Selected work

Representative Papers

A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation

Aug 17, 2026

This study addresses the challenges of missing and degraded sequences in multiparametric prostate MRI by proposing MSCNet, a cross-modal generation framework. Leveraging sequence-conditioned generation, this method achieves high-quality image reconstruction and was validated through a ten-center blinded reader study. Results demonstrate that reconstructed images attained a structural similarity index (SSIM) of 0.818 and a cancer diagnosis area under the curve (AUC) of 0.841. Notably, diagnostic performance was non-inferior to that of original images and exhibited robust cross-center transferability. These findings confirm the clinical utility of MSCNet as an effective auxiliary diagnostic tool, providing a reliable solution for managing incomplete mp-MRI data in clinical practice.

0 citationsRead paper

DSVM-UNet : Enhancing VM-UNet with Dual Self-distillation for Medical Image Segmentation

Jan 27, 2026

This work addresses the challenge of balancing model performance and computational efficiency in medical image segmentation by proposing a structure-preserving dual self-distillation mechanism built upon the VM-UNet architecture. The method enhances semantic awareness through alignment of global and local feature representations, leveraging Vision Mamba’s strength in efficiently modeling long-range dependencies. Without altering the underlying network structure, the approach achieves state-of-the-art performance on the ISIC2017, ISIC2018, and Synapse datasets while maintaining high computational efficiency, thereby effectively reconciling accuracy and resource consumption.

0 citationsRead paper

Recent publications

Latest Papers

A cross-modal generative model for incomplete and degraded prostate MRI with multicentre clinical validation

Aug 17, 2026

This study addresses the challenges of missing and degraded sequences in multiparametric prostate MRI by proposing MSCNet, a cross-modal generation framework. Leveraging sequence-conditioned generation, this method achieves high-quality image reconstruction and was validated through a ten-center blinded reader study. Results demonstrate that reconstructed images attained a structural similarity index (SSIM) of 0.818 and a cancer diagnosis area under the curve (AUC) of 0.841. Notably, diagnostic performance was non-inferior to that of original images and exhibited robust cross-center transferability. These findings confirm the clinical utility of MSCNet as an effective auxiliary diagnostic tool, providing a reliable solution for managing incomplete mp-MRI data in clinical practice.

0 citationsRead paper

DSVM-UNet : Enhancing VM-UNet with Dual Self-distillation for Medical Image Segmentation

Jan 27, 2026

This work addresses the challenge of balancing model performance and computational efficiency in medical image segmentation by proposing a structure-preserving dual self-distillation mechanism built upon the VM-UNet architecture. The method enhances semantic awareness through alignment of global and local feature representations, leveraging Vision Mamba’s strength in efficiently modeling long-range dependencies. Without altering the underlying network structure, the approach achieves state-of-the-art performance on the ISIC2017, ISIC2018, and Synapse datasets while maintaining high computational efficiency, thereby effectively reconciling accuracy and resource consumption.

0 citationsRead paper