/Software_data

Welcome to YGan Lab

Our research agenda aims to develop cutting-edge medical data analytic and human computer interaction techniques to unlock the value of big medical image data, obtain new insights, generate actionable guidance, and facilitate clinical decision making.

Here is a list of projects with data and code source.

Journal manuscript

  • Multi-scale GCN-assisted two-stage network for joint segmentation of retinal layers and discs in peripapillary OCT images, Biomedical Optics Express, 12(4), pp.2204-2220. [code]

    Li, J., Jin, P., Zhu, J., Zou, H., Xu, X., Tang, M., Zhou, M., Gan, Y., He, J., Ling, Y. and Su, Y.

    A novel graph convolutional network (GCN)-assisted two-stage framework is proposed to simultaneously label the nine retinal layers and the optic disc.

  • Inpainting for saturation artifacts in optical coherence tomog-raphy using dictionary-based sparse representation, IEEE Photonics Journal, vol. 13, no. 2, pp. 1–10. [code]

    H. Liu, S. Cao, Y. Ling, and Y. Gan

    In this paper, we propose a novel method to localize and correct saturation artifacts in SD-OCT images. Specifically, we formulate the artifact removal problem as an image inpainting problem and adopt sparse representation framework to solve it.

  • On the spectral leakageand speckle formation in fourier-domain optical coherence tomography, Optics Letter, 2020. [code]

    Y. Ling, M. Wang, X. Yao, Y. Gan, L. Schmetterer, C. Zhou, and Y. Su

    We report on the investigation of spectral leakage’s impact on the reconstruction of Fourier-domain optical coherence tomography (FD-OCT). We discuss the shift-variant nature introduced by the spectral leakage and develop a novel spatial-domain FD-OCT image formation model.

  • Analyzing three-dimensional ultrastructure of human cervical tissue using optical coherence tomography, Biomedical Optics Express, vol. 6, no. 4, pp. 1090–1108, Oct. 2015. [code]

    Y. Gan, W. Yao, K. M. Myers, J. Y. Vink, R. J. Wapner, and C. P. Hendon

    We used optical coherence tomography to investigate the directionality and dispersion of collagen fiber bundles in the human cervix. An image analysis tool has been developed, combining a stitching method with a fiber orientation measurement, to study axially sliced cervix samples.

Conference proceedings​

  • Co-seg: An image segmenta-tion framework against label corruption, IEEE 18th International Symposium on BiomedicalImaging (ISBI), 2021. [data & code]

    Z. Huang, H. Zhang, A. Laine, E. Angelini, C. P. Hendon, and Y. Gan

    In this paper, we propose a novel deep learning framework, namely Co-Seg, to collaboratively train segmentation networks on datasets which include low-quality noisy labels.

  • Hydranet: A multi-branch convolutional neuralnetwork architecture for MRI denoising, SPIE Medical Imaging, 2021. [data & code]

    S. Gregory, H. Cheng, S. Newman, and Y. Gan