Category: Statistics

Med Phys. 2026 Mar;53(3):e70361. doi: 10.1002/mp.70361.

ABSTRACT

BACKGROUND: Breath-hold PET imaging helps reduce respiratory motion artifacts in thoracoabdominal scans. However, its clinical application is limited by the short acquisition time, which results in significant image noise and poor lesion detectability. Enhancing image quality under such conditions remains a technical challenge.

PURPOSE: To improve the image quality of 15-s breath-hold PET scans, we investigated a deep learning-based framework using a diffusion probabilistic model. The goal was to suppress noise and enhance lesion visibility while maintaining quantitative accuracy under severely limited acquisition durations.

METHODS: We propose TAM-DiffPET, a denoising diffusion probabilistic model (DDPM) augmented with Temporal Attention Modulation (TAM) to refine intermediate feature representations by injecting diffusion time-step embeddings and temporal contextual cues. The model was trained on paired PET datasets comprising 15-s breath-hold scans and 5-min free-breathing scans from 230 patients at Ren Ji Hospital; 180 cases were used for training and 50 for quantitative and qualitative evaluation. Performance was assessed using PSNR, SSIM, and voxel-wise SUV distributions within lesion ROIs. Visual and statistical comparisons were conducted against U-Net, CycleGAN, and a vanilla DDPM.

RESULTS: The proposed method demonstrated superior performance compared to existing deep learning-based approaches. Quantitatively, it achieved the highest PSNR (40.2 ± 1.2 dB) and SSIM (0.995 ± 0.004), significantly outperforming previous deep learning based methods such as U-Net, CycleGAN, and DDPM. Voxel-wise SUV error distributions showed lower standard deviation and mean absolute error within lesion ROIs. Visual assessments revealed enhanced lesion contrast, sharper anatomical boundaries, and reduced background noise. Difference maps confirmed minimal deviation from the 5-min reference scans. Furthermore, SUV distribution analysis across representative patients confirmed that the proposed method preserves tracer uptake consistency, offering improved fidelity in clinical lesion quantification.

CONCLUSION: Our diffusion-based framework effectively denoises breath-hold PET images acquired under ultrashort durations, offering improved visual clarity and quantitative fidelity. These results support its clinical utility in motion-prone scenarios, such as thoracic or abdominal imaging, and suggest its potential for enhancing diagnostic accuracy while reducing scan time and radiation burden on patients.

PMID:41761600 | DOI:10.1002/mp.70361

J Gerontol A Biol Sci Med Sci. 2026 Feb 27:glag062. doi: 10.1093/gerona/glag062. Online ahead of print.

ABSTRACT

BACKGROUND: The INVEST in Bone Health Trial examined the effects of weight loss (WL), WL plus resistance training (WL+RT), or WL plus weighted vest use (WL+VEST) on musculoskeletal health. This secondary analysis evaluated changes in muscle area and density using computed tomography (CT) and lean and fat mass using dual-energy X-ray absorptiometry (DXA).

METHODS: 150 participants (50/group) were randomized to 12-months of WL, WL+RT, or WL+VEST, undergoing CT and DXA scans at baseline, six- and 12-months. DXA measured lean and fat mass, while CT assessed muscle and intermuscular adipose tissue (IMAT) cross-sectional area (CSA) and density. Mixed linear models evaluated changes and treatment effects, and partial Pearson’s correlations examined relationships between weight change and CT/DXA outcomes.

RESULTS: Participants (66.4±4.6 years,75% female,69% white) were living with overweight (14.7%) or obesity (85.3%). All groups achieved similar and significant weight loss (∼10%). At 12-months, WL+RT increased mid-thigh muscle CSA (0.5%, p < 0.05), improved muscle density (3.7-5.9%, p < 0.03), and reduced IMAT (20-22%, p < 0.05) and fat masses (22-26.8%, all p < 0.061). At the trunk, WL+VEST showed a trend toward muscle preservation and improved density (4.2%, p = 0.08) compared to WL, but had minimal impact on other measures. Differences between WL+VEST and WL were insignificant (all p > 0.05), but group comparisons showed improvements for WL+RT. Weight loss correlated with increased muscle density (r < 0, p < 0.001) but reduced muscle CSA and IMAT (r > 0, p < 0.001), indicating improved quality but reduced quantity.

CONCLUSION: Our findings underscore the significance of weight loss-associated muscle loss, highlighting progressive RT as a minimally effective preservation strategy.

PMID:41761582 | DOI:10.1093/gerona/glag062

Wound Repair Regen. 2026 Mar-Apr;34(2):e70136. doi: 10.1111/wrr.70136.

ABSTRACT

High-frequency ultrasound (HFUS) allows non-invasive visualization of skin microarchitecture, offering quantitative assessment of dermal composition and vascularity, but its systematic use to track temporal changes in postoperative wound healing is still limited. This study aimed to describe and validate HFUS morphologic and vascular features corresponding to the biological phases of cutaneous surgical wound healing. A total of 730 patients who underwent surgical excision of skin lesions were evaluated at different postoperative intervals using high- and ultra-high-frequency ultrasound (48-70 MHz). Dermal thickness, echogenicity and vascularity were analysed with B-mode and colour Doppler imaging through quantitative and semi-quantitative methods and reproducibility was assessed using intraclass correlation coefficients (ICC) and Cohen’s κ statistics. Cross-sectional analysis demonstrated a progressive structural and vascular evolution consistent with canonical healing phases: dermal thickness decreased from 2.45 ± 0.38 mm at T0 to 1.58 ± 0.21 mm at T4, while echogenicity increased from 0.5 [0-1] to 2.5 [2, 3], reflecting collagen compaction and maturation. Vascularity peaked at T2 (2.2 ± 0.5) and declined to 0.8 ± 0.3 by T4, paralleling the regression of angiogenesis. Measurement reproducibility was excellent (ICC = 0.91; κ = 0.82). HFUS morphologic patterns closely mirrored the biological sequence from inflammatory oedema through granulation and fibroplasia to collagen remodelling, providing real-time in vivo correlates of tissue repair. These findings support HFUS as a reliable, quantitative and reproducible tool for monitoring postoperative wound healing and as a potential imaging biomarker framework for early detection of abnormal scar evolution.

PMID:41761379 | DOI:10.1111/wrr.70136

Hum Genomics. 2026 Feb 28. doi: 10.1186/s40246-026-00913-2. Online ahead of print.

NO ABSTRACT

PMID:41761370 | DOI:10.1186/s40246-026-00913-2

Perioper Med (Lond). 2026 Feb 28. doi: 10.1186/s13741-026-00664-7. Online ahead of print.

NO ABSTRACT

PMID:41761352 | DOI:10.1186/s13741-026-00664-7

J Neuroeng Rehabil. 2026 Feb 27. doi: 10.1186/s12984-025-01819-1. Online ahead of print.

ABSTRACT

BACKGROUND: Home exercise programs prescribed to children with cerebral palsy (CP) are often associated with low adherence. Interactive technologies can help motivate and guide children through exercise programs at home, reducing onus on parents. This study sought to understand the impact of movement-tracking feedback on children’s engagement and parents’ experiences within an interactive computer play home exercise program (ICP-HEP), Bootle Boot Camp.

METHODS: A multi-case mixed methods study was conducted with three children with CP and their parents. In the quantitative single case experimental design with alternating treatments phase, children used the ICP-HEP with and without movement-tracking feedback for four weeks, and exercise adherence, exercise fidelity (movement performance quality), perceived level of fun and helpfulness for the body (i.e., 5-point rating scales and survey) were evaluated. The version (feedback/no feedback) with the highest exercise adherence was carried out for two additional weeks. Dyadic (child/parent) qualitative interviews followed. Quantitative data were analyzed using visual and statistical approaches. Qualitative data were analyzed using directed content analysis. Quantitative and qualitative results were merged through narrative weaving and joint displays.

RESULTS: Accuracy of the movement tracking and feedback provided varied among children, exercises, and play environments. Feedback may have contributed positively to exercise adherence for two children, with a significant enhancement (p < 0.001) for one of these children, and no observed negative impacts for the third child. Parents and one child perceived feedback as generally being useful for learning about movement quality, however when perceived to be inaccurate, it may have been ignored. While children had varied perspectives on how fun and helpful feedback was, it was valued by all parents. All children experienced some frustration due to sporadic technical issues. All children/parents preferred Bootle Boot Camp over conventional home programs, and suggested game refinements to enhance this ICP-HEP experience.

CONCLUSION: Use of an interactive therapy game has the potential to support children’s adherence to and children’s/parents’ experiences with home exercise, with feedback impacting children differently based on personal and environmental factors. This study serves as a foundation for future game refinements and larger-scale testing that will continue to explore the impact of feedback within an ICP-HEP.

TRIAL REGISTRATION: NCT05998239.

PMID:41761323 | DOI:10.1186/s12984-025-01819-1

Hum Genomics. 2026 Feb 27. doi: 10.1186/s40246-026-00928-9. Online ahead of print.

ABSTRACT

BACKGROUND: The interplay between genetic factors and COVID-19 susceptibility and severity underscores the critical roles of genetic variations in the responses to the virus. Specifically, genetic variations in genes such as AAK1 and ADAM17 may influence the molecular pathways that determine how the virus enters cells and how the immune system responds, thereby affecting disease outcomes. Identifying these potential genetic variants clarifies individual responses to infectious diseases and helps aid in developing effective targeted therapeutic strategies.

METHODS: We performed targeted next-generation sequencing focusing on specific single-nucleotide variants (SNVs) in AAK1, ADAM17, and CD209 genes, which are implicated in the entry of SARS-CoV-2 into host cells. The study was conducted in a Middle Eastern cohort, comprising 96 COVID-19 patients with varying disease severities and 69 healthy controls. The correlation between the prevalence of the investigated genetic variants and the serum level of inflammatory cytokines within the studied cohort was also evaluated.

RESULTS: Our analysis revealed statistically significant differences in genetic variants between COVID-19 patients and healthy controls. Notably, a 5’UTR variant, rs12692386: chr2:9695906, A > G, in the ADAM17 gene showed a significant association (p = 0.039). Additionally, two variants in the AAK1 gene, an intronic variant chr2:69732672, C > A (p = 0.029) and a missense variant rs1275698668: chr2:69747984, G > C (p = 0.017), were identified, suggesting their potential role in influencing disease susceptibility and severity. The gender-stratified analysis between the two groups showed a significant difference in the AAK1-SNV-rs1275698668 (p = 0.027) in female susceptibility, suggesting a protective effect against SARS-CoV-2 infection. The AAK1-SNV-rs1275698668 showed a significant difference between the three severity groups (p = 0.045). Prediction in-silico tools suggest that 2:g.69747984G > C and 2:g.9,695,906 A > G have potential functional/regulatory impacts on the ADAM17 and AAK1 genes, respectively. Moreover, different correlation patterns between the identified genetic variants and inflammatory cytokine levels (including CD40 ligand, IL-1b, GM-CSF, and IL-4) were observed in COVID-19 patients.

CONCLUSIONS: Our findings suggest potential genetic biomarkers in AAK1 and ADAM17 genes that could affect the disease severity and circulating cytokine levels in COVID-19 patients.

PMID:41761307 | DOI:10.1186/s40246-026-00928-9

BMC Res Notes. 2026 Feb 28. doi: 10.1186/s13104-026-07754-1. Online ahead of print.

ABSTRACT

OBJECTIVES: This paper aims to provide comprehensive documentation of the Occupational Job Strain Index (JSI), developed using data from the Norwegian nationwide Survey of Living Conditions conducted in 2006, 2009, 2013, 2016, and 2019. The JSI, based on Karasek’s Demand-Control Model, offers detailed information on working conditions, addressing a gap in national register data. The surveys, including information on the psychosocial working environment, contributed to the development of the JSI through self-reported information and psychosocial exposure items by Statistics Norway (SSB).

DATA DESCRIPTION: This paper includes the scripts documenting the construction of the Psychosocial Job Exposure Matrix and the Occupational Job Strain Index (Data File 1), along with a script containing the key for converting 4-digit STYRK-08 codes into 4-digit STYRK-98 codes and the procedure for merging the index with register data (Data File 2). Data File 3 contains the complete Psychosocial Job Exposure Matrix and Occupational Job Strain Index. Data File 4 includes a full codebook.

PMID:41761288 | DOI:10.1186/s13104-026-07754-1

BMC Health Serv Res. 2026 Feb 27. doi: 10.1186/s12913-026-14233-7. Online ahead of print.

ABSTRACT

BACKGROUND: In China, national health reforms are interpreted, piloted and implemented slightly differently by each local health authority. This is the first study of China’s 2023 pharmacy reform allowing 360 million urban employee basic medical insurance (UEBMI) members the right to purchase prescription medicine using their insurance at retail pharmacies. For Wuhan, we assess the impact of the 2023 medicine purchasing reform on medicine buying behavior; evaluate the benefits of the changes to UEBMI members; and provide new insights into China’s retail pharmacies’ business model.

METHODS: From March 2022 to December 2023, we gained unique access to daily UEBMI members’ medicine purchases, covering 34,956 claims at two representative Wuhan pharmacies. The impact of the new retail pharmacy purchasing policy on medicines purchasing behavior were quantitatively evaluated using descriptive statistics and interrupted time series analysis (ITSA).

RESULTS: No significant pre-reform trend in medicine purchases was observed, confirming a stable baseline prior to the 2023 reform. However, medicine purchases from retail pharmacies showed a significant upward trend. There was no evidence of medicine over-use or fraud. The benefit to members was measured by the absence of any significant upward per capita total expenditure trend and a significant decrease in the proportion of medicine expenditures from members personal medical savings accounts. Retired UEBMI members benefited more than currently employed UEBMI members from allowing UEBMI funds for prescription purchases from retail pharmacies.

CONCLUSIONS: The 2023 reforms allowed UEBMI members to purchase medicine from retail pharmacies, providing medicine accessibility and affordability for members, especially retired members; did not trigger an increase in members’ medicine consumption; attenuated the over-use of hospital pharmacies. Future study should empirically test whether pharmacies’ business model transformed away from a singular focus on price competition towards a combination of price competition and health services.

PMID:41761281 | DOI:10.1186/s12913-026-14233-7

Nutr J. 2026 Feb 27. doi: 10.1186/s12937-025-01236-z. Online ahead of print.

NO ABSTRACT

PMID:41761272 | DOI:10.1186/s12937-025-01236-z