搜索结果：outperform

Lesion-symptom mapping results can vary substantially as a function of specific analysis parameters, but the extent to which individual methodological choices interact to modulate the sensitivity and specificity of results is not clear. Here, we employed a large-scale simulation approach to inform practical recommendations for lesion-symptom mapping studies. Routine clinical imaging from 959 stroke survivors (mean age = 72.5, 49.3% female) was used to conduct 384,780 lesion-symptom mapping analyses based on simulated behavioural data. Each simulated analysis used different combinations of plausible sample inclusion criteria, analysis parameters (e.g., correction factors), analysis types (e.g., univariate vs. multivariate), and underlying target correlates. Simulated analysis accuracy (percent coverage of target correlates, Dice similarity coefficient, and false positives) was compared across designs. Overall, analysis accuracy varied widely and was substantially modulated by the specific design used. Analyses that maximised lesion coverage by including large and diverse samples reliably outperformed analyses using more restricted samples. Analyses using direct total lesion volume controls outperformed analyses using other (or no) volume corrections across all accuracy measures. False discovery rate corrections yielded the best performance in terms of target coverage, while permutation corrections yielded the best Dice coefficients. While multivariate approaches were more accurate than univariate analyses in terms of Dice coefficient, univariate analyses generated higher target hit rates and percent target coverage. These results identify specific analysis designs suitable for studies aiming to maximise their sensitivity and/or specificity to underlying critical correlates, while highlighting the inferential strengths and weaknesses of these complementary approaches.

Right ventricular failure (RVF) remains a frequent complication following Left Ventricular Assist Device (LVAD) implantation. Conventional risk scores rely on preoperative data acquired under stable conditions, which may overlook early signs of RV decompensation that emerge under perioperative stress. This study evaluates whether perioperative haemodynamic markers prior to cardiopulmonary bypass (CPB) improve RVF prediction compared to preoperative assessment obtained by right heart catheterisation (RHC), echocardiographic parameters and scoring systems such as the EUROMACS-RHF score. We conducted a retrospective single-centre study of 203 LVAD patients (2013-2023). The primary outcome was RVF. Preoperative data included clinical variables, echocardiography, and RHC. Perioperative data included vasoactive-inotropic score (VIS), pulmonary artery catheter and arterial-line monitoring between induction and start of CPB. Two multivariable models were developed using LASSO logistic regression: (1) Preoperative model; (2) Pre-CPB model. Model performance was assessed using area-under-the-curve (AUC), and compared against EUROMACS-RHF using DeLong tests. 36 patients (17.7%) developed RVF. No preoperative RHC-derived parameter independently predicted RVF. The Preoperative model achieved moderate discrimination (AUC: 0.747). Incorporating intraoperative mean arterial pressure (MAP), pulmonary artery pulsatility index (PAPi) and maximum VIS, yielded a Pre-CPB model with strong discrimination (AUC: 0.834). The Pre-CPB model significantly outperformed both EUROMACS-RHF (p = 0.0299) and the Preoperative model (p = 0.049), with no significant difference between our Preoperative model and EUROMACS-RHF (p = 0.222). Perioperative markers prior to CPB significantly outperform preoperative prediction of RVF post-LVAD. Incorporating these into real-time risk stratification can improve decision-making, enabling earlier RV support in the operating theatre for high-risk LVAD patients.

This study aimed to generate medical qualification exam questions and their corresponding answers from real-world electronic health records (EHRs) with large language models (LLMs), and to compare their output to that of human medical experts. Utilizing a multicenter bidirectional anonymized database China Elderly Comorbidity Medical Database (CECMed), a total of 8 LLMs: ERNIE 4, ChatGLM 4, Doubao, Hunyuan, Spark 4, Qwen, Llama 3, and Mistral were tasked with generating open-ended questions and answers based on a subset of sampled admission reports. LLMs generated the medical question and answer through few-shot prompting. An independent expert panel scored the AI-generated outputs based on multiple criteria, including coherence, sufficiency of key information, information correctness, factual consistency, evidence of statement, and professionalism, using 5-point Likert scales. For question generation, ERNIE 4 achieved the highest cumulative score (16.47). Human experts surpassed LLMs in sufficiency of key information (3.67) but lagged in information correctness (3.63 vs. LLMs' 4.03-4.57). The information correctness of ERNIE was significantly higher than the human's [0.93 (0.62, 1.24), p&#x2009;<&#x2009;0.01]. For answer generation, humans led overall (14.49), while Doubao outperformed the other LLMs in coherence (3.57), factual consistency (3.60), and professionalism (3.53). The coherence of human's was significantly better than that of 8 LLMs, especially outperformed Llama [0.8 (0.37, 1.23), p&#x2009;<&#x2009;0.01] and Mistral [0.87 (0.45, 1.28), p&#x2009;<&#x2009;0.01]. Conventional medical education requires clinicians to formulate questions and answers based on prototypes from EHRs, which is heuristic and time-consuming. This study shows that mainstream LLMs could generate questions and answers with real-world EHRs at levels close to clinicians. Although current LLMs performed dissatisfactorily in some aspects, medical students and interns may find LLMs a useful auxiliary tool to support their learning. https://clinicaltrials.gov/study/NCT06316544, identifier: NCT06316544.

The Coronary Artery Disease Reporting and Data System (CAD-RADS) standardizes coronary CT angiography (CCTA) reporting, but not all reports contain CAD-RADS classifications. We benchmarked 54 large language model (LLM) configurations across 50 distinct models, including recent proprietary and open-weight reasoning models, for zero-shot CAD-RADS classification. We retrospectively analyzed 500 anonymized CCTA reports from four hospitals across three U.S. regions. Expert cardiovascular radiologists provided the reference standard (human inter-rater [Formula: see text]). Fifty-four model configurations (50 distinct LLMs; four configurable models tested in both thinking and non-thinking modes) spanning Llama 2 7B through recent thinking models (DeepCogito v2, Gemini 3 Pro) processed reports using identical zero-shot prompts. Performance was measured with unweighted Cohen's &#x3ba;. Two LLMs met both pre-specified non-inferiority criteria ([Formula: see text] margin and entire 95% CI within the human inter-rater agreement band, [Formula: see text]-0.956): Claude 4.6 Opus ([Formula: see text], 95% CI 0.817-0.889) and the open-weight Gemma 4 31B ([Formula: see text], 95% CI 0.810-0.882), which ranked second overall. Both met the criteria at the pre-specified [Formula: see text] margin; at [Formula: see text] no model qualified. On the reports originally dictated without a CAD-RADS statement (n=343), the top models reached [Formula: see text]. Performance declined with longer thinking chains (proxy for case complexity), but thinking-mode outperformed non-thinking mode on matched difficult reports. Gemma 4 31B remained non-inferior at 3-bit quantization and fits on a 24&#xa0;GB consumer GPU. Current LLMs extract the CAD-RADS stenosis severity category from unstructured CCTA reports with agreement approaching the human inter-rater band, without task-specific training. Our data shows the remarkable rise of open models. A 31B open-weight model matched top proprietary systems and runs on consumer GPUs, enabling privacy-preserving local deployment for clinical data mining.

Sustainable energy conversion depends on the development of effective and economical electrocatalysts. In this work, we highlight the development of cobalt oxide (Co3O4) as an electrocatalyst by employing a scalable and economical Successive Ionic Layer Adsorption and Reaction (SILAR) method onto an electrically activated pencil graphite (Ac-PGE) as an affordable substrate for monitoring the oxygen evolution reaction (OER). According to electrochemical impedance spectroscopy, the SILAR process produced uniform deposition and improved surface activation, which resulted in a considerably reduced charge transfer resistance (R ct) of 0.08 k&#x3a9;. The OER overpotential was observed at 240 mV at 10 mA cm-2 with a Tafel slope of 47.57 mV dec-1, and a turnover frequency of 0.082 s-1 at the activated electrode. LSV and OCP demonstrate that Co3O4@Ac-PGE performs better electrochemically than the other electrodes under investigation (In-PGE, Ac-PGE, and Co3O4@In-PGE). Additionally, after 8 hours, it maintained more than 93% of its initial activity, demonstrating exceptional endurance. Overall, it was observed that the Co3O4@Ac-PGE electrode developed by the SILAR method outperforms a number of traditional and noble-metal-based catalysts and offers a practical, long-lasting, and financially sustainable approach to effective water-splitting and renewable energy conversion. The structural and surface properties of the modified electrodes were investigated using energy-dispersive X-ray spectroscopy (EDX), field emission scanning electron microscopy (FESEM), and X-ray photoelectron spectroscopy (XPS). This work shows a scalable and cost-effective strategy to design an efficient electrocatalyst by using SILAR for OER, which can contribute towards Green Hydrogen production.

Chronic cardiopulmonary diseases, including chronic obstructive pulmonary disease, interstitial lung disease, and coronary artery disease, represent a major global health burden. Low-dose computed tomography (LDCT) combined with artificial intelligence (AI) quantitative imaging enables the identification of cardiopulmonary imaging abnormalities in asymptomatic individuals. To evaluate early risk factors for cardiopulmonary imaging abnormalities in asymptomatic middle-aged and elderly population using single-inspiratory phase LDCT combined with AI-based whole-lung quantitative analysis. A retrospective collection was conducted on 1035 asymptomatic individuals aged &#x2265; 40 years who underwent routine single-inspiratory-phase LDCT screening at Zhuzhou 331 Hospital in 2025. An AI platform was utilized to automatically extract airway wall area percentage, low-attenuation area percentage, interstitial lung abnormality, and coronary artery calcification score. Based on risk stratification criteria, the population was divided into a high-risk group (n = 689) and a low-risk group (n = 346) and randomly stratified into a training set (n = 724) and a validation set (n = 311) at a 7:3 ratio. Independent sample t-test or &#x3c7; 2 test was applied to analyze between-group differences. Binary logistic regression was used to screen independently associated factors, and a multivariate logistic regression model was constructed after excluding collinear variables using variance inflation factor (VIF < 5), followed by the establishment of a nomogram prediction model. The receiver operating characteristic curve and DeLong test were employed to evaluate model discrimination. The Hosmer-Lemeshow test and calibration curves were used to assess goodness of fit. Decision curve analysis (DCA) was applied to evaluate clinical net benefit, and internal validation was performed using the bootstrap method (1000 resamplings). Univariate analysis showed that smoking history, abnormal metabolic status, body mass index (BMI), age, and gender were significantly associated with cardiopulmonary imaging-defined high-risk status (P < 0.05), while work style showed a marginal association in univariate analysis (P = 0.043) but did not retain statistical significance in the multivariate model (P = 0.851). After VIF collinearity screening (all VIF < 5) and multivariate logistic regression analysis with forced entry of six candidate variables, smoking history [odds ratio (OR) = 1.968 per level, 95% confidence interval (CI): 1.665-2.325, P < 0.001], abnormal metabolic status (OR = 3.266, 95%CI: 2.259-4.723, P < 0.001), BMI (OR = 1.150 per unit, 95%CI: 1.079-1.226, P < 0.001), and age (OR = 1.028 per year, 95%CI: 1.005-1.052, P = 0.018) were identified as independently associated factors for cardiopulmonary imaging-defined high-risk status, while male gender demonstrated an (inverse) association (OR = 0.427, 95%CI: 0.276-0.660, P < 0.001) after adjustment for smoking and other covariates. The nomogram-based risk stratification model integrating the above five independently associated factors achieved an area under the curve (AUC) of 0.833 (95%CI: 0.787-0.879) in the validation set, significantly outperforming the baseline model containing only age and gender (AUC = 0.647, 95%CI: 0.582-0.712; DeLong test: Delta AUC = 0.186, z = 5.256, P < 0.001). The Hosmer-Lemeshow goodness-of-fit test confirmed satisfactory calibration (training set: &#x3c7; 2 = 10.40, P = 0.238; validation set: &#x3c7; 2 = 6.50, P = 0.591). Calibration curves and DCA demonstrated good agreement and positive clinical net benefit. Bootstrap internal validation (1000 resamples) confirmed model robustness (mean AUC = 0.792, 95%CI: 0.760-0.824). The LDCT-based nomogram model combined with AI quantitative imaging analysis may assist in identifying individuals with cardiopulmonary imaging abnormalities in asymptomatic screening populations, potentially guiding further diagnostic evaluation. Prospective studies are warranted to establish its role in improving clinical outcomes.

The differentiation of benign and malignant breast nodules, particularly those categorized as Breast Imaging Reporting and Data System (BI-RADS) 3-4, remains a clinical challenge due to the subjectivity and operator dependence of conventional ultrasound assessment. This study aimed to evaluate the diagnostic value of intratumoral and peritumoral ultrasound radiomics features in distinguishing benign from malignant BI-RADS 3-4 breast nodules and to construct interpretable machine learning models. Ultrasound images and parameters (BI-RADS classification) of breast nodules were retrospectively collected from female patients at two institutions between January 2021 and June 2024: 571 patients (880 nodules) from Institution 1 (Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine) and 333 patients (351 nodules) from Institution 2 (Shanghai Second People's Hospital). Included patients had BI-RADS 3-4 nodules confirmed by pathology or, for BI-RADS 3 nodules, stable findings on follow-up for at least 2 years. Nodules from Institution 1 were randomly divided into a training set (n = 615) and an internal validation set (n = 265) at a 7:3 ratio, while patients from Institution 2 constituted an external test set (n = 351). Radiomics features of intratumoral and peritumoral regions (3 and 5 pixels wide) were extracted using PyRadiomics. Features were screened using the independent t-test, Spearman correlation, and least absolute shrinkage and selection operator (LASSO) regression. Machine learning models-including random forest (RF), multilayer perceptron (MLP), extra trees (ET), support vector machine (SVM), logistic regression (LR), k-nearest neighbor (KNN), eXtreme Gradient Boosting (XGBoost), Adaptive Boosting (AdaBoost), Gradient Boosting Decision Tree (GBDT), and Light Gradient Boosting Machine (LightGBM)-were trained and compared using area under the curve (AUC) and other metrics. The best-performing model was then used to compare intratumoral versus peritumoral regions. SHapley Additive exPlanations (SHAP) was applied to interpret feature importance. Across the training, internal validation, and external test sets, SVM demonstrated the most stable and balanced performance. The 3 pixels transitional zone region of interest support vector machine (EI3_SVM) model achieved a higher AUC than the tumor core region of interest (T) model in the external test set (0.875 vs. 0.787, p < 0.01), with accuracy 78.1%, sensitivity 42.2%, specificity 97.0%, and Brier score 0.166, outperforming other peritumoral models. Most models incorporating peritumoral features demonstrated AUCs that were higher than or comparable to those of the T model. SHAP analysis indicated that the high specificity was mainly driven by texture and shape features. Integrating intratumoral and peritumoral radiomics features significantly improves the diagnostic accuracy and objectivity for differentiating benign and malignant breast nodules. This approach may aid clinical decision-making, reduce unnecessary biopsies, and support early precision diagnosis of breast cancer.

Metalloporphyrin-based complexes featuring stable tetrapyrrole structures and tunable coordination environments are promising electrocatalysts for CO2 reduction. Modifying the first-coordination atoms that bonded with the central metal or changing the outer-sphere functional groups was found to affect the catalytic performance. However, the combined effect of these two strategies remains unexplored. Herein, using nickel tetraphenyl porphyrin (NiN4-TPP, 1) as the original catalyst, we first investigated the effect of first-coordination modification by replacing one N atom with a S atom, yielding NiN3S-TPP (2). Density functional theory calculations revealed that, as an electrocatalyst, NiN3S-TPP outperforms NiN4-TPP in converting CO2 to CO. Notably, the S-substituted catalyst 2 exhibits a positively shifted reduction potential&#xe5f8;a trend consistent with experimental observations. This is attributed to the lowered LUMO energy level after introducing the S atom. Moreover, the S-substitution stabilizes the [*-COOH] intermediate and shifts the rate-determining step from the protonation of [1-CO2]- to form [1-COOH]0 to the absorption of CO2 on [2]-, reducing the barrier. Charge analysis reveals that the S atom donates electrons to the metal center and porphyrin ligands of the catalyst. Then, upon binding to COOH, these reserved electrons are transferred to COOH, thus facilitating the interaction and stabilizing the intermediate. Moreover, peripheral substituent modifications with the cationic -N-(Me)3 + functional group yield great enhancement in the catalytic performance. It shifts the reduction potential to be more positive and further stabilizes the [*-COOH] intermediate via through-space electrostatic interaction. This work demonstrates how first- and peripheral-coordination modifications synergistically enhance CO2 reduction catalysis, offering a strategy for designing efficient porphyrin-based electrocatalysts through core coordination environment regulation and through-space interactions.

The global infection rates caused by multidrug-resistant bacteria continue to rise, while the pipeline for novel antibiotics is increasingly drying up, highlighting the urgent need to develop new antimicrobial agents. Drawing inspiration from the amphiphilic structure of cationic antimicrobial peptides (AMPs), a collection of amphiphilic guanidinium salts incorporating cinnamic acid skeleton was designed and prepared. Bioactivity screening revealed that compound 10c exhibited excellent inhibitory effects against Gram-positive bacteria, with MIC values ranging from 1 to 2&#x202f;&#x3bc;g/mL, comparable to the clinically used drug vancomycin. Further evaluation demonstrated that 10c possessed negligible hemolytic activity, infrequent resistance acquisition, low cytotoxicity, fast bactericidal action, and good plasma stability, indicating strong potential for further development. Additionally, 10c not only effectively prevented biofilm formation but also significantly disrupted pre-formed biofilms. Mechanistic studies revealed that 10c achieved selective membrane targeting by specifically interacting with phosphatidylglycerol present in the bacterial cell membrane. This interaction triggered membrane depolarization, increased membrane permeability, leading to elevated intracellular ROS levels and escape of cellular contents, ultimately accelerating bacterial death. More importantly, 10c significantly reduced bacterial burden and mitigated tissue inflammation, outperforming vancomycin in a murine skin abscess model. In summary, these results indicated that compound 10c was a membrane-active antimicrobial candidate with promising potential for further development.

Mental chronometry tasks commonly presume that real-time action imagery indicates high action imagery ability (AIA). However, factors that systematically influence imagery durations undermine the validity of AIA scores based on deviations from real-time action imagery. Therefore, we introduce and validate a novel approach for measuring AIA using mental chronometry. This approach relies on the relationship between a movement's difficulty and its duration (Fitts' law), with AIA scores quantifying the degree to which this relationship is preserved in action imagery. In a Chronometric Radial Fitts' Task (CRFT), participants (N&#x2009;=&#x2009;30) executed and imagined both tapping (with a stylus) and clicking (with a computer mouse) radially arranged targets. Single execution and imagery durations were isolated through simultaneous space bar presses. Targets varied in difficulty according to Fitts' law, which held across all four conditions (for both tools, when executing and imagining the movements). The AIA scores derived from the proposed Fitts' law approach outperformed those derived from earlier mental chronometry approaches, being the only scores indicating convergent validity across tools. Using the stylus, larger absolute and relative deviations from real-time action imagery were consistently associated with higher subjective AIA, reinforcing validity concerns. Participants struggled to incorporate the computer mouse's properties into the internal model of the movement, resulting in imagery with the computer mouse relying on internal models similar to the stylus (simpler tool). In addition to offering a refined approach for measuring manipulation- and timing-related AIA, our study highlights the importance of tool familiarity in mental chronometry.

The covalent linkage of preformed supramolecular entities represents a powerful yet synthetically challenging strategy for constructing sophisticated functional systems, crucial for advances in enzyme mimicry, catalysis, and materials. Addressing the core difficulty of precise, site-selective coupling, we report a direct one-step C-C coupling of carbazole-based macrocycles 2 under Scholl reaction conditions to afford bismacrocycle 1. Interestingly, when using the smaller ring precursor 4 under similar reaction conditions, we obtain the intermolecular dehydrogenative product 3. DFT computations provide insight into the distinct formation pathways of monomacrocycle 3 and bismacrocycle 1. Moreover, we investigated the photophysical properties of compounds 1-4, confirming that all are suitable hosts for C70, with their binding constants determined. Notably, bismacrocycle 1 demonstrates exceptional photocatalytic activity for debrominative borylation, outperforming other macrocycles (2-4) in catalytic efficiency. Finally, we also examined the anticancer activities of these newly synthesized macrocycles. Collectively, this work provides a versatile synthetic strategy for a complex bismacrocycle and establishes its significant dual potential as a selective host, a powerful photocatalyst, and a photocytotoxic anticancer candidate.

White matter lesion (WML) identification, assessment, and characterization using magnetic resonance imaging (MRI) are fundamental for diagnosis and monitoring of multiple sclerosis (MS). Portable ultra-low field (pULF) MRI at 64 millitesla (mT) has been shown to visualize WML with at least one dimension greater than 4 mm. An automated WML segmentation tool catered to pULF-MRI can provide standardized and accurate quantitative measurements of WML volume. In this study, we sought to investigate and compare the accuracy of machine-learning (ML) and deep-learning (DL) pULF MRI segmentation tools. Same-day paired pULF (64mT) and high-field (HF, 3T) MRI scans from 84 adults with MS or suspected-MS (mean age&#xb1;SD: 48&#xb1;13, 62 females) included T2-FLAIR and T1w images. Reference WML segmentations were manually annotated on pULF T2-FLAIR for all scans, with WML confirmed with registered HF T2-FLAIR. HF reference WML segmentations were created. Four automated segmentation methods were applied to pULF scans: Method for Inter-Modal Segmentation Analysis (MIMoSA), an ML algorithm trained on HF WML masks; WMH-SynthSeg, a convolutional neural network model with flexible segmentation capabilities across field strengths and resolution; nnU-Net, a DL algorithm trained on pULF reference WML masks; and Pseudo-Label Assisted nnU-Net (PLAn), a DL algorithm pre-trained on HF reference WML masks and refined with 64mT reference WML masks. Two models were trained with nnU-Net, one using T2-FLAIR images only (nnU-Net-FL) and one using T1w and T2-FLAIR images (nnU-Net-FL/T1). The same was done with PLAn, creating PLAn-FL and PLAn-FL/T1. The six automated WML segmentation outputs were compared to the manual segmentations to determine Dice Similarity Coefficient (DSC) scores. Associations of WML volume estimates with clinical measures were investigated. DSC scores with pULF reference WML masks from PLAn-FL (DSC mean&#xb1;SD: 0.50&#xb1;0.24) outperformed MIMoSA (0.24&#xb1;0.20, p<0.0001), WMH-SynthSeg (0.30&#xb1;0.18, p<0.0001), nnU-Net-FL (0.41&#xb1;0.24, p<0.0001), and nnU-Net-FL/T1 (0.41 &#xb1; 0.26, p = 0.0004). Worse Expanded Disability Status Scale (EDSS) and Scripps Neurologic Rating Scale (SNRS) scores were correlated with higher WML volumes in the pULF and HF reference masks. They were also correlated with WML volumes derived from WHM-SynthSeg, nnU-Net-FL, nnU-Net-FL/T1, PLAn-FL, and PLAn-FL/T1, but not MIMoSA. After adjusting for age, WHM-SynthSeg, nnU-Net FL, nnU-Net-FL/T1, PLAn-FL, and PLAn-FL/T1 had significant associations with EDSS and SNRS scores. nnU-Net and PLAn performed best in segmenting WML on pULF-MRI at 64 mT, providing accurate quantitative estimates of WML burden. Moreover, WML volumes estimated by these algorithms were associated with clinical measures of disability, underscoring their utility for reflecting clinical and radiological disease severity. Given pULF-MRI's mobility and lower cost, these findings highlight its relevance in clinical trials, particularly in involving more participants who face logistical constraints and barriers. Deep learning algorithms were trained on 20 paired, same-day pULF and HF MRIAutomated pULF segmentation tools were evaluated using 84 paired, same day MRIAutomated pULF WML segmentations align with reference pULF and HF segmentationsTrained deep learning algorithms demonstrate strong pULF MRI lesion segmentationAutomated pULF WML estimates are associated with clinical disability scores.

Amyloid oligomers are intermediate aggregates of misfolded proteins that exhibit stronger cellular interactions and enhanced interfacial adhesion capabilities compared with protofibrils and mature fibrils. However, the molecular basis of their adhesion remains insufficiently understood because of their structural complexity and mutability. Here, we present a surface-induced oligomerization method for fabricating an amyloid-like oligomer monolayer (AOM). Owing to its well-defined structure, stability, and accessibility, AOM serves as an ideal model system for elucidating the adhesion mechanisms of amyloid oligomers. AOM demonstrated a 34-fold increase in adhesion strength relative to native lysozyme. Systematic adhesion analysis on chemically defined surfaces combined with molecular simulations uncovered that hydrophobic interactions and hydrogen bonding are the dominant contributors to adhesion, supported by electrostatic and van der Waals forces. Furthermore, AOM can be used as a universal surface modification platform, outperforming conventional techniques in terms of stability and reliability, particularly on inert polymer surfaces. Because of its biocompatibility, the AOM significantly enhances cell adhesion on diverse substrates, with improvements ranging from 1.8 to 32 times compared to blank substrates. Collectively, this work not only provides a strategy for mechanistic insight into oligomer-mediated adhesion but also establishes a robust bioinspired surface modification platform for advanced material applications.

To evaluate a support structure for women victims of violence from the perspective of the survivors themselves. We conducted semi-structured interviews with 17 women attending a Maison des Femmes (MdF), a hospital-based multiprofessional structure providing medical, psychological, social, and legal support to women victims of violence, within a model currently being scaled up nationally in France. When enrolling in the MdF, participants reported seeking safety and support and expecting a therapeutic relationship tailored to their individual needs. They valued access to a wide range of professionals, coordinated by a single trusted interlocutor responsible for their care pathway. Overall, participants reported improved well-being since initiating MdF care. Experiences of group therapy were mixed: some participants described it as fostering supportive social interactions (n&#x2009;=&#x2009;12), whereas others felt outperformed by other group members (n&#x2009;=&#x2009;5), leading to feelings of discouragement. Frustration arose from unmet needs related to housing and income support, the slow pace of legal proceedings, and long waiting times for appointments within the structure. When considering the planned end of care, typically after 6-12&#x2009;months, participants expressed uncertainty and fear of losing a vital source of support. Although women survivors of violence reported improved well-being after joining the MdF, persistent unmet social and legal needs and anxiety surrounding the end of care were identified. These findings highlight the importance of early information and anticipatory planning regarding care duration and post-discharge support.

Magnetic Resonance Fingerprinting (MRF) enables rapid quantitative imaging, but high-resolution 3D reconstructions remain computationally expensive due to the NUFFTs required at every iteration, and the commonly used Locally Low Rank (LLR) regularization becomes ineffective at high acceleration. Learned 3D priors could address these limitations, but training them at scale is challenging due to memory and runtime constraints. This work proposes SPUR-iG, a fully 3D deep unrolled subspace reconstruction framework that provides fast, high-quality reconstruction for high-resolution non-Cartesian 3D MRF, while keeping training time computationally tractable. SPUR-iG leverages implicit GROG-based data consistency (DC), which grids non-Cartesian k-space using a learned family of kernels, enabling efficient FFT-based DC with minimal artifacts. To make 3D unrolled training more efficient, we introduce a staged training strategy that keeps computation tractable while progressively improving reconstruction quality. We evaluate the method on a large in vivo dataset, as well as on cross-vendor out-of-distribution data. At 1&#x2009;mm isotropic resolution, SPUR-iG outperforms LLR and a state-of-the-art hybrid 2D-3D unrolled baseline in subspace coefficient quality and T 1 / T 2 accuracy. Whole-brain reconstructions complete in under 15&#x2009;s, providing up to a 111 &#xd7; speedup for 2-min scans relative to LLR. Notably, SPUR-iG reconstructions from 30-s acquisitions achieve mean T 1 accuracy that matches or exceeds the mean accuracy of LLR reconstructions from 2-min acquisitions. SPUR-iG introduces a fully 3D unrolled reconstruction framework for MRF that improves both reconstruction speed and accuracy, making high-resolution accelerated 3D MRF more practical for research and clinical use.

To evaluate genetic testing practices (exome sequencing, commercial panel, and in-house genetic panels) from a large tertiary hospital for determining gaps, and to identify clinical associations with pathogenic genetic variants. This retrospective cohort study included patients (age&#xa0;<&#x2009;18&#x2009;years) from a neurology department for whom genetic testing was requested for neurological disorders, epilepsy, and movement disorders (2020-2023). Logistic regression was used to identify clinical features predictive of positive results. The clinical benefits of genetic testing were studied. Three hundred and ninety patients underwent genetic testing by exome sequencing (n&#x2009;=&#xa0;125), commercial panel (n&#x2009;=&#xa0;143), in-house epilepsy (n&#x2009;=&#xa0;78), and movement disorder (n&#x2009;=&#xa0;44) gene panels. Exome sequencing had the highest pathogenic yield (n&#x2009;=&#xa0;49, 39%), followed by epilepsy (n&#x2009;=&#xa0;22, 28%), movement disorder (n&#x2009;=&#xa0;11, 25%), and commercial (n&#x2009;=&#xa0;22, 15%) panels. Variants of uncertain significance were highest in commercial (64%) and epilepsy (34%) panels. Among the exome sequencing cohort, the predominant clinical features were developmental delay (89%), intellectual disability (51%), and epilepsy (35%). Pathogenic variants in the exome sequencing cohort were more likely in patients with severe developmental delay (33%, p&#x2009;=&#xa0;0.03) and hypotonia (39%, p&#x2009;=&#xa0;0.05). There was significant utility of genetic testing in informing clinical decision making (49% of pathogenic variants in exome sequencing cohort). Exome sequencing outperforms gene panels in confirming genetic diagnoses in paediatric neurological disorders, and highlights the need for building local diagnostic genetic-testing resources.

The purpose of this study is to ascertain the incidence of fatigue and to determine the cross-sectional associate and related variables of fatigue in patients with Parkinson's disease (PD). A sample of 130 healthy individuals (39.23% male, mean age 64.19 &#xb1; 8.70&#x202f;years) and 130 patients with PD (50.00% males, mean age 65.06&#x202f;&#xb1;&#x202f;9.04) were enrolled in the study. All participants were evaluated using the Fatigue Severity Scale (FSS), Mini-Mental State Examination (MMSE), Hamilton anxiety scale (HAMA), and Hamilton depression scale (HAMD). Patients with PD were assessed with the Unified Parkinson's Disease Rating Scale (UPDRS), Hoehn & Yahr staging, the MMSE, HAMA, HAMD, the Parkinson's Disease Sleep scale-2 (PDSS-2), and Parkinson's Disease Questionnaire-39 (PDQ-39). The PD group exhibited significantly higher scores in the FSS (p&#x202f;<&#x202f;0.001) and demonstrated a greater susceptibility to fatigue compared to the control group. The FSS score in female patients was significantly higher than that in male patients (p&#x202f;=&#x202f;0.003). Patients experiencing fatigue exhibited significantly higher scores on the UPDRS (p&#x202f;<&#x202f;0.001), Hoehn & Yahr staging (p&#x202f;<&#x202f;0.001), NMSS (p&#x202f;<&#x202f;0.001), HAMD (p&#x202f;<&#x202f;0.001), HAMA (p&#x202f;<&#x202f;0.001), PDSS-2 (p&#x202f;=&#x202f;0.012), and PDQ-39 (p&#x202f;<&#x202f;0.001) compared to those without fatigue. Correlation analysis revealed significant associations between fatigue and NMSS, HAMD, HAMA, PDSS, PDQ-39, UPDRS, and Hoehn & Yahr stage. In the FSS-based multiple linear regression, HAMA and PDQ-39 remained cross-sectional associated with higher fatigue severity, whereas NMSS showed a borderline effect. To directly evaluate multivariable discrimination for fatigue occurrence, we constructed a combined logistic model including NMSS, HAMD, and PDQ-39. The combined model showed good discrimination (AUC 0.879, 95% CI 0.806-0.941) and acceptable calibration (Brier score 0.125), outperforming any single scale. Patients with Parkinson's disease experience a higher incidence of distressing fatigue compared to healthy individuals. The significant prevalence of fatigue among these patients is linked to various motor and non-motor symptoms.

The low-dimensional nature of conventional polyoxometalate-based materials imposes intrinsic constraints on charge transport, limiting their photoelectrochemical performance. Herein, we report a dimensionality-controlled molecular engineering strategy that integrates phenylphosphonate-functionalized Fe{(PhP)4Mo6}2 clusters with the rigid conjugated bib ligand to construct two hourglass-type phosphomolybdate-based structures: one 3D covalent framework (compound 1) and one 0D supramolecular structure (compound 2), enabling specific photoelectrochemical (PEC) responses toward Cr(vi) reduction and levofloxacin (LVF) oxidation. Benefiting from enhanced carrier separation and charge transfer, compound 1 achieves detection limits of 0.11 nM for Cr(vi) and 0.16 nM for LVF, with sensitivities of 555.58 and 362.73 &#xb5;A &#xb5;M-1, outperforming most polyoxometalate-based sensors and rivaling noble-metal platforms. Compound 1 also showed excellent anti-interference and reliable performance in real water and milk samples. This work offers a new molecular engineering strategy for high-performance sensing materials of trace environmental pollutants.

Hungry bone syndrome (HBS) is a frequent and clinically significant complication following parathyroidectomy (PTX) in patients with secondary hyperparathyroidism (SHPT), often leading to prolonged hypocalcaemia and increased healthcare burden. Existing prediction models are limited by small sample sizes and inability to capture complex clinical interactions. This study aimed to develop and validate a clinically aligned, two-stage machine learning (ML) framework to predict HBS after PTX. A retrospective cohort of patients undergoing PTX for SHPT between 2008 and 2025 at a tertiary centre was analysed. A two-stage ML framework was constructed: stage 1 used preoperative variables to generate a risk score, and stage 2 integrated this score with intraoperative features. Multiple ML models were evaluated using area under the receiver operating characteristic curve (AUROC), calibration metrics and resampling techniques. A total of 882 patients were included, with an HBS incidence of 69.9%. EasyEnsemble and logistic regression demonstrated the highest discrimination (AUROC 0.712), outperforming the k-nearest neighbours baseline. EasyEnsemble achieved the best overall performance (accuracy 0.707, F1 score 0.666) and calibration (Brier score 0.186). Key predictors included elevated preoperative alkaline phosphatase, higher intact parathyroid hormone levels and lower serum calcium. This two-stage ML framework demonstrated acceptable predictive performance and aligns with clinical decision-making processes. It enables early identification of high-risk patients and may support individualised perioperative management to mitigate HBS and its complications.