Tag: pubmed

Proc Natl Acad Sci U S A. 2025 Dec 23;122(51):e2523012122. doi: 10.1073/pnas.2523012122. Epub 2025 Dec 18.

ABSTRACT

Like human decision-making under real-world constraints, artificial neural networks may balance free exploration in parameter space with task-relevant adaptation. In this study, we identify consistent signatures of criticality during neural network training and provide theoretical evidence that such scaling behavior arises naturally from information-driven self-organization: a dynamic balance between the maximum entropy principle that promotes unbiased exploration and mutual information constraint that relates updates with task objective. We numerically demonstrate that the power-law exponent of updates remains stable throughout training, supporting the presence of self-organized criticality. Furthermore, we show that the loss landscape exhibits exponential ruggedness under small perturbations, transitioning to power-law ruggedness at larger scales, in the absence of mini-batch noise, indicating an intrinsic geometric landscape. We also observe a power-law distribution in the intervals between large updates, indicating an intermittent learning process. Together, these findings suggest that neural network learning reflects a nonequilibrium process governed by the fundamental trade-off between randomness and relevance, highlighting its dynamic nature and offering insights into the interpretability of AI systems.

PMID:41410766 | DOI:10.1073/pnas.2523012122

Eur J Appl Physiol. 2025 Dec 18. doi: 10.1007/s00421-025-06095-4. Online ahead of print.

ABSTRACT

PURPOSE: Fatigue can exacerbate gait abnormalities in unilateral transtibial amputees (TTA), but specific changes in walking patterns and compensatory strategies under fatigue remain unclear. This study aimed to investigate gait alterations in unilateral TTA following a fatigue-inducing protocol.

METHODS: Ten male unilateral TTA and ten age-matched able-bodied controls underwent three-dimensional gait analysis at self-selected speed under non-fatigued and fatigued conditions. Spatiotemporal parameters, joint kinematics, and ground reaction forces were measured. Two-way ANOVA was used for statistical comparisons.

RESULTS: Within the TTA group, the residual limb demonstrated a shorter stance phase (62.3 ± 2.4% vs. 65.1 ± 1.4%) and single support (34.7 ± 1.9% vs. 37.1 ± 2.8%), but greater step length (38.27 ± 2.49% vs. 35.76 ± 1.92%), peak hip flexion(35.0 ± 5.6° vs. 28.6 ± 3.0°) and knee flexion(67.3 ± 7.2° vs. 56.4 ± 5.6°) than the intact limb. Fatigue further increased step length (37.42 ± 2.41% vs. 36.61 ± 2.63%) and hip flexion (33.2 ± 5.3° vs. 30.5 ± 5.3°), while reducing hip extension (10.9 ± 4.8° vs. 12.8 ± 4.2°) and hip horizontal range of motion (17.5 ± 6.0° vs. 19.5 ± 6.6°). Compared to controls, TTA group had longer stride time (1.15 ± 0.04s vs. 1.08 ± 0.08s) and greater hip horizontal range of motion (18.5 ± 6.4° vs. 12.5 ± 3.0°), but lower cadence (105.1 ± 4.1 steps/min vs. 111.6 ± 8.1 steps/min).

CONCLUSION: Fatigue amplifies pre-existing gait asymmetries in unilateral TTA and elicits compensatory strategies, including increased reliance on the intact limb and greater proximal joint mobility. Targeted interventions to enhance residual limb function and hip flexor strength and endurance may help reduce fatigue-related asymmetry and fall risk.

PMID:41410758 | DOI:10.1007/s00421-025-06095-4

Eur J Trauma Emerg Surg. 2025 Dec 18;51(1):363. doi: 10.1007/s00068-025-03042-y.

ABSTRACT

PURPOSE: Post-traumatic pulmonary embolism (PE) may develop directly in the lungs, termed “de novo” pulmonary embolism (DNPE). Severe chest trauma has been identified as a potential risk factor for DNPE due to localized inflammation, occult vascular injury, and low-flow states (venous stasis). The primary outcome was to examine the association between DNPE and chest trauma, while the secondary outcome was to characterize patients in the DNPE group.

METHODS: We conducted a retrospective cohort study of patients with trauma aged ≥ 15 years admitted to Songklanagarind Hospital, a level 1 trauma center, from 2013 to 2023. All patients diagnosed with post-traumatic PE were reviewed for clinical parameters, imaging findings, and treatments. Patients without ultrasonographic evaluation for DVT were excluded.

RESULTS: Among 43,908 patients with trauma, PE was diagnosed in 78 (0.18%). After excluding four patients without DVT assessment, 74 patients remained. Of these, 49 (66%) were diagnosed with DNPE and 25 (34%) with PE + DVT. Compared with patients with PE + DVT (32%), patients with DNPE (38.8%) showed no significant difference in the incidence of chest trauma (p = 0.567). The location of PE significantly differed (p = 0.005) between the groups, with DNPE showing more peripheral involvement (79.6%) and PE + DVT showing more central emboli (52%). No patient in the DNPE group underwent pulmonary thromboembolectomy.

CONCLUSION: DNPE is more common among patients with trauma, but its association with chest trauma was not statistically significant. DNPE may result from undetected pelvic DVT or other mechanisms requiring further investigation.

PMID:41410754 | DOI:10.1007/s00068-025-03042-y

Pak J Biol Sci. 2025 Dec;28(9):576-586. doi: 10.3923/pjbs.2025.576.586.

ABSTRACT

<b>Background and Objective:</b> Carvacrol, a naturally occurring phenolic monoterpenoid compound found in various essential oils, exhibits outstanding pharmacological characteristics, essentially anticancer, antiproliferative and pro-apoptotic effects. This study evaluates the therapeutic potential of carvacrol as a topical agent in preventing and treating DMBA-induced oral cancer in rats through <i>in vivo</i> assessment of its anticancer efficacy and histopathological effects. <b>Materials and Methods:</b> This study was designed to investigate the chemopreventive and therapeutic potential of topically applied carvacrol in a rat model of Oral Squamous Cell Carcinoma (OSCC) induced by 7,12-dimethylbenz[a]anthracene (DMBA). The expression of Proliferating Cell Nuclear Antigen (PCNA) and B-cell lymphoma/leukemia-2 (Bcl-2), markers of proliferation and apoptosis, respectively, was estimated by immunohistochemical staining. Data were analyzed using one-way ANOVA with Tukey’s <i>post hoc</i> test and Pearson’s correlation, following normality confirmation (p>0.05), with results summarized using descriptive statistics. <b>Results:</b> Noteworthy declines in both PCNA and Bcl-2 expression were noticed in groups treated with carvacrol, either concurrently with DMBA or following its application. The group receiving carvacrol alternately with DMBA showed the most noticeable suppression in both markers. <b>Conclusion:</b> These results highlight carvacrol’s dual character in quashing carcinogenesis and stimulating apoptotic cell death, supporting its potential as a safe, natural therapeutic agent for OSCC intervention.

PMID:41410137 | DOI:10.3923/pjbs.2025.576.586

Biotechnol Bioeng. 2025 Dec 18. doi: 10.1002/bit.70128. Online ahead of print.

ABSTRACT

Extracellular vesicles (EVs), including exosomes, are abundant in bovine milk and Lactobacillus culture media but difficult to isolate with high efficiency and purity. In response, a micro-electro-mechanical systems (MEMS)-based membrane filter was developed to address these limitations. Under equivalent conditions, the developed filter outperformed commercial filters, achieving a 2.17-fold higher EV recovery rate compared to the commercial polyethersulfone (PES) membrane from a 5 mL high-concentration sample, and yielding a total of 50 mL of EV solution at a concentration of 5.52 × 10¹⁰ particles/mL. The membrane was engineered to achieve a minimum pore size of 32 nm and a minimum thickness of 290 nm through separate fabrication processes. Among these, the MEMS160 membrane, which features uniformly distributed 168 nm pores on a 318 nm thick structure, demonstrated enhanced performance by effectively reducing fouling, as confirmed by blocking-model assessments. Biological evaluations showed that EVs isolated using the developed filter retained notable purity and bioactivity. Specifically, milk-derived EVs increased the proliferation of human fibroblasts (Hs68) and human follicle dermal papilla cells (HFDPCs) by up to 25% and 50%, respectively, while Lactobacillus-derived EVs increased proliferation by up to 11% and 53% at certain concentrations. Furthermore, co-treatment with an anti-aging peptide (AIMP1-derived peptide) had a synergistic effect on both cell types. Similar trends were seen in canine and feline fibroblasts. Milk-derived EVs boosted proliferation by up to 25% in canine and 31% in feline cells, while Lactobacillus-derived EVs increased it by up to 46% and 34%, respectively. These effects reached statistical significance. These results show the filter’s potential for large-scale EV isolation and dermatological applications, requiring high purity and yield.

PMID:41410131 | DOI:10.1002/bit.70128

Biom J. 2025 Dec;67(6):e70104. doi: 10.1002/bimj.70104.

ABSTRACT

In the competing risks setting, the $t$ -year absolute risk for a specific time $t$ (e.g., 2 years), also called the cumulative incidence function at time $t$ , is often interesting to estimate. It is routinely estimated using the nonparametric Aalen-Johansen estimator. This estimator handles right-censored data and has desirable large sample properties, as it is the nonparametric maximum likelihood estimator (NPMLE). Inference for comparing absolute risks, via either a risk difference or a risk ratio, can therefore be done via usual asymptotic normal approximations and the delta method. However, the small sample performances of this approach are not fully satisfactory. Especially, (i) coverage of confidence intervals may be inaccurate and (ii) comparisons made using a risk ratio and a risk difference can lead to inconsistent conclusions, in terms of statistical significance. We, therefore, introduce an alternative empirical likelihood approach. One advantage of this approach is that it always leads to consistent conclusions when comparing absolute risks via a risk ratio and a risk difference, in terms of significance. Simulation results also suggest that small sample inference using this approach can be more accurate. We present the computation of confidence intervals and p-values using this approach and the asymptotic properties that justify them. We provide formulas and algorithms to compute constrained NPMLE, from which empirical likelihood ratios and inference procedures are derived. The novel approach has been implemented in the timeEL package for R, and some of its advantages are demonstrated via reproducible analyses of bone marrow transplant data.

PMID:41410124 | DOI:10.1002/bimj.70104

J Int Neuropsychol Soc. 2025 Dec 18:1-6. doi: 10.1017/S1355617725101732. Online ahead of print.

ABSTRACT

OBJECTIVE: This study aimed to explore the correlates of zero, one, and multiple performance validity test (PVT) failures on cognitive test performance in patients with various degrees of severity of traumatic brain injury.

METHOD: 306 participants completed the Trail Making Test as part of a neuropsychological evaluation within 1-36 months post-injury. They were assigned to zero, one, or ≥ two fail groups on the basis of at least two independent PVTs. Group differences in Trail Making Test performance were analyzed with analysis of variance, with post hoc contrasts with the Bonferroni correction for multiple comparisons. Groups were also compared on various background characteristics.

RESULTS: Participants who passed all PVTs had statistically significantly better performance on both parts of the Trail Making Test as compared to those who failed either one or multiple PVTs, with the latter two groups not differing statistically significantly from each other. PVT failure was relatively more common in participants who were female, had an uncomplicated mild TBI, were involved in financial compensation-seeking, and were seen at a longer time point since injury.

CONCLUSION: Failure of even only one PVT is associated with lower neuropsychological test performance in patients with traumatic brain injury, especially when empirically validated criteria are used that are stratified by injury severity. Such failure does not always reflect malingering but must be interpreted and addressed in the context of patient background characteristics.

PMID:41410118 | DOI:10.1017/S1355617725101732

Biom J. 2025 Dec;67(6):e70102. doi: 10.1002/bimj.70102.

ABSTRACT

Functional data analysis has received significant attention due to its frequent occurrence in modern applications, such as in the medical field, where electrocardiograms or electroencephalograms can be used for a better understanding of various medical conditions. Due to the infinite-dimensional nature of functional elements, the current work focuses on dimension reduction techniques. This study shifts its focus to modeling the conditional quantiles of functional data, noting that existing works are limited to quantitative predictors. Consequently, we introduce the first approach to partial dimension reduction for the conditional quantiles under the presence of both functional and categorical predictors. We present the proposed algorithm and derive the convergence rates of the estimators. Moreover, we demonstrate the finite sample performance of the method using simulation examples and a real dataset based on functional magnetic resonance imaging.

PMID:41410116 | DOI:10.1002/bimj.70102

Psychol Med. 2025 Dec 18;55:e384. doi: 10.1017/S0033291725102651.

ABSTRACT

BACKGROUND: Early life adversity (ELA) is common and cross-sectionally associated with brain gray matter structure, including cortical thickness, cortical surface area, and subcortical volumes in childhood. However, to which degree ELA influences the trajectory of gray matter macrostructural and microstructural development during childhood and adolescence remains largely unexplored.

METHODS: We included 6414 participants from the Adolescent Brain Cognitive Development study at ages 9-11, where 1923 were followed to ages 11-13. We used linear mixed-effects models to test for associations between MRI-derived longitudinal measures of gray matter macro- (cortical thickness, surface area, subcortical volume) or microstructure (T1w/T2w ratio) and trauma exposure, parental acceptance, household abuse, and being resilient or susceptible to trauma in terms of developing an internalizing disorder.

RESULTS: At ages 9-11, higher levels of parental acceptance, trauma exposure, and being trauma resilient were associated with lower levels of cortical thickness. In contrast, being trauma susceptible was negatively related to hippocampal volume and cortical surface area. Longitudinally, more parental acceptance at baseline was associated with more cortical thinning between ages 9-11 and 11-13, while more household abuse was associated with less change in T1w/T2w ratio over time.

CONCLUSIONS: Parental acceptance and trauma resilience are linked to accelerated pace of apparent cortical thinning in youth aged 9-13 years, while household abuse is associated with slower microstructural development, as reflected by smaller longitudinal changes in the T1w/T2w ratio. Threat and deprivation may be distinctly associated with gray matter developmental trajectories in late childhood.

PMID:41410108 | DOI:10.1017/S0033291725102651

J Adv Nurs. 2025 Dec 17. doi: 10.1111/jan.70456. Online ahead of print.

ABSTRACT

BACKGROUND: Delirium is a common complication following cardiac surgery and significantly affects patient prognosis and quality of life. Recently, the application of artificial intelligence (AI) has gained prominence in predicting and assessing the risk of postoperative delirium, showing considerable potential in clinical settings.

OBJECTIVE: This scoping review summarises existing research on AI-based prediction models for post-cardiac surgery delirium and provides insights and recommendations for clinical practice and future research.

METHODS: Following the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines, eight databases were searched: China National Knowledge Infrastructure, Wanfang Database, China Biomedical Literature Database, Virtual Information Platform, PubMed, Web of Science, Medline, and Embase. Studies meeting the inclusion criteria were screened, and data were extracted on surgery type, delirium assessment tools, predictive factors, and AI-based prediction models. The search covered database inception through January 12, 2025. Two researchers independently conducted the literature review and data analysis.

RESULTS: Ten studies from China, Canada, and Germany involving 11,702 participants were included. The reported incidence of postoperative delirium ranged from 5.56% to 34%. The most commonly used assessment tools were Confusion Assessment Method for the Intensive Care Unit, Diagnostic and Statistical Manual of Mental Disorders-5, and Intensive Care Delirium Screening Checklist. Key predictive factors included age, cardiopulmonary bypass time, cerebrovascular disease, and pain scores. AI-based prediction models were primarily developed using R (6/10, 60%) and Python (4/10, 40%). Model performance, as measured by the area under the curve, ranged from 0.544 to 0.92. Among these models, Random Forest (RF) was the most effective (5/10, 50%), followed by XGBoost (3/10, 30%) and Artificial Neural Networks (2/10, 20%).

CONCLUSION: AI-based models show promise for predicting postoperative delirium in cardiac surgery patients. Future studies should prioritise integrating these models into clinical workflows, conducting rigorous multicenter external validation, and incorporating dynamic, time-varying perioperative variables to enhance generalizability and clinical utility.

REPORTING METHOD: This review followed the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines.

PATIENT OR PUBLIC CONTRIBUTION: This study did not include patient or public involvement in its design, conduct, or reporting.

PMID:41410092 | DOI:10.1111/jan.70456