Category: Statistics

Int Orthop. 2026 Jun 22. doi: 10.1007/s00264-026-06921-0. Online ahead of print.

ABSTRACT

PURPOSE: Accurate acetabular cup placement is essential in total hip arthroplasty (THA). We hypothesized that the newly introduced Computed Tomography (CT)-based portable navigation system would demonstrate accuracy comparable to that of the imageless portable navigation system. The aim of this study was to compare cup placement accuracy between the CT-based and imageless portable navigation systems of the same platform in THA performed in the lateral decubitus position.

METHODS: This retrospective cross-sectional study included 36 patients who underwent primary THA via a direct lateral approach in the lateral decubitus position. In all cases, both imageless and CT-based portable navigation systems were used concurrently. Postoperative cup alignment was evaluated using three-dimensional CT (3D-CT). The primary outcome was the absolute error in cup inclination and anteversion, defined as the difference between intraoperative navigation values and postoperative 3D-CT measurements in the functional. Secondary outcomes included outlier rates and registration success rates.

RESULTS: No statistically significant differences were observed between the imageless and CT-based portable navigation systems in the mean absolute error for inclination (2.2 ± 1.8° vs. 2.3 ± 1.8°, p = 0.93) or anteversion (2.3 ± 2.3° vs. 2.6 ± 2.5°, p = 0.41). There were no significant differences in outlier proportions. The registration success rate was 92% (36/39) due to three technical failures.

CONCLUSION: In this preliminary study, the CT-based portable navigation system demonstrated cup placement accuracy comparable to that of the imageless portable navigation system. Although the CT-based system may provide additional spatial information intraoperatively, its impact on clinical outcomes remains unclear and requires further longitudinal investigation.

PMID:42332259 | DOI:10.1007/s00264-026-06921-0

Pediatr Res. 2026 Jun 22. doi: 10.1038/s41390-026-05217-8. Online ahead of print.

ABSTRACT

BACKGROUND: Cystic fibrosis (CF) results from mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) gene, causing multisystem disease and impaired quality of life. CFTR modulators have improved pulmonary and nutritional outcomes, yet potential metabolic effects such as dyslipidemia remain a concern, particularly in adults. Pediatric data are limited. This study evaluated the effects of CFTR modulators on lipid and lipoprotein profiles in children with CF.

METHODS: This retrospective study was conducted at a tertiary pediatric pulmonology center. Body mass index (BMI) Z-scores, lung function tests, and serum lipid profiles were compared between baseline and six months after initiation of CFTR modulator therapy.

RESULTS: Twenty-six patients were included finally. The median age was 11 (5.9-15.6) years; 53.8% were female, and 80.8% received elexacaftor/tezacaftor/ivacaftor. After six-month therapy, significant improvements were observed in BMI (p = 0.047), FEV₁ (p = 0.005), and FVC Z-scores (p = 0.001). Although HDL, LDL, VLDL, triglycerides, and total cholesterol increased numerically, none reached statistical significance. No difference was found between BMI changes and lipid alterations.

CONCLUSION: Over six months of continuous CFTR modulator therapy, we observed improvements in pulmonary and nutritional outcomes without evidence of clinically meaningful lipid deterioration.

IMPACT: This study adds real-world pediatric data regarding lipid profile changes during CFTR modulator therapy. While these therapies significantly improve pulmonary and nutritional outcomes, we observed no evidence of lipid deterioration over six months of treatment. These findings contribute to the evolving understanding of the systemic effects of CFTR modulators in childhood and underscore the importance of ongoing cardiometabolic evaluation as survival improves.

PMID:42332244 | DOI:10.1038/s41390-026-05217-8

Pediatr Res. 2026 Jun 22. doi: 10.1038/s41390-026-05199-7. Online ahead of print.

ABSTRACT

BACKGROUND: We sought to evaluate oxidative changes in premature infants receiving 100% oxygen compared with 30% during deferred cord clamping (DCC).

METHODS: Premature infants born at 22^0/7 to 28^6/7 weeks received DCC in conjunction with either 30% (LO Group) or 100% (HI Group) oxygen. Blood was extracted from a preserved umbilical segment and a postnatal sample was collected from umbilical vascular lines within two hours of birth. Reduced-to-oxidized glutathione (GSH/GSSG) ratios were analyzed using liquid chromatography coupled to tandem mass spectrometry.

RESULTS: Sixty-eight infants had data available for analysis. The median (IQR) gestational age of infants was 26^4/7 (24^6/7, 28^2/7) weeks in both groups. Among infants receiving 100% versus 30% oxygen, median (IQR) GSH/GSSG ratio were not statistically different in arterial cord blood [7.5 (0.6, 290) vs 37 (1.1, 265), p = 0.52] or venous cord blood [8.4 (2,50) vs 76 (5, 210), p = 0.12] or postnatal samples [14 (2, 290) vs 8 (2, 280), p = 0.98)].

CONCLUSION: Briefly providing 30% vs. 100% oxygen for 90 seconds during DCC showed no significant difference in GSH/GSSG ratios, but redox effects remain unclear given variability, sample size and limited power. Further studies are needed to ascertain potential oxidative damage during neonatal resuscitation and deferred cord clamping. THIS TRIAL IS REGISTERED ON CLINICALTRIALS.

GOV ID: NCT04413097 IMPACT: The effect of oxygen administration during deferred cord clamping on redox status is unclear due to large variability in GSH and GSSG values and small sample size. These data provide some insights about umbilical arterial and venous oxygen levels and the effect of placenta on GSH/GSSG in preterm infants. Further basic and clinical studies are needed to better ascertain the potential for oxidative damage during neonatal resuscitation and deferred cord clamping.

PMID:42332243 | DOI:10.1038/s41390-026-05199-7

J Imaging Inform Med. 2026 Jun 22. doi: 10.1007/s10278-026-02061-4. Online ahead of print.

ABSTRACT

Accurate kidney ultrasound segmentation is fundamental for clinical measurement and computer-aided diagnosis. However, domain shifts across devices and centers-manifested as differences in grayscale intensity, contrast, and speckle texture statistics-can substantially degrade model generalization, while acquiring new pixel-level annotations is costly. To address this, we propose a statistical spectral-similarity-guided ultrasound-to-ultrasound translation method to improve kidney segmentation performance without target-domain annotations. Motivated by frequency-domain analysis of renal ultrasound data, we observe that mid-to-low frequency components, which encode global organ structure, exhibit high consistency across domains, whereas mid-to-high frequency components, dominated by device-dependent speckle and texture statistics, vary substantially. Based on dataset-level frequency statistics, our method automatically identifies spectrally similar frequency bands shared by the source and target domains and derives structural guidance from them. This guidance is injected as a soft condition throughout a diffusion-based image generation process, enabling translation to target-device appearance while preserving anatomical structure. The translated images, paired with source-domain labels, are then used to train a segmentation network without requiring any target-domain annotations. Experiments on two public renal ultrasound datasets (OKUS and UNK) and an in-house multi-center dataset demonstrate superior structural preservation in image translation and consistently improved downstream segmentation performance, with particularly large reductions in boundary error. In the challenging OKUS to UNK adaptation scenario, our method boosts the mean Dice score by up to 20.52% (from 56.05% to 76.57%) and drastically reduces the 95% Hausdorff Distance (HD95) boundary error by 71.96 mm compared to the direct transfer baseline. Furthermore, consistent performance gains are achieved across the in-house multi-center dataset. These results indicate that the proposed spectral-similarity-based guidance effectively handles ultrasound domain shifts, substantially improving robustness and generalization for kidney segmentation under zero-shot and cross-center settings.

PMID:42332239 | DOI:10.1007/s10278-026-02061-4

J Imaging Inform Med. 2026 Jun 22. doi: 10.1007/s10278-026-02029-4. Online ahead of print.

ABSTRACT

Breast density is a breast cancer risk factor. The accurate quantification of breast density requires reliable segmentation of dense tissue in mammograms, but it is a challenging task due to large variations in tissue appearance across hospitals and imaging devices. We propose MammoDenseSegNet, a new deep encoder-decoder convolutional neural network designed to enhance segmentation performance through two complementary modules: a) Adaptive dual attention module, which captures long-range spatial and channel interdependencies to provide focused attention on relevant dense tissue areas regardless of their location; and b) Multi kernel receptive field module, which enlarges the network’s receptive field at the bottleneck layer to aggregate multi-scale contextual features. Additionally, a multi-scale dice loss with deep supervision guides learning across decoder levels to improve robustness. We evaluated MammoDenseSegNet on two public digital mammogram datasets (VinDR-Mammo and EMBED) and one private dataset, spanning a variety of breast densities and imaging artifacts in a total of 1499 images from 606 women. Statistical analysis was done using generalized linear models accounting for correlation among images from the same women and adjusting for potential confounders (proc genmod, proc mixed, SAS v.9.4, SAS Institute, Cary, NC). MammoDenseSegNet demonstrated consistently high performance across various conditions (with Recall ranging from 0.64 to 0.90 and Dice from 0.63 to 0.91) and significantly (p < 0.001) outperformed the publicly available state-of-the-art algorithm based on the VGG16 (with Recall from 0.04 to 0.91 and Dice from 0.06 to 0.82 across the same conditions). The improvement was largest for low-density tissue, where the baseline algorithm practically fails (with the mean Recall of 0.14 and Dice of 0.16) while MammoDenseSegNet remained clinically useful (with the mean Recall of 0.66 and Dice of 0.63).

PMID:42332236 | DOI:10.1007/s10278-026-02029-4

Ann Dyslexia. 2026 Jun 22. doi: 10.1007/s11881-026-00372-3. Online ahead of print.

ABSTRACT

Many children with oral language difficulties also experience challenges with word reading, as evidenced in the high comorbidity rate between developmental language disorders and dyslexia. The current study investigated a sample of students (n = 357) with language and literacy difficulties classified into latent classes based on pre-intervention performance on word reading and listening comprehension measures. Specifically, this study sought to determine the extent to which latent classes responded to an evidence-based, narrative language program at immediate post-test and a five-month follow-up time point, and whether performance varied as a function of class membership. Findings revealed that students in all latent profiles showed statistically significant improvements in narrative language at post-test, a primary target of the intervention. However, intervention effects varied at a five-month follow-up time point, with students with the greatest listening comprehension and word reading difficulties showing the most notable gains. Instructional response also varied according to performance on a narrative writing measure. These findings suggest that assessing initial performance on key variables may be able to help educators predict student response and adapt intervention plans to more effectively meet individual needs.

PMID:42332233 | DOI:10.1007/s11881-026-00372-3

Bone Marrow Transplant. 2026 Jun 23. doi: 10.1038/s41409-026-02934-w. Online ahead of print.

ABSTRACT

Gastrointestinal graft-versus-host disease (GI-GVHD) is a major complication after allogeneic hematopoietic stem cell transplantation (allo-HSCT), often requiring invasive endoscopy for diagnosis. Fecal calprotectin (FC) offers a non-invasive marker of intestinal inflammation, but its utility in GI-GVHD needs clarification. In this prospective observational study of 165 adult allo-HSCT recipients, FC was measured on days +7, +14, and +21 post-transplant, at GI-GVHD onset, and 7 days post-treatment, alongside clinical, endoscopic, and histological data. GI-GVHD developed in 52.7% of patients, with histological confirmation in 90.3% of cases. FC lacked predictive value on day +7 (AUC = 0.50) but showed moderate-to-high accuracy on days +14 (AUC = 0.69) and +21 (AUC = 0.77), with an optimal day +21 cutoff of 52.5 µg/g (sensitivity 75%, specificity 87%). At diagnosis, median FC was 120 µg/g, correlating with endoscopic severity (r = 0.31; p = 0.02) but not clinical or histological grades. FC declined significantly post-treatment (120 to 51.5 µg/g; p = 0.04), though concurrent infections elevated levels without compromising discriminative ability. FC serves as a dynamic biomarker for predicting, diagnosing, and monitoring GI-GVHD, but requires integrated clinical interpretation due to limited specificity amid other inflammations.

PMID:42332220 | DOI:10.1038/s41409-026-02934-w

Comput Biol Chem. 2026 Jun 20;124(Pt 2):109188. doi: 10.1016/j.compbiolchem.2026.109188. Online ahead of print.

ABSTRACT

In the present research work, graph theory has been used to represent the molecular structures of antiviral drugs corresponding to the influenza strain treatment using degree-based topological descriptors and Laplacian energy to understand their structural and physicochemical behavior. The graph theory-based descriptors are used to construct the quantitative structure property relationship and provide input to an ANN model used to predict various physicochemical properties among the selected antiviral drugs. The comparison shows that the ANN model outperforms the Linear Regression model by demonstrating higher R² and lower RMSE. The suggested ANN-QSPR model was verified through the 5-fold cross-validation process, showing high prediction efficiency and better robustness regarding various physicochemical properties of antiviral compounds. Moreover, an MCDM approach using the TOPSIS method has been employed to assess and rank the antiviral drugs using both structural and physicochemical aspects. The integrated framework in the present research work offers a comprehensive mathematical and computational platform to perform drug analysis and decision-making in antiviral drug design tasks.

PMID:42330573 | DOI:10.1016/j.compbiolchem.2026.109188

Eur J Cancer. 2026 Jun 19;244:116898. doi: 10.1016/j.ejca.2026.116898. Online ahead of print.

ABSTRACT

BACKGROUND: Tobacco use remains a leading cause of preventable morbidity and mortality and is commonly initiated in adolescence. We evaluated whether Smokerface-Poster, a low-intensity, appearance-based school campaign promoting a photoaging app, could attenuate smoking uptake among early adolescents.

METHODS: In this two-arm cluster-randomized controlled trial, 126 German secondary schools (9797 grade 6/7 students; 96.0% were 11-13 years old; 51.4% male; 48.6% female) were allocated to intervention or control. Intervention schools displayed two classroom posters for 24 months. Smoking behavior was assessed at baseline and 24 months. As primary outcome, we investigated the between‑group difference in the change in 30-day smoking prevalence, with a number needed to treat (NNT) of < 100 students predefined as clinically relevant per protocol.

RESULTS: Baseline smoking prevalence was 7.4% in the control vs. 7.9% in the intervention group. At 24 months post-intervention, smoking prevalence increased by 19.2 %age points in control vs. 18.1 %age points in the intervention group (number needed to treat=93; adjusted ratio of odds ratios 0.87, 95% CI 0.69-1.09; p = 0.228). Favorable, non-significant patterns were also observed for anti-smoking intentions and attitudes.

CONCLUSION: Although the between-group difference was not statistically significant, the intervention reached the predefined threshold for clinical relevance, with an NNT of 93. This suggests that, for every 93 students exposed to the Smokerface-Poster campaign, one fewer adolescent would be expected to smoke over the two-year follow-up. Given its low cost of < €50 per 100 students, the intervention appears to be a promising approach to supporting school-based smoking prevention.

PMID:42330568 | DOI:10.1016/j.ejca.2026.116898

Appl Physiol Nutr Metab. 2026 Jun 22. doi: 10.1139/apnm-2026-0025. Online ahead of print.

ABSTRACT

Canada’s Food Guide (CFG) was revised in 2019 to emphasize food quality through a balanced diet rich in fruits, vegetables, whole grains, and plant-based proteins. Although the Canadian Food Scoring System (CFSS) was developed to assess the alignment of individual foods and beverages with CFG 2019, no validated tool exists to evaluate multi-ingredient recipes. In this study, we modified the CFSS to create the CFSS for recipes (CFSSr), a nutrient profiling model that classifies recipes into five alignment categories (“very poor” to “excellent”) based on food group composition and nutrient-of-concern thresholds from Canadian front-of-package labelling (FOPL) regulations. The CFSSr was applied to 93,021 unique generic recipes from the 2015 Canadian Community Health Survey – Nutrition. The majority (64%) of recipes were rated “poor” or “very poor” according to CFG 2019, with dessert and dessert topping recipes scoring lowest and nuts & seeds and snack recipes scoring highest. A sub-analysis of home-prepared recipes (n=21,774) showed a similar pattern, with 57.6% rated “poor” or “very poor”. Home-prepared recipes showed statistically significantly better alignment than otherwise-prepared recipes (median CFSSr: 45.0 vs. 43.8, p<0.001), though the effect size was negligible, suggesting that preparation location alone does not ensure dietary quality but rather the cooking method (e.g., frying or added ingredients like salt and fats during preparation). These findings underscore a persistent gap between the nutritional quality of common recipes and CFG recommendations, highlighting the need for targeted policies, food literacy initiatives, and tools to promote healthier recipe choices and preparation methods.

PMID:42330547 | DOI:10.1139/apnm-2026-0025