Brain tumors are one of the major health threats to humans, and their complex pathological features and anatomical structures make accurate segmentation and detection crucial. However, existing models based on Transformers and Convolutional Neural Networks (CNNs) still have limitations in medical image processing. While Transformers are proficient in capturing global features, they suffer from high computational complexity and require large amounts of data for training. On the other hand, CNNs perform well in extracting local features but have limited performance when handling global information. To address these issues, this paper proposes a novel network framework, MUNet, which combines the advantages of UNet and Mamba, specifically designed for brain tumor segmentation. MUNet introduces the SD-SSM module, which effectively captures both global and local features of the image through selective scanning and state-space modeling, significantly improving segmentation accuracy. Additionally, we design the SD-Conv structure, which reduces feature redundancy without increasing model parameters, further enhancing computational efficiency. Finally, we propose a new loss function that combines mIoU loss, Dice loss, and Boundary loss, which improves segmentation overlap, similarity, and boundary accuracy from multiple perspectives. Experimental results show that, on the BraTS2020 dataset, MUNet achieves DSC values of 0.835, 0.915, and 0.823 for enhancing tumor (ET), whole tumor (WT), and tumor core (TC), respectively, and Hausdorff95 scores of 2.421, 3.755, and 6.437. On the BraTS2018 dataset, MUNet achieves DSC values of 0.815, 0.901, and 0.815, with Hausdorff95 scores of 4.389, 6.243, and 6.152, all outperforming existing methods and achieving significant performance improvements. Furthermore, when validated on the independent LGG dataset, MUNet demonstrated excellent generalization ability, proving its effectiveness in various medical imaging scenarios. The code is available at https://github.com/Dalin1977331/MUNet.

Tongue diagnosis in Traditional Chinese Medicine (TCM) plays a crucial role in clinical practice. By observing the shape, color, and coating of the tongue, practitioners can assist in determining the nature and location of a disease. However, the field of tongue diagnosis currently faces challenges such as data scarcity and a lack of efficient multimodal diagnostic models, making it difficult to fully align with TCM theories and clinical needs. Additionally, existing methods generally lack multi-label classification capabilities, making it challenging to simultaneously meet the multidimensional requirements of TCM diagnosis for disease nature and location. To address these issues, this paper proposes TongueNet, a multimodal deep learning model that integrates tongue image data with text-based features. The model utilizes a Hierarchical Aggregation Network (HAN) and a Feature Space Projection Module to efficiently extract and fuse features while introducing consistency and complementarity constraints to optimize multimodal information fusion. Furthermore, the model incorporates a multi-scale attention mechanism (EMA) to enhance the diversity and accuracy of feature weighting and employs a Kolmogorov-Arnold Network (KAN) instead of traditional MLPs for output optimization, thereby improving the representation of complex features. For model training, this study integrates three publicly available tongue image datasets from the Roboflow platform and enlists multiple experts for multimodal annotation, incorporating multi-label information on disease nature and location to align with TCM clinical needs. Experimental results demonstrate that TongueNet outperforms existing models in both disease nature and disease location classification tasks. Specifically, in the disease nature classification task, it achieves 89.12% accuracy and an AUC of 83%; in the disease location classification task, it achieves 86.47% accuracy and an AUC of 81%. Moreover, TongueNet contains only 32.1 M parameters, significantly reducing computational resource requirements while maintaining high diagnostic performance. TongueNet provides a new approach for the intelligent development of TCM tongue diagnosis.

Introduction: This study investigated the association between AIM2 cg11003133 DNA methylation and Rheumatoid Arthritis (RA), evaluating its diagnostic potential for RA and subtypes. Methods: MethylTarget™ sequencing targeted AIM2 cg11003133 (chr1:159076528-159076740) in RA, Ankylosing Spondylitis (AS), Psoriatic Arthritis (PsA), gout, Systemic Lupus Erythematosus (SLE), Dermatomyositis (DM), primary Sjögren's Syndrome (SS), and Healthy Controls (HC) patients. Logistic regression, random forest, and XGBoost models were applied, with Spearman’s correlation used to assess associations. Results: RA and RF/CCP-positive patients showed significantly higher methylation at cg11003133_79/91 compared to HC and AS (FDR < 0.05), but lower levels compared to DM. Methylation at cg11003133_139 was elevated in RA compared to AS/SS (FDR = 0.04/0.03). Anti- TNF-α non-responders had higher cg11003133_79/91 methylation levels compared to HC/AS non-responders (FDR < 0.05). RF-negative RA patients had higher cg11003133_91 methylation than AS patients who failed anti-TNF-α treatment (FDR < 0.05). Haplotype CCCC correlated positively with CRP (r = 0.14, P = 0.006); TTTT was significantly negatively correlated with erythrocyte sedimentation rate, CRP, and the presence of diabetes (r = -0.18, -0.15, and -0.14; P < 0.001, 0.003, and 0.008, respectively). XGBoost and RF models achieved AUCs of 0.9911 and 0.9975 for RA versus non-RA, and 1 for RF/CCP double-negative versus double-positive RA. Discussion: AIM2 cg11003133 methylation is strongly linked to RA, aligning with its role in inflammasome activation. While promising for diagnostics, larger validation is needed. Conclusions: AIM2 cg11003133 methylation may serve as a diagnostic biomarker for RA and subtypes.

Analyze the correlation between the DNA methylation levels of Protocadherin 17 (PCDH17) cg03865667 and rheumatoid arthritis (RA), and evaluate its potential as a biomarker for diagnosing RA. Peripheral blood samples were collected from a cohort of 370 individuals, comprising patients diagnosed with RA, ankylosing spondylitis (AS), psoriatic arthritis (PsA), gout, systemic lupus erythematosus (SLE), dermatomyositis (DM), primary Sjögren’s syndrome (SS), and healthy controls (HC), for subsequent analysis.DNA methylation sequencing techniques were employed to evaluate the methylation levels of the PCDH17 cg03865667 locus. Relative to the HC, AS, and SS groups, PCDH17 cg03865667 was significantly downregulated in RA patients (P = 0.0403; p = 0.0290; p = 0.044). Compared to the HC group, the methylation levels at CpG sites 57,631,544, 57,631,571, and 57,631,581 were significantly downregulated in RA patients (p = 0.0078; p = 0.0123; p = 0.0309). For the TTCCTT and TTTCTT haplotypes, methylation levels were significantly lower in RA patients than in HC (p = 0.0188; p = 0.0053), particularly for the TTTCTT haplotype. Significant differences were observed between the CCP(-) RF(-) group, the CCP(+) / RF(+) group, and the HC group among the RA subgroups. No significant differences were found within the double-positive subgroup.The average methylation level of PCDH17 was negatively correlated with C-reactive protein (CRP) (r=-0.28, p = 6.9e-4). This study indicates that PCDH17 cg03865667 methylation may function as a potential biomarker for RA diagnosis. The subgroup analysis suggests that it may serve as a potential biomarker for diagnosing RA, indicating a capacity to enhance the diagnostic accuracy for seronegative RA and reduce the likelihood of missed and incorrect diagnoses.

Background: Autoimmune hepatitis (AIH) is a chronic liver disease characterized by immune-mediated liver inflammation. Despite its global prevalence, the pathogenesis of AIH remains poorly understood, and there is a lack of specific biomarkers and targeted treatments. This study aimed to investigate the role of hsa_circ_0109623, hsa-miR-146b-3p, and Sortilin 1 (SORT1) in AIH and their potential as therapeutic targets. Methods: We collected liver tissue samples and peripheral blood mononuclear cells from patients with AIH and healthy controls and performed RT-PCR, western blotting, flow cytometry, and other molecular biology techniques to analyze the expression of hsa_circ_0109623, hsa-miR-146b-3p, and SORT1. We also used bioinformatics tools to predict the interaction between these molecules and conducted luciferase reporter assays to confirm their binding. Results: hsa_circ_0109623 was significantly upregulated in patients with AIH and positively correlated with inflammatory activity. We also found that hsa_circ_0109623 could enhance CD4+ T-cell activation and promote the expression of proinflammatory cytokines. Conversely, hsa-miR-146b-3p was downregulated in patients with AIH and negatively correlated with the expression of hsa_circ_0109623 and SORT1. In addition, hsa-miR-146b-3p acted as a sponge for hsa_circ_0109623, inhibiting CD4+ Th1 cell polarization and cytokine production. SORT1 was also upregulated in patients with AIH and acted as a sponge for hsa-miR-146b-3p, promoting CD4+ Th1 cell polarization and cytokine expression. Furthermore, hsa_miR_146b-3p/SORT1 can regulate the STAT1/STAT4 signaling pathway mediating the progression of AIH. Conclusions: The hsa_circ_0109623/hsa-miR-146b-3p/SORT1 axis plays a crucial role in the pathogenesis of AIH by regulating CD4+ T-cell activation and cytokine production. These molecules may serve as potential biomarkers and therapeutic targets for AIH. Further research is needed to validate these findings and explore their clinical applications.