My 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. Our lab develops generative AI tool for image super-resolution and image translation. We develop agentic AI and explainable AI tools for clinical data collection, diagnostic report, and hypothesis generation. We also develop advanced image processing tools (segmentation, object detection, and classification) to enable trustworth image analysis that is robust to data/label corruption using explainable AI
Generative AI for Image Translation
We are developing a generative AI framework to enhance optical and digital resolution of OCT images (NSF-CAREER) and enable virtual staining (NSF-CRII) and animal to human image patch translation (NIH-R01).
[1] X. Li, S. Cao, H. Liu, X. Yao, B. C. Brott, S. H. Litovsky, X. Song, Y. Ling, and Y. Gan, “Multi-scalereconstruction of undersampled spectral-spatial oct data for coronary imaging using deep learning,”IEEE Transactions on Biomedical Engineering, vol. 69, no. 12, pp. 3667–3677, 2022
[2]X. Li, H. Liu, X. Song, C. C. Marboe, B. C. Brott, S. H. Litovsky, and Y. Gan, “Structural constrained and pathology aware convolutional transformer-GAN for virtual histology staining of human coronary OCT images,” Journal of Biomedical Optics, vol. 29, no. 3, p. 036 004, 2024.
[3]J. Kang, X. Li, M. Attia, C. Gouya, and Y. Gan, “Multi-source fundus photography image quality assessment in retinopathy of prematurity using deep learning,” in ARVO Annual Meeting, Seattle, WA, May 2024
Agentic AI
We are developing foundation models for medical information collection, scence recognition (NIH-R21), and scentific discovery (Scialog Collaborative Award), enabling explainable AI solutions in clinics, behavioral study, and chemistry.
[1] X. Li, X. Hou, N Ravi, Z Huang, and Y. Gan, "A Two-Stage Proactive Dialogue Generator for Efficient Clinical Information Collection Using Large Language Model" , Expert System with Applications, 2025
[2]X. Hou, S. Shen, X. Li, X. Gao, Z. Huang, S. Holiday, M. Cribbet, S. White, E. Sazonov, Y. Gan, “Screen Detection from Egocentric Image Streams Leveraging Multi-View Vision Language Model” IEEE Trans Multimedia, 2025.
[3]X. Hou, X. Li, N. Ravi, Z. Huang, and Y. Gan, “Reinforcement-learning-based proactive medical dialogue system for health status and medical image collection,” in Medical Imaging 2025: Image Processing, 2025.
Intelligent System for Rehabilitation
We are developing intelligent robot system self-learning based treadmill control and hologram-enhanced virtual reality for lower body rehabilitation (NSF-PFI).
[1] S. Cao, J. Zhao, F. Hu, and Y. Gan, “Real-time, free-viewpoint holographic patient rendering fortelerehabilitation via a single camera: A data-driven approach with 3d gaussian splatting for real-world adaptation,” IEEE Transactions on Visualization and Computer Graphics, 2025
[2]S. Cao, M. Ko, C. Li, D. Brown, X. Wang, F. Hu, and Y. Gan, “Single-belt vs. split-belt: Intelligenttreadmill control via micro-phase gait capture for post-stroke rehabilitation,” IEEE Transactions onHuman-Machine Systems, vol. 53, no. 6, pp. 1006–1016, 2023.
[3]F. Hu, Y. Zuo, and Y. Gan, “Robotic upper trunk support device,” pat., US Patent App. 63/549,236, 2024.
Robust Image Informatics Tool
We are developing customized image processing tool to extract critical cell informations for food science (USDA/NIFA-Data Science), cardiac imaging, tissue engineering, and cellular/molecular biology (Burroughs Wellcome Fund). We address the issue of network uncertainty, label corruption, and few shot learning with the aim to develop a trustworthy solution for image informatics.
[1] M. Ahmed, H. Mustafa, M. Wu, M. Babaei, L. Kong, N. Jeong, and Y. Gan, “Few shot learning for avocado maturity determination from microwave images,” Journal of Agriculture and Food Research, vol. 15, p. 100 977
[2]Z. Huang, Y. Gan, T. Lye, Y. Liu, H. Zhang, A. Laine, E. Angelini, and C. Hendon, “Cardiac adiposetissue segmentation via image-level annotations,” IEEE Journal of Biomedical and Health Informatics,vol. 27, no. 6, pp. 2932–2943, 2023
[3]Z. Huang, Y. Gan, T. Lye, H. Zhang, A. Laine, E. D. Angelini, and C. Hendon, “Heterogeneity mea-surement of cardiac tissues leveraging uncertainty information from image segmentation,” in MedicalImage Computing and Computer Assisted Intervention – MICCAI 2020
[4] Mahdi Babaei, Aaron Shamouil, J. Wang, D. Khare, T. Wang, M. Shih, X. Yu, and Y. Gan, “Automated cell properties toolbox from 3d bioprinted hydrogel scaffolds via deep learning and optical coherence tomography,” Biomed. Opt. Express, vol. 16, no. 5, pp. 2061–2076, 2025.
[5]B. Miao, Z. Hu, R. Mezzadra, L. Hoeijmakers, A. Fauster, S. Du, Z. Yang, M. Sator-Schmitt, H. Engel,X. Li, C. Broderick, ..., Y. Gan, ..., and C. Sun, “CMTM6 shapes antitumor t cell response throughmodulating protein expression of CD58 and PD-L1,” Cancer Cell, 2023
Other Biomedical Application
Image denoising
We developed denoising framework to denoise MRI images and ultrasound images.
Object detections
We developed region proposal network to identify diseased region in human coronary and football player in sports data analysis.
Cardiac characterization
Cervical collagen fiber image analysis and image informatics to better understanding of preterm birth
Breast cancer identification for surgical margin detection