Search Results (1,828)

Reliance on standalone accuracy limits credible assessment of fundus-focused large vision–language models (LVLMs), as high scores often stem from linguistic shortcuts rather than real visual reasoning. This work develops the Cognitive Audit Framework (CAF), a four-module automated auditing pipeline that dissects model reasoning flaws: Visual–Linguistic Decoupling (textual dependency via modality ablation), Hierarchical Logical Consistency (lesion–diagnosis contradiction detection), Reasoning Fidelity Gap (chain-of-thought unfaithfulness scoring), and Contextual Robustness (positional bias under option permutation). Experiments on six 7B–31B LVLMs over FunBench reveal a notable gap between benchmark accuracy and reasoning quality: high accuracy coexists with measurable textual dependency, logical inconsistencies across diagnostic levels, limited chain-of-thought faithfulness, and non-trivial positional sensitivity. CAF serves as a reproducible complement to pure accuracy metrics for validating clinical competence of ophthalmic multimodal models. Full article

Autonomous driving systems (ADSs) are reliable only when heterogeneous sensors, estimation algorithms, and safety mechanisms are engineered as a single coherent safety-critical measurement system rather than as loosely coupled modules. Production stacks integrate cameras, LiDAR, automotive radar, and GNSS/IMU, yet deployment remains constrained by modality-specific failure modes, calibration and synchronization drift, and out-of-distribution (OOD) conditions that violate modeling assumptions. These limitations induce overconfidence and downstream decision errors whenever planning assumes certainty sharper than sensing can justify. This survey introduces a sensor-centric framework linking measurement physics, uncertainty propagation, fusion integrity, safety assurance, and risk-aware planning and control. We formalize what each modality physically measures; unify probabilistic, evidential, and conformal uncertainty representations; analyze filtering, factor-graph, BEV, transformer, and state-space fusion architectures with an emphasis on robustness and graceful degradation; and generalize aviation-style integrity concepts (RAIM/ARAIM) to multi-modal autonomy. The distinctive contribution is a single sensor-to-assurance throughline in which every uncertainty representation is tied to its measurement physics, every fusion architecture is evaluated against an explicit integrity-monitoring requirement generalized from RAIM/ARAIM, and every safety-standard clause is mapped to a concrete architectural mechanism. We map these mechanisms onto ISO 26262, ISO 21448 (SOTIF), ISO/PAS 8800, ANSI/UL 4600, and the UNECE framework, and connect perception uncertainty to decision-making through chance-constrained MPC and formal safety filters (RSS, CBF). Industry case studies and emerging V2X and generative-simulation approaches close the loop to deployable safety arguments. Full article

Lymphatic malformations (LMs) are rare congenital low-flow vascular anomalies that frequently involve the head and neck and may be managed with surgery, laser therapy, sclerotherapy, or multimodal approaches depending on lesion type, size, depth, and relationship with adjacent structures. Ultrasound-guided sclerotherapy with doxycycline is an established treatment option for macrocystic lesions, whereas the practical role of high-frequency superficial ultrasound as a technical adjunct has been less specifically discussed. We report the case of a 32-year-old man presenting with a painless left submandibular swelling of approximately two years’ duration. Magnetic resonance imaging showed a well-encapsulated cystic lesion measuring 56 × 35 mm in the left submandibular region, extending into the internal paralaryngeal space and causing mild compression of the laryngeal wall. Previous fine-needle aspiration cytology had not conclusively established the lymphatic nature of the lesion; therefore, an incisional biopsy was performed and confirmed a macrocystic LM. The patient underwent day-surgery intralesional doxycycline sclerotherapy under real-time high-frequency ultrasound guidance using an 18 MHz hockey-stick transducer. After aspiration of the main cystic compartment through a 25-gauge needle, 100 mg of doxycycline diluted to 10 mg/mL in normal saline was slowly injected under continuous visualization. The procedure was well tolerated under topical local anesthesia, without pain, complications, or adverse effects. A partial clinical reduction was observed after the first session; the treatment was repeated after three months, resulting in apparent complete clinical resolution at one-year follow-up; no post-treatment imaging was available to confirm radiological resolution. This case highlights the potential technical value of high-frequency superficial ultrasonography, particularly for needle positioning, improved delineation of superficial locules, and real-time monitoring of sclerosant distribution. Full article

Road transport is one of the most significant sources of environmental pollution and greenhouse gas emissions; therefore, the development of sustainable mobility is becoming an important direction of urban transport policy. The objectives of the European Union’s transport policy encourage cities to plan and implement measures that reduce the environmental impact of transport, improve transport conditions, and increase the availability of mobility alternatives. The aim of this study is to evaluate the planning of sustainable mobility development in Lithuanian cities by analysing sustainable urban mobility plans, the measures proposed in them, and their links to the needs of urban transport systems. The study applied descriptive statistics, comparative analysis, and document content analysis methods. The urban plans of Lithuanian cities were evaluated according to the following criteria: the time scope and relevance of the plan, the completeness of the analysis of the existing transport system, the assessment of the environment and quality of life in cities, and the compliance of the planned sustainable mobility measures with the needs of the city. The results of the study show that only a portion of Lithuanian cities have prepared sustainable urban mobility plans, and their contents and analytical bases differ. Some of the plans do not provide a sufficiently detailed and relevant analysis of the current situation; therefore, the need for the selected measures is not always clearly justified. The cities analysed generally envisage or apply measures to improve public transport, develop pedestrian and bicycle infrastructure, regulate traffic, create electric vehicle infrastructure, and promote multimodality. It was concluded that sustainable mobility planning in Lithuanian cities is uneven, and its assessment depends not only on the diversity of the envisaged measures but also on the analytical quality of planning documents, the justification of measures, and the consistency of envisaged implementation measures. The study highlights the need to strengthen data-based sustainable mobility planning and to more clearly link the measures envisaged in the plans with the specific challenges of urban transport systems. Full article

Glioblastoma (GBM) is one of the most aggressive human cancers, with therapeutic failure driven by pronounced intratumoral heterogeneity, microenvironmental plasticity, immune suppression, blood–brain barrier (BBB)-related pharmacological constraints, and adaptive resistance mechanisms. A major limitation in GBM research is the lack of a human-relevant experimental system able to reproduce these dynamic features while generating interpretable, multimodal datasets. In this context, we propose a testable organ-on-chip (OoC)-extracellular vesicle (EV)-deep learning (DL) framework in which patient-derived GBM cells, endothelial cells, astrocytes, pericytes, stromal cells, and immune components are organized within perfused microphysiological systems. EVs are selectively and temporally harvested from defined compartments, and imaging, barrier-function, sensor, and EV-cargo data are integrated through modality-specific and multimodal DL architectures. This framework is intended not as an immediately validated clinical tool but as an experimental roadmap for linking EV-mediated communication to measurable phenotypes such as BBB disruption, invasion, immune reprogramming, and drug response. We critically discuss the technical requirements of BBB-on-chip systems, EV source attribution, immune-component integration, DL model selection, data scarcity, overfitting, batch effects, domain shift, regulatory barriers, cost, throughput, and reproducibility. By repositioning OoC-EV-DL integration as a staged translational strategy rather than a clinically established solution, this work aims to define a realistic and biologically grounded route for advancing precision oncology in GBM. Full article

Background: Intracranial pressure (ICP) monitoring remains the cornerstone of neurocritical care in severe traumatic brain injury (TBI), yet its prognostic value as a standalone metric is limited. The Pressure Reactivity Index (PRx), a continuous measure of cerebrovascular reactivity derived from ICP and arterial blood pressure, may offer additional or complementary prognostic information. This systematic review aimed to compare the prognostic performance of PRx-derived metrics versus standard ICP monitoring for mortality and functional outcome in patients with TBI. Methods: A systematic search of PubMed, Web of Science, and Scopus was conducted for studies published between January 2000 and December 2025. Studies were eligible if they included adult TBI patients with continuous multimodal monitoring and reported comparative prognostic data for PRx- and ICP-based metrics. Risk of bias within the studies was appraised via the QUIPS tool, and the GRADE system was used to rate the strength of the evidence. Due to methodological heterogeneity, findings were synthesized narratively. Results: Nine studies were included. Applying a maximum-cohort estimation to account for overlapping registries, the pooled sample comprised a minimum of 1240 unique patients. In the majority of included studies, direct within-cohort head-to-head comparisons demonstrated that specific PRx-derived metrics—such as the individualized ICP threshold (iICP), Longest Continuous Duration of Autoregulatory Impairment (LCAI), Lower Limit of Reactivity (LLR), and time-integrated burdens (%Time > Threshold)—yielded stronger prognostic discrimination compared to standard ICP thresholds for both mortality (PRx: AUC 0.747–0.648 and ICP: AUC 0.660–0.614) and functional outcome. When added to established predictive models, PRx-derived metrics provided clinically meaningful incremental improvements in prognostic accuracy, with descriptive incremental AUC gains ranging from +0.039 to +0.170 across the six studies reporting model augmentation. Due to heterogeneity in baseline models, PRx-derived metrics, and patient populations, these findings are presented strictly as a descriptive range. Conclusions: PRx and PRx-derived cerebrovascular reactivity metrics-namely iICP, LCAI, LLR, and time-integrated burdens of autoregulatory failure—show potential to offer additive prognostic value beyond standard ICP monitoring in severe TBI. However, because current evidence is strictly observational and likely influenced by institutional confounders, it cannot currently support definitive clinical recommendations. Further prospective, multicenter studies utilizing standardized thresholds are necessary to confirm these associative findings and isolate their true prognostic value. Full article

Rapid and non-contact surface inspection is essential for quality control in modern production. Optical coherence tomography (OCT) can image a surface without contact, but turning those images into roughness parameters usually requires specialized processing software. This study examined whether a multimodal large language model (LLM) could estimate roughness parameters directly from OCT B-scans as a screening tool. The study was designed as a controlled macro-scale proof of concept using periodic, analytically defined phantoms rather than as validation on stochastic industrial micro-roughness. Five test surfaces with exactly known geometries were designed, 3D-printed, and scanned with a spectral-domain OCT system. For each surface, roughness values were computed from the theoretical shape, extracted from the OCT image using MATLAB, and also estimated by the LLM from the same image. The repeatability of the LLM was checked by running the same prompt ten times per surface. On a sawtooth profile, the LLM estimates varied by 3.8% for Ra, 4.2% for Rq, 3.5% for Rp, 2.8% for Rv, and 3.1% for Rt. Across all five surfaces, the variation in Ra and Rq was around 3–5%, and for Rt, it stayed below 5%. The results show that a generative AI approach can produce repeatable roughness estimates that are useful for comparative screening. This method offers a flexible option for surface comparison and AI-assisted quality control when calibrated measurements are not required. Full article

Critically ill obese patients with gynecological cancer represent a high-risk population with complex nutritional needs. Although excess adiposity may suggest adequate energy reserves, it often conceals sarcopenia, micronutrient deficiencies, and functional malnutrition, contributing to impaired wound healing, immune dysfunction, prolonged mechanical ventilation, increased susceptibility to infections, and adverse oncologic outcomes. Obesity-associated inflammation, insulin resistance, and tumor-driven catabolism further exacerbate metabolic stress and complicate nutritional management in the intensive care setting. Accurate nutritional assessment requires a multimodal approach incorporating body composition analysis, functional measures, and laboratory parameters, as conventional indices such as body mass index may underestimate nutritional risk. Nutritional support should be individualized and may include early enteral nutrition to preserve gut integrity, supplemental or total parenteral nutrition when gastrointestinal function is compromised, high-protein regimens, and targeted micronutrient replacement. Immunonutrition, including arginine, glutamine, omega-3 fatty acids, and nucleotides, has emerged as a promising strategy to modulate inflammation, enhance immune function, and support tissue repair. This narrative review summarizes current evidence regarding obesity-related metabolic dysfunction, nutritional assessment, enteral and parenteral nutrition, and immunonutrition in obese critically ill patients with gynecological cancer. It highlights the challenges associated with sarcopenic obesity and hidden malnutrition while providing a clinically relevant overview for intensivists, gynecologic oncologists, surgeons, and nutrition specialists. Early recognition of nutritional risk and implementation of individualized multimodal nutritional strategies may improve recovery and clinical outcomes. However, high-quality ICU-specific studies remain limited, and further prospective research is needed to establish evidence-based nutritional protocols and evaluate their impact on survival, treatment tolerance, and quality of life in this vulnerable population. Full article

Earthquake risk communication often remains centered on event parameters and structural collapse, while local site effects, building response and non-structural elements vulnerability shape how earthquakes are experienced and what people can do to reduce risk. This study examines whether a multi-modal, experience-based strategy focused on these dimensions, which are referred to as “last mile” of seismic risk, can improve public understanding and support actionable preparedness behaviors. The case study is the exhibition “Terremoti: Attenti agli Elementi!—Dettagli che salvano la vita” (Earthquakes: Beware of the Elements!—Details that Save Lives), designed for school audiences and the general public. Its effectiveness was assessed through five multiple-choice questions administered before (N = 183) and after (N = 174) the visit to the Genoa Science Festival; responses were analyzed overall and by topic and demographic group. Correct answers increased significantly from pre- to post-visit, with the largest gains concerning local site effects (+43.29%) and household prevention measures (+49.45%), whereas building vulnerability (+14.97%) and building dynamic response (+0.49%) showed more limited improvement. These exploratory results suggest that seismic risk communication is more effective when abstract concepts are translated into observable, manipulable, and everyday experiences, and support a shift toward a “last-mile” framework of seismic risk communication. Full article

Extreme weather events exacerbated by global warming pose severe threats to urban safety, underscoring the urgent need for highly accurate precipitation nowcasting. Short-term local heavy precipitation remains a particular challenge for traditional forecasting due to its suddenness and high disaster potential. To address this, we propose a multi-modal fusion framework that integrates ground-based GNSS-derived Precipitable Water Vapor (GNSS-PWV) and ground-based Radar Composite Reflectivity (CR). While GNSS-PWV keenly captures pre-convective atmospheric water vapor accumulation, radar CR details the morphological distribution of hydrometeors. Specifically, we developed the Spatio-Temporal Enhanced Attention Swin U-Net (STEA-Swin) model to synergize these heterogeneous datasets over the Beijing–Tianjin–Hebei region. High-precision PWV was retrieved from 250 Continuously Operating Reference Stations (CORS) using the dual-frequency ionosphere-free Precise Point Positioning (PPP) method, achieving a strong correlation (>0.97) with ERA5 reanalysis data. Validated against measured data from the 2025 flood season, the STEA-Swin model achieved a Probability of Detection (POD) of 0.68 for torrential rain events at a +1 h forecast lead time. Notably, compared to single-source models, the Critical Success Index (CSI) and POD for torrential rain improved by 18.5% and 21.5%, respectively. These findings demonstrate that coupling deep learning with ground-based GNSS-derived atmospheric thermodynamic information can significantly enhance early warning capabilities, providing a promising technical approach for regional disaster prevention and climate resilience. Full article

Objectives: This study proposes an implementation-oriented design framework for multimodal conversational agents handling patient-generated health data and reports an exploratory experiment evaluating its instantiation in hypertension self-monitoring, focusing on user experience of conversational data-entry workflows. Methods: The framework operationalizes four complementary dimensions (social intelligence, communication style, anthropomorphic characteristics, and technological mapping) and was instantiated in two agents integrated into an eHealth platform. Each agent supports users by providing prompts, interpreting responses, checking data plausibility, and confirming submission. A three-arm, single-session feasibility experiment ($n=18$, $n=6$ per group) compared a conventional app interface with text-based and voice-based conversational agents. Evaluation triangulated three sources of evidence: open-ended qualitative responses analyzed through descriptive content analysis, session-level researcher observation notes, and the User Experience Questionnaire (UEQ) reported descriptively with one-way ANOVA and ${\eta }^2$ effect sizes. Results: All three modalities were acceptable to participants and produced UEQ scores in the positive range. Hesitation was observed in 2 of 6 Control participants, 1 of 6 Text participants, and 3 of 6 Voice participants, with self-reports indicating that voice-related difficulties were modality-specific (diction, command phrasing) and resolved within the session. Qualitative themes of acceptability and innovation, perceived effort, and modality-specific facilitators emerged across the corpus. Between-group ANOVAs did not reach statistical significance ($p>0.05$), as expected for an underpowered design, yet ${\eta }^2$ values were medium for Attractiveness, Efficiency, Dependability, and Pragmatic Quality and large for Stimulation and Hedonic Quality, converging with the qualitative innovation and engagement signal in the conversational conditions. Conclusions: The framework and feasibility experiment provide preliminary, hypothesis-generating evidence on the potential of multimodal conversational interfaces in healthcare. However, no clinical, behavioral, or longitudinal outcomes were assessed. The four design dimensions can be tentatively associated with themes recognizable in user discourse, and the observed effect-size pattern motivates adequately powered longitudinal studies that incorporate behavioral and clinical endpoints alongside user experience measures. Full article

Mixed reality (MR) technologies enable users to experience computer-generated content within the physical environment through spatial computing and head-mounted displays. By supporting real-time interaction through speech, gesture, gaze, and movement, MR offers new opportunities for game design beyond productivity and educational applications. However, relatively few studies have examined interaction modalities in MR games. In this paper, we present the design and deployment of four MR games on the Microsoft HoloLens 2: three that use monomodal input (speech, gaze, or gesture) and one that uses multimodal input (speech, gaze, and gesture). We conducted a study with ten participants and evaluated player experience using subjective self-reports of task load, emotional engagement, and comfort alongside objective measures, namely brain activity data collected with a five-channel electroencephalogram (EEG) device. Our preliminary findings suggest two clusters of interaction modalities based on subjective measures, a pattern that is also reflected in the objective EEG measures. Our analysis combining subjective and EEG data indicates that interaction modality influences task load and emotional engagement. Additionally, our functional connectivity analysis showed links in activity across the prefrontal, temporal, and occipital brain regions for different input modalities in the MR games. Full article

Background: Brain-derived neurotrophic factor (BDNF) is crucial in the nociception and mechanisms underlying chronic and neuropathic pain. The evaluation of circulating BDNF in patients with multimodal analgesia has not been reported previously. We hypothesized that opioid-based multi-analgesia induces changes in BDNF values and that BDNF correlates with pain intensity in neuropathic pain. Methods: Adult patients who met low back pain (LBP) criteria and received multimodal opioid-based therapy were included. The control group included patients with LBP who did not receive any pharmacotherapy. Plasma measurements obtained with the ELISA test were analyzed. The study was registered at Clinical Trials.gov (NCT 04227223). Results: Patients with multimodal opioid-based analgesia had significantly higher BDNF values compared to the monotherapy: 3.6 ng/mL vs. 2.7 ng/mL, p = 0.01. No statistical differences were observed compared to the non-pharmacologically treated group: 3.6 ng/mL vs. 5.0 ng/mL, p = 0.75. The median BDNF values were lowest in the mild-pain group, and significant differences were observed between the severe and moderate-pain groups (p = 0.006) and the mild-pain group (p = 0.0001). BDNF was significantly higher in the neuropathic-pain group compared to the group of patients without neuropathic pain (p = 0.0005). A significant correlation was demonstrated between the BDNF and numerical rating pain score (NRS) in the neuropathic-pain component (rho = 0.6, p = 0.001). Conclusions: Multimodal opioid-based analgesia decreases plasma BDNF concentrations less than opioid monotherapy, which offers an opportunity to limit opioid-induced adverse effects. BDNF influences pain intensity and predicts neuropathic pain in multimodal opioid-based analgesia. Full article

The following narrative review discusses the use of deep learning and 3D modeling in facial reconstruction from skeletal remains, focusing on accuracy, algorithmic bias, and evidential reliability. Forensic facial reconstruction (FFR) is a multidisciplinary field combining anthropology, medicine, and visual sciences to approximate the facial appearance of unidentified individuals from skeletal remains. Traditional manual methods, based on anatomical knowledge and facial soft tissue thickness (FSTT) measurements, are limited by subjectivity, labor intensity, and inter-expert variability. This narrative review summarizes contemporary AI-assisted approaches, with emphasis on convolutional neural networks (CNNs), generative adversarial networks (GANs), variational autoencoders (VAEs), and diffusion models, which enable probabilistic prediction of facial morphology while accounting for demographic variables such as sex, age, and population ancestry. Key challenges affecting reconstruction accuracy—including dataset limitations, population-specific variability, and algorithmic bias—are discussed, alongside quantitative validation methods and concerns regarding model transparency. Legal and ethical considerations, such as privacy, biometric data protection, and the need for explainable AI (XAI) frameworks, are highlighted. Future perspectives include hybrid expert–AI workflows, the development of globally representative datasets, and the integration of multimodal data sources, including DNA phenotyping, 3D morphometrics, and biomechanical modeling. These advances aim to create standardized, interpretable, and biologically informed frameworks that enable AI to support expert judgment and enhance the reliability of forensic facial reconstructions. Full article

Background: Hepatocellular carcinoma (HCC) is a leading indication for liver transplantation (LT), representing a curative treatment option for selected patients. A remaining clinical challenge is the recurrence of HCC after transplantation, impacting long-term graft and patient survival. The impact of different bridging therapies (BTs) such as transarterial chemoembolization (TACE), local ablation or liver resection on recurrence rates remains unclear. We assessed post-transplant HCC recurrence and survival focusing on the role of pre-transplant bridging therapies. Methods: Adult recipients undergoing LT for HCC at Hannover Medical School from January 2007 to September 2022 were retrospectively analyzed (n = 185). Recurrence was defined as confirmed intra or extrahepatic HCC after LT. Overall survival (OS) and recurrence-free survival (RFS) were analyzed using Kaplan–Meier estimation and log-rank testing; multivariable Cox proportional hazards regression was used to identify independent factors influencing OS. Results: Pre-transplant BT was administered in 85.4% of patients, consisting of only TACE, (n = 20; 10.8%), local ablation, (n = 32; 17.3%), liver resection (n = 27; 14.6%) or a multimodal approach (n = 50; 27%). Post-transplant HCC recurrence rate was 9.2% with a median time to recurrence of 845 days (range 126–3978 days). Patients with post-transplant HCC recurrence had a significantly higher prevalence of viral hepatitis (70.6% vs. 57.1%; p = 0.01), higher pre-transplant AFP peak levels (37.5 vs. 10 ng/mL; p = 0.03), larger tumor sizes (median 3.95 cm vs. 2.6 cm; p = 0.03) and more poorly differentiated tumors (G3; 25.0% vs. 5.3%, p = 0.04). Kaplan–Meier analysis showed significant overall differences in OS and RFS among bridging therapy groups (p = 0.03). In the subgroup of early HCC < 3 cm, local ablation was associated with significantly improved OS compared to TACE (p = 0.035). Last measured pre-transplant AFP < 15 ng/mL was a significant predictor of both improved OS (p = 0.006) and RFS (p = 0.008), whereas peak AFP did not reach significance after correction. Multivariable Cox regression revealed HCC recurrence, high recipient BMI and low LabMELD as independently associated with reduced OS after LT. Median OS after HCC recurrence was 13 months. Conclusions: Our monocentric retrospective data indicate that post-transplant HCC recurrence is uncommon but remains a challenge regarding life expectancy and is influenced by pre-transplant bridging therapy. In the subgroup of early HCC < 3 cm, local ablation was associated with significantly improved OS compared to TACE. Last measured pre-transplant AFP < 15 ng/mL was associated with both improved OS and RFS, suggesting that treatment response may also represent a prognostically relevant factor. Further prospective validation of contemporary locoregional and systemic bridging approaches, especially in the context of tumor biology and treatment response, is warranted. Full article