminative difference maps. Extensive experiments demonstrate that MaskUCD achieves state-of-the-art performance and superior robustness compared to existing advanced methods. Full article