HCP, 多模态脑MRI, MDM
收藏资源简介:
HCP、多模态脑MRI和MDM数据集被用于验证该框架的有效性。HCP数据集包含大量高质量的多模态脑MRI图像,多模态脑MRI数据集包含多种模态的脑MRI图像,MDM数据集是内部的多壳体扩散MRI数据集。这些数据集包含大量高质量的MRI图像,可以用于视觉通路分割的研究。框架采用了一种新的半监督多参数特征分解框架,该框架可以有效地处理复杂的跨序列关系,并从不同的MRI序列中捕获互补信息。此外,该框架还开发了一种基于一致性的样本增强模块,以解决标记数据有限的问题。
The HCP, multimodal brain MRI, and MDM datasets were used to validate the effectiveness of the proposed framework. The HCP dataset comprises a large number of high-quality multimodal brain MRI images. The multimodal brain MRI dataset includes brain MRI images of various modalities, whereas the MDM dataset is an internal multi-shell diffusion MRI dataset. All these datasets contain abundant high-quality MRI images, making them suitable for research on visual pathway segmentation. The proposed framework incorporates a novel semi-supervised multi-parameter feature decomposition approach, which can effectively handle complex cross-sequence dependencies and capture complementary information from different MRI sequences. Furthermore, a consistency-based sample augmentation module is developed within this framework to address the issue of limited labeled data.

- 1Cross-Sequence Semi-Supervised Learning for Multi-Parametric MRI-Based Visual Pathway Delineation中国科学院深圳先进技术研究院 · 2025年



