Tag: pubmed

Biomed Eng Online. 2026 Feb 7. doi: 10.1186/s12938-026-01528-3. Online ahead of print.

ABSTRACT

BACKGROUND: Machine learning (ML) applied to diffusion tensor imaging (DTI) has emerged as a promising tool for detecting microstructural brain alterations in movement disorders. However, existing studies vary widely in design, sample size, imaging pipelines, and analytic rigor, resulting in high methodological heterogeneity that limits quantitative comparability.

OBJECTIVES: This exploratory meta-analysis and narrative synthesis aimed to characterize performance trends, methodological diversity, and sources of variability among ML models trained on DTI data for classifying movement disorders, rather than to infer a single pooled diagnostic effect. This was designated exploratory because extreme heterogeneity prevented confirmatory pooled effect inference, so the analysis focused on describing performance distributions and methodological patterns rather than estimating a unified diagnostic effect.

METHODS: A systematic search of PubMed, Web of Science, and Scopus identified human studies applying ML algorithms to DTI for diagnostic or classification purposes. Accuracy, sensitivity, specificity, and the area under the curve (AUC) were extracted, with multiple imputation used for incomplete metrics with missingness rates below 40%. Random-effects modeling was employed to provide descriptive summaries, and subgroup analyses were conducted to explore trends across disorders, model architectures, and imaging modalities. Study qualities were assessed with JBI tools.

RESULTS: Forty-six studies (2016-2024) were included, spanning Parkinson’s disease, Tourette syndrome, and essential tremor. Reported performance was generally high (median AUC ≈ 0.91), but between-study heterogeneity was extreme (I² = 94.7%), indicating that studies were estimating distinct effects. Disorder-specific subgroup AUCs varied markedly: Essential Tremor (0.95), Parkinson’s (0.90), Tourette’s (0.88), and Other (0.79). Deep learning and radiomics-based models have reported higher accuracies, but they were often trained on small, single-center cohorts (37-139 participants), which limits their external validity. Pooled statistics were presented descriptively to illustrate performance ranges despite high heterogeneity, and were not interpreted as confirmatory effect sizes.

CONCLUSIONS: ML models using DTI demonstrate high internal performance across studies, although generalizability remains limited across multiple movement disorders; however, current evidence remains exploratory due to small sample sizes, methodological fragmentation, and a lack of standardized imaging pipelines. Rather than confirmatory inference, these findings provide a descriptive map of emerging trends in ML-DTI diagnostics. Future progress will depend on data harmonization initiatives, multicenter collaborations, and federated learning frameworks that can support reproducible, generalizable, and clinically interpretable models.

PMID:41654900 | DOI:10.1186/s12938-026-01528-3

Harm Reduct J. 2026 Feb 7. doi: 10.1186/s12954-026-01411-3. Online ahead of print.

ABSTRACT

INTRODUCTION: In response to the overdose epidemic, novel strategies including Overdose Response Hotlines and Applications (ORHAs) have been introduced to help mitigate the crisis. These technologies enable individuals with a phone to access harm reduction support via smartphones and applications. Such supports include overdose monitoring, access to social services, mental health referrals, and more. This study analyzed data from the Canadian National Questionnaire on Overdose Monitoring (CNQOM), a large bilingual national survey, to evaluate the perspectives of people who use unregulated substances currently (PWUS-C), people who used unregulated substances previously (PWUS-P), and addiction service provider (ASP) on the importance of specific ORHA features.

METHODS: One component of the CNQOM pertained to the importance of specific ORHA service elements. Examined categories included accessibility and technological features, overdose response functionality, data privacy and philosophies of care, additional support services, and substance usage. Each group responded to 33 questions on a 5-point Likert scale, and the data was analyzed using descriptive statistics involving percentages and ordinal logistical regression analysis.

RESULTS: The study involved 971 participants: 840 PWUS-C, 298 PWUS-P, and 169 ASP. The majority of respondents from the key groups considered all ORHA elements important. Generally, the groups ranked the elements in a similar order of importance, with only minor variations. The highest-ranked elements in each element category with regards to importance were: 24/7 availability (84% of PWUS-C, 88% of PWUS-P, and 90% of ASP), the ability of EMS to resuscitate individuals during an overdose (81% of PWUS-C, 83% of PWUS-P, 85% ASP), non-judgmental support (87% of PWUS-C, 87% of PWUS-P, and 91% of ASP), access to mental health support (82% of PWUS-C, 84% of PWUS-P, and 90% of ASP), and feeling safer when using substances (80% of PWUS-C, 81% of PWUS-P, and 88% of ASP).

CONCLUSION: This paper highlights the importance multiple groups place on various elements of ORHAs, reflecting critical elements that should be considered when standardizing these virtual harm reduction technologies. The results of this study provide insight into opportunities to enhance virtual platforms, making them more responsive, accessible, and trusted as harm reduction resources.

PMID:41654899 | DOI:10.1186/s12954-026-01411-3

BMC Psychol. 2026 Feb 7. doi: 10.1186/s40359-026-04066-8. Online ahead of print.

ABSTRACT

BACKGROUND: To investigate the longitudinal relationship between informal social participation and life satisfaction among older adults relocated for poverty alleviation in China, as well as the mediating role of perceived stress and sleep duration.

METHODS: Overall, 1345 participants [mean age 71.52 (SD:7.19) years; 48.4% female] were included in the longitudinal study. The participants were surveyed using perceived stress scale-14 (PSS-14) and satisfaction with life scale (SWLS). AMOS Statistics 26 was used to test for common method bias (CMB). SPSS Statistics 26 was used to conduct descriptive statistics, and correlation analysis. Besides, four longitudinal cross-lagged models and bootstrap methods were employed to investigate whether there is a mutual influence among informal social participation, perceived stress/sleep duration, and life satisfaction by AMOS Statistics 26.

RESULTS: This study did not have a severe problem of CMB. The results indicated informal social participation predicted perceived stress and sleep duration 6 months later; perceived stress predicted life satisfaction 6 months later; and informal social participation at T1 predicted life satisfaction at T3 through perceived stress at T2. However, informal social participation at T1 did not predict life satisfaction at T3 through sleep duration at T2.

DISCUSSION: These results indicate that for relocated older adults, informal social participation enhances life satisfaction not by improving sleep duration, but primarily through reducing perceived stress. The key pathway is that social participation lowers stress levels, which in turn leads to greater long-term life satisfaction. Hence, our findings could serve to prompt the administrators of community to be aware of the significance of stress alleviation and regard it as a key intervention target in programs designed to enhance the well-being of relocated older adults.

PMID:41654879 | DOI:10.1186/s40359-026-04066-8

J Intensive Care. 2026 Feb 7. doi: 10.1186/s40560-026-00866-7. Online ahead of print.

ABSTRACT

BACKGROUND: Tracheal intubation in critically ill adults is frequently complicated by severe physiological adverse events, particularly cardiovascular instability. Although fentanyl is commonly used for induction, observational data suggest that its use may increase the risk of post-intubation hypotension. However, the overall randomized evidence remains unclear. In this systematic review and meta-analysis of randomized controlled trials (RCTs), we hypothesized that induction regimens including fentanyl or its analogs would increase the risk of peri-intubation cardiovascular instability in critically ill patients.

METHODS: We comprehensively searched PubMed, Embase, the Cochrane Library, ClinicalTrials.gov, and the WHO ICTRP from inception through October 31, 2025. Eligible studies were RCTs comparing an induction regimen including fentanyl or its analogs with one without them in critically ill adults undergoing tracheal intubation. The primary outcome was peri-intubation cardiovascular instability. Secondary outcomes included peri-intubation hypoxemia, successful intubation on the first attempt, duration of mechanical ventilation, ICU length of stay, and mortality. Random-effects meta-analyses were performed for all outcomes. Trial sequential analysis (TSA) was conducted for the primary outcome. Certainty of evidence was assessed using the GRADE approach.

RESULTS: We included five RCTs and 515 participants. Two studies were judged to be low risk of bias, two raised some concerns, and one was at high risk of bias. Comparators included various induction agents and placebo. Definitions of peri-intubation cardiovascular instability also varied. The evidence was very uncertain regarding the effect of fentanyl on the risk of peri-intubation cardiovascular instability (risk ratio, 1.41; 95% confidence interval, 0.83-2.40; risk difference, 9.2% more; 95% confidence interval, 3.8% fewer to 31.3% more; very low certainty). In TSA, the required information size (n = 5586) was not reached, indicating the lack of statistical power. The certainty of evidence for pooled secondary outcomes was generally low or very low.

CONCLUSIONS: The effect of fentanyl on peri-intubation cardiovascular instability remains highly uncertain, with pooled estimates compatible with substantial harm, substantial benefit, or no effect. Current randomized evidence is insufficient to guide routine clinical practice, given their very low or low certainty and susceptibility to random error.

TRIAL REGISTRATION: PROSPERO (registration number: CRD420251241214).

PMID:41654873 | DOI:10.1186/s40560-026-00866-7

Cardiovasc Diabetol. 2026 Feb 8. doi: 10.1186/s12933-026-03096-1. Online ahead of print.

ABSTRACT

BACKGROUND: The estimated glucose disposal rate (eGDR), an established measure of peripheral insulin sensitivity, contributes to stratifying the risk of cardio-cerebrovascular events. Nevertheless, the association between long-term eGDR exposure and stroke incidence throughout all stages (0-4) of cardiovascular-kidney-metabolic (CKM) syndrome remains unknown.

METHODS: A cohort of 5248 individuals was drawn from the China Health and Retirement Longitudinal Study (CHARLS). For each participant, eGDR values for the years 2012 and 2015 were calculated using the equation: 21.158 – [0.090 × WC (cm)] – [3.407 × HTN (presence = 1)] – [0.551 × HbA1c (%)]. Cumulative eGDR was calculated as (eGDR₂₀₁₂ + eGDR₂₀₁₅)/2* time (2015-2012). K-means clustering was used to analyse eGDR values from both 2012 and 2015 to identify distinct change patterns. To assess associations with stroke risk, we utilised multivariable logistic regression and restricted cubic spline models.

RESULTS: During the 2015-2018 follow-up period, a total of 336 incident stroke cases were documented. Four distinct eGDR change patterns were identified. In fully adjusted models, compared with the participants in the persistent low pattern (Class 2), those in the moderate-high stable (OR 0.43, 95% CI: 0.31-0.58), stable high (OR 0.29, 0.19-0.43), and rapid decrease (OR 0.66, 0.47-0.91) patterns exhibited significantly lower stroke risk. Furthermore, each 1-unit increase in cumulative eGDR was associated with a 5% reduction in stroke odds (OR 0.95, 0.93-0.96). Restricted cubic spline analysis confirmed a linear inverse relationship between cumulative eGDR and stroke risk (P < 0.001; P for nonlinearity = 0.259).

CONCLUSION: Cumulative eGDR is inversely associated with stroke risk across all CKM syndrome stages (0-4). This observation suggests that prolonged eGDR surveillance may be associated with improved risk stratification in this population.

PMID:41654871 | DOI:10.1186/s12933-026-03096-1

J Ovarian Res. 2026 Feb 7. doi: 10.1186/s13048-025-01901-1. Online ahead of print.

ABSTRACT

BACKGROUND: Germline mutations of BRCA1 and BRCA2 may impair DNA repair in the ovarian cortex, leading to increased oocyte apoptosis, thus, affecting ovarian reserve. Aim of this study was to assess follicular density in ovarian biopsies from women with breast cancer carrying BRCA1 and BRCA2 mutations who underwent ovarian tissue cryopreservation (OTC) at our center.

METHODS: This was a single center, observational, cross-sectional study carried out in a tertiary level referral center for fertility preservation treatment. Exclusion criteria were: patients aged < 18 years or > 38 years, patients who had already undergone chemotherapy/pelvic radiotherapy at the time of OTC, patients without data on follicular density and those with unknown BRCA mutational status. Follicular density was defined as the number of primordial, intermediate primordial, small primary, large primary, secondary, preantral, and antral follicles per 1 mm² of cortical section area.

RESULTS: Out of 216 patients, 21 women reported germline mutation: 9 (4.2%) were carriers of the BRCA1 mutation and 13 (6%) of the BRCA2 mutation. The mean age at OTC was 31.5 ± 3.6 years, and the median age was 32.4 years (range, 21-38). No significant difference in follicular density was observed among women without BRCA mutations, those with BRCA1 mutations, and those with BRCA2 mutations. The median follicular density was 4.0/mm² (range 0-74.5) in BRCA-negative women, 3.5/mm² (range 0-20) in women with BRCA1 mutations, and 4.0/mm² (range 0-32) in women with BRCA2 mutations (p = 0.272 and p = 0.703, respectively). After adjusting for age, no statistically significant differences in follicular density were observed according to BRCA1 and BRCA2 mutation status: the median follicular density was 4.6/mm² in BRCA-negative women, 3.1/mm² in women with BRCA1 mutations, and 3.6/mm² in women with BRCA2 mutations (p = 0.428 and p = 0.385, respectively).

CONCLUSIONS: No significant difference in follicular density was observed between women with BRCA1/BRCA2 mutations and those without. Our findings suggest that the presence of a BRCA mutation does not have a significant negative clinical impact on the follicular population of the ovarian cortex. Larger studies are needed to further validate these findings.

PMID:41654853 | DOI:10.1186/s13048-025-01901-1

BMC Complement Med Ther. 2026 Feb 7. doi: 10.1186/s12906-026-05289-3. Online ahead of print.

NO ABSTRACT

PMID:41654841 | DOI:10.1186/s12906-026-05289-3

BMC Public Health. 2026 Feb 7. doi: 10.1186/s12889-026-26526-x. Online ahead of print.

ABSTRACT

BACKGROUND: Academic performance is often highly prioritized among college students, sometimes at the expense of their health. Despite growing interest in this relationship, limited research with college students has explored how diet quality (DQ) varies by gender, first-generation status, and grade-point average (GPA). The purpose of this paper was to: (1) examine the relationship between DQ and academic performance in college students and (2) identify potential differences based on gender, first-generation status, and varying GPAs.

METHODS: In this cross-sectional study, undergraduate students (n = 301), mean age 21.2 (SD ± 2.49), completed the validated Short Healthy Eating Index (sHEI) based on the USDA’s Healthy Eating Index (HEI) per 2015-2020 Dietary Guidelines for Americans, to examine DQ. Academic performance was assessed using self-reported GPA. Students were predominantly non-Hispanic White (63%), Female (61%), and 75% had at least one parent graduate college. Descriptive statistics, correlation, and one-way ANOVAs were used to analyze the data using SPSS V.29. GPA was categorized into 3 groups: high, mid, and low GPA groups. Results were significant when p < 0.05.

RESULTS: DQ scores ranged from 21% to 68%, with a mean of 44% (SD: ±2.494). There were no significant associations between GPA and total DQ. However, significant associations were found between gender and specific dietary components. Further, total protein scores were greater among students with a high GPA compared to low and mid-GPA groups (F = 5.214, p = 0.006). Plant-based protein was greater among students who had at least one parent graduate college compared to first-generation students (F = 3.435, p = 0.034). Students living independently had lower total protein scores compared to those living with family (F = 4.841, p = 0.029). Additionally, students without a current job had higher dairy scores than those employed (F = 4.280, p = 0.039).

CONCLUSION: Overall, college students reported poor DQ; however, personal (e.g., gender) and environmental factors (e.g., living arrangements) were associated with one’s DQ. Further investigation is needed to facilitate the development of effective interventions that encourage healthier dietary habits among college students to improve their overall health and wellness.

PMID:41654838 | DOI:10.1186/s12889-026-26526-x

J Orthop Surg Res. 2026 Feb 7. doi: 10.1186/s13018-026-06727-1. Online ahead of print.

ABSTRACT

PURPOSE: To explore the clinical significance of the artificial intelligence (AI)-assisted three-dimensional (3D) planning system AI-HIP in total hip arthroplasty (THA) and evaluate its accuracy and efficacy in clinical practice.

METHODS: Preoperative planning was done using the AI-HIP system in the AI group and two-dimensional (2D) template measurements in the conventional group. The two groups were compared for postoperative radiographic results, perioperative monitoring indicators, and the degree of consistency between preoperative planning and actual implant size. Postoperative Harris scores, hip joint range of motion (ROM), and Barthel index were used to evaluate clinical effectiveness.

RESULTS: None of the patients who ultimately completed the 6-months follow-up experienced adverse events such as hip dislocation and infection during follow-up. Compared to the conventional group, the AI group had significantly higher Harris scores (P = 0.026), hip ROM (P = 0.018), Barthel index (P = 0.042) at 6 months postoperatively, and conformity rates of the acetabular (P = 0.001) and femoral components (P < 0.001) between intraoperative application of prosthesis model and preoperative planning. Additionally, the AI group had significantly shorter operation time (P = 0.041), less intraoperative blood loss (P = 0.012), and smaller discrepancy between bilateral acetabular offset (P = 0.032) and vertical distance of hip center of rotation (P = 0.011). However, no statistically significant intergroup differences were observed for the acetabular abduction angle, anteversion angle, femoral offset and leg length discrepancy.

CONCLUSION: Preoperative planning for THA using the AI-HIP system has a high accuracy rate and allows for effective reconstruction of the rotation center and acetabular offset, reduction of surgical time, and early recovery of joint function. Further research is needed to confirm its potential clinical value.

CLINICAL REGISTRATION NUMBER: ChiCTR210004826, Date:28/03/2021, https://www.chictr.org.cn/showproj.html? proj=52846.

PMID:41654837 | DOI:10.1186/s13018-026-06727-1

BMC Oral Health. 2026 Feb 7. doi: 10.1186/s12903-026-07727-7. Online ahead of print.

ABSTRACT

OBJECTIVE: This study evaluated multiple deep learning approaches for automated classification of dental conditions in panoramic radiographs, comparing custom convolutional neural networks (CNNs), hybrid CNN-machine learning models, and fine-tuned pre-trained architectures, comparing the performance of custom convolutional neural networks (CNNs), hybrid CNN-machine learning models, and fine-tuned pre-trained architectures for detecting fillings, cavities, implants, and impacted teeth.

METHODS: A dataset of 1,512 panoramic X-ray images with 11,137 manually annotated bounding boxes for four dental conditions (fillings, cavities, implants, and impacted teeth) was analyzed, with regions of interest extracted using expert annotations for subsequent AI-based classification. Class imbalance was addressed through random downsampling, creating a balanced dataset of 894 samples per condition. Multiple approaches were evaluated via 5-fold cross-validation: a custom CNN, hybrid models combining CNN features with traditional classifiers (Support Vector Machine, Decision Tree, Random Forest), and fine-tuned pre-trained networks (VGG16, Xception, ResNet50). Performance was assessed using accuracy, precision, recall, and F1-score metrics.

RESULTS: The hybrid CNN-Random Forest model achieved the highest accuracy of 85.4 ± 2.3% with macro-F1 score of 0.843 ± 0.028, representing an 11% point improvement over the custom CNN (74.29% accuracy, 0.724 macro-F1). VGG16 demonstrated superior pre-trained architecture performance (82.3 ± 2.0% accuracy, 0.817 macro-F1), followed by Xception (80.9 ± 2.3%) and ResNet50 (79.5 ± 2.7%). CNN + Random Forest exhibited exceptional fillings detection (F1: 0.860 ± 0.033) with balanced multi-class performance. Systematic misclassifications between morphologically similar conditions revealed inherent diagnostic challenges.

CONCLUSION: Hybrid CNN-based approaches combining feature extraction with Random Forest classification provide superior discriminative capability for dental condition detection on manually annotated regions compared to standalone architectures. While computationally efficient hybrid models show promise as supportive diagnostic tools, observed misclassification patterns indicate these AI systems should serve as adjuncts to clinical expertise, requiring prospective validation studies.

PMID:41654817 | DOI:10.1186/s12903-026-07727-7