The segmentation of glioma is crucial for early diagnosis, according to a World Health Organization (WHO) 2021 report. For glioma diagnosis, 3D multi-modal brain MRI/CT imaging has become an essential tool, offering detailed information. Nowadays, deep learning frameworks have been applied to various medical imaging problems, including brain glioma segmentation. Recently, foundation models like Segment Anything Model (SAM) have emerged as pivotal tools in computer vision tasks. These models are trained using large (real-world) datasets, offering a generalized understanding of visual data and semantic key features. Therefore, the effective utilization of foundation models in medical imaging is a significant area of current research. However, the differences in data distribution between multi-modal medical images and real-world images present challenges in directly applying foundation models to medical imaging. Additionally, utilizing multi-modal images to extract crucial information and its fusion poses further challenges. To address these issues, we propose a framework using foundation model and novel strategies for multi-modal fusion. Our fusion adapters effectively integrate the information from different modalities to enhance glioma segmentation in multi-modal MRI scans. Our method outperforms current state-of-the-art methods for accurate segmentation of the glioma using private and publicly available brain MRI datasets, proving the effectiveness of our approach across different datasets and imaging modalities.

2025-01-24·Current Vascular Pharmacology

Association of ST2, Galectin-3, and NT- Probnp in Elderly Hypertensive Patients and Heart Failure with a Preserved Ejection Fraction

Article

作者: Yu, Hui ; Wang, Lin ; Xiao, Fan Kai ; Yang, Fan ; Li, Ping

Purpose:The objective of this study was to explore the relationship among serum levels of the growth-stimulating expressed gene 2 protein (ST2), Galectin-3(GAL-3), N-terminal pro-B-type natriuretic peptide (NT-proBNP) in elderly hypertensive patients and heart failure with preserved ejection fraction (HFpEF).Materials and Methods:Eighty-five elderly hypertensive patients with HFpEF were registered as the HFpEF group, and 46 hypertensive patients without HF were registered as the Non-HF group. The levels of serum sST2 (soluble ST2), Galectin-3, and NT-proBNP were measured, and related indexes of heart function were performed with echocardiography in two groups, respectively.The obtained variables were applied to statistical software for analysis.Results:Age, BMI, SBP, DBP, TC, LDL-C, HCY, sST2, Galectin-3, NT- proBNP, LVEDD, IVSD, LVEF, and E/A were obviously different between the two groups (p < 0.05). The levels of sST2, Galectin- 3 and NT- proBNP in the HFpEF group were higher than in the Non-HF group (P < 0.05). ANOVA results indicated that sST2, Galectin-3, and NT- proBNP levels increased gradually with the increasing NYHA grades (P<0.05). BMI, SBP, DBP, TC, LDL-C, FBG,UA, HCY, LVEDD, IVSD, LVEF, and E/A were significant differences in patients with different NYHA classes (P < 0.05). Spearman indicated that sST2, Galectin-3, and NT-proBNP were positively correlated with BMI, SDP, DBP, LDL-C, FBG, and HCY (P < 0.05). Logistic analysis indicated that BMI, SBP, DBP, FBG, HCY,sST2, Galectin-3, NT-proBNP, LVEDD, LVEF, and E/A were risk factors for hypertension with HFpEF. (P < 0.05). ROC indicated that the AUC of the diagnostic performance of sST2, Galectin-3, and NT-proBNP were all above 0.7, which may have some forecasting value for elderly hypertensive patients with HFpEF.Conclusion:The levels of sST2, Galectin-3, and NT-proBNP were closely related to cardiac function grades. sST2, Galectin-3, and NT-proBNP have similar diagnostic performance and predictive value for elderly hypertensive patients with HFpEF. sST2 was more sensitive than NT-proBNP. It is recommended that measurements of sST2, Galectin-3 and NT-proBNP levels in elderly hypertensive patients may be useful in classifying early HFpEF.

2024-10-29·2024 IEEE 13th Global Conference on Consumer Electronics (GCCE)

Knowledge Distillation Using Segment Anything to U-Net Model for Lightweight High accuracy Medical Image Segmentation