RESUMEN
Background: In recent years, diseases caused by abnormal immune-inflammatory responses have become increasingly severe. Dietary intervention involving omega-3 polyunsaturated fatty acids (ω-3 PUFAs) has emerged as a potential treatment. However, research investigating the relationship between ω-3, ω-6 PUFAs, and ω-6 to ω-3 ratio with inflammatory biomarkers remains controversial. Methods: To investigate the correlation between the intake of ω-3 and ω-6 PUFAs and the ratio of ω-6: ω-3 with biomarkers of inflammation, the National Health and Nutrition Examination Survey (NHANES) data (1999 to 2020) was utilized. The systemic immune-inflammation index (SII), platelet-lymphocyte ratio (PLR), neutrophil-lymphocyte ratio (NLR), and white blood cell (WBC) were selected as study subjects. Dietary data for ω-3 and ω-6 PUFAs were collected via two 24-h dietary recall interviews. SII index and other indicators were obtained from the blood routine data. The multiple linear regression and restricted cubic spline models were utilized to evaluate the association of ω-3, ω-6 PUFAs intake, and ω-6: ω-3 ratio to SII and secondary measures. Results: This study involved a total of 43,155 American adults. ω-3 and ω-6 PUFAs exhibited negative correlations with SII, PLR, NLR, and WBC. The correlation between ω-6: ω-3 ratio and SII, PLR, NLR, and WBC was not significant. Furthermore, the dose-response relationship showed that the relationship between the intake of ω-3 and ω-6 PUFAs and SII was an "L" pattern. Conclusion: Intake of dietary ω-3 and ω-6 PUFAs reduces the levels of several inflammatory biomarkers in the body and exerts immunomodulatory effects.
RESUMEN
Due to the cross-domain distribution shift aroused from diverse medical imaging systems, many deep learning segmentation methods fail to perform well on unseen data, which limits their real-world applicability. Recent works have shown the benefits of extracting domain-invariant representations on domain generalization. However, the interpretability of domain-invariant features remains a great challenge. To address this problem, we propose an interpretable Bayesian framework (BayeSeg) through Bayesian modeling of image and label statistics to enhance model generalizability for medical image segmentation. Specifically, we first decompose an image into a spatial-correlated variable and a spatial-variant variable, assigning hierarchical Bayesian priors to explicitly force them to model the domain-stable shape and domain-specific appearance information respectively. Then, we model the segmentation as a locally smooth variable only related to the shape. Finally, we develop a variational Bayesian framework to infer the posterior distributions of these explainable variables. The framework is implemented with neural networks, and thus is referred to as deep Bayesian segmentation. Quantitative and qualitative experimental results on prostate segmentation and cardiac segmentation tasks have shown the effectiveness of our proposed method. Moreover, we investigated the interpretability of BayeSeg by explaining the posteriors and analyzed certain factors that affect the generalization ability through further ablation studies. Our code is released via https://zmiclab.github.io/projects.html.