Arrow Research search

Author name cluster

Yuhang Xia

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

2 papers
1 author row

Possible papers

2

EAAI Journal 2026 Journal Article

Neck computed tomography angiography generation from computed tomography via Mamba U-shaped convolutional network-based diffusion with content and style conditions

  • Yuhang Xia
  • Hongqing Zhu
  • Tong Hou
  • Ning Chen
  • Kai Chen
  • Zhong Zheng
  • Bingcang Huang

Computed tomography angiography (CTA) is a widely used imaging technique for diagnosing vascular diseases, particularly in detecting conditions such as aortic aneurysms and dissections. However, traditional CTA relies on iodinated contrast agents (ICAs), which pose risks to patients and present clinical challenges such as prolonged procedure times, high costs, and increased adverse reaction risks due to repeated use. To address these issues, this paper proposes a novel deep learning-based approach for generating CTA-like images from computed tomography (CT) scans, reducing dependence on ICAs and overcoming the limitations of conventional medical procedures. Specifically, we employ a conditional denoising diffusion probabilistic model and the Mamba U-shaped convolutional network as the artificial intelligence (AI) methods for implementation. To enhance image quality and precision, we introduce content and style conditioning modules to ensure anatomical and stylistic consistency with real CTA images. Content extractors capture key anatomical features from both CT and CTA images, and frequency-domain processing further enhances vascular structures. The style extractor ensures consistency in appearance with real CTA images. Experimental results demonstrate the superior performance of the proposed AI approach. On clinical private datasets of the neck and abdomen, the structural similarity index measure (SSIM) reached scores of 0. 948 and 0. 906, respectively, indicating the successful generation of high-quality images across multiple anatomical regions. This work represents a significant advancement toward non-invasive, safer, and higher-quality CTA image generation.

Details DOI

AAAI Conference 2025 Conference Paper

Deconfound Semantic Shift and Incompleteness in Incremental Few-shot Semantic Segmentation

  • Yirui Wu
  • Yuhang Xia
  • Hao Li
  • Lixin Yuan
  • Junyang Chen
  • Jun Liu
  • Tong Lu
  • Shaohua Wan

Incremental few-shot semantic segmentation (IFSS) expands segmentation capacity of the trained model to segment new-class images with few samples. However, semantic meanings may shift from background to object class or vice versa during incremental learning. Moreover, new-class samples often lack representative attribute features when the new class greatly differs from the pre-learned old class. In this paper, we propose a causal framework to discuss the cause of semantic shift and incompleteness in IFSS, and we deconfound the revealed causal effects from two aspects. First, we propose a Causal Intervention Module (CIM) to resist semantic shift. CIM progressively and adaptively updates prototypes of old class, and removes the confounder in an intervention manner. Second, a Prototype Refinement Module (PRM) is proposed to complete the missing semantics. In PRM, knowledge gained from the episode learning scheme assists in fusing features of new-class and old-class prototypes. Experiments on both PASCAL-VOC 2012 and ADE20k benchmarks demonstrate the outstanding performance of our method.

PDF Details DOI
v2026.09.13