We are pleased to share that three of Siyeop Yoon’s papers have been accepted at the MICCAI 2024 main conference! MICCAI is one of the most respected conferences in the field of medical image analysis.

1. “Efficient Volumetric Conditional Score-based Residual Diffusion Model for PET Denoising” – First authored
This work aimed to improve 3D PET imaging quality through advanced diffusion models. This work presents a computationally efficient 3D diffusion model for volumetric processing and significantly improves the image quality of PET scans.

2. “Conditional Score-Based Diffusion Model for Cortical Thickness Trajectory Prediction” – Co-first authored with the brilliant Qing Xiao, PhD student.

This study leverages diffusion models in 1D+temporal dimensions to predict cortical thickness changes over time. It addresses challenges with sparse and incomplete longitudinal data and provides accurate stochastic CTh predictions based on baseline information.

3.”Hallucination Index: An Image Quality Metric for Generative Reconstruction Models” Co-authored, (First-author Matthew Tivnan, Instructor, MGH/HMS)

This work introduces a novel metric to assess the quality of images generated by reconstruction models. This is crucial for advancing generative AI in medical imaging, ensuring the reliability and accuracy of generated images.

“These papers highlight the versatility and transformative potential of generative models across different dimensions and applications. A huge thank you to my co-authors, collaborators, and mentors for their invaluable support and contributions. Looking forward to meeting at MICCAI 2024!” – Siyeop Yoon

Congratulation Siyeop, Qing and Matthew! All the great work you have been doing is paying off!

If you are interested in looking at other papers written by Siyeop Yoon, please check out the below: Please also check out these papers below if interested in more of what our lab members have researched:

  1. High-resolution 3D CT synthesis from bidirectional X-ray Images using 3D Diffusion Model
    • Authors: Siyeop Yoon, Jayanth Pratap, Wen-Chih Liu, Matthew Tivnan, Hui Ren, Abhiram Bhashyam, Quanzheng Li, Neal Chen, and Xiang Li
    • Conference: 21st IEEE International Symposium on Biomedical Imaging, Athens, Greece, 27-30 May, 2024
    • Description: This work presents a novel approach for synthesizing high-resolution 3D CT images from bidirectional X-ray images using a 3D diffusion model. This method aims to provide 3D information earlier and facilitating faster treatment planning.
  2. Zero-Shot Novel View Synthesis of Wrist X-Rays using Latent Diffusion Model
    • Authors: Jayanth Pratap, Siyeop Yoon, Wen-Chih Liu, Quanzheng Li, Abhiram Bhashyam, Neal Chen, and Xiang Li
    • Conference: 21st IEEE International Symposium on Biomedical Imaging, Athens, Greece, 27-30 May, 2024
    • Description: This study introduces a latent diffusion model for zero-shot novel view synthesis of wrist X-rays. The proposed model can generate novel X-rays for every specific view and potentially remove the repeated X-ray imaging for fracture diagnosis and treatment.
  3. Realistic Tumor Generation using 3D Conditional Latent Diffusion Model
    • Authors: Rui Hu, Siyeop Yoon, Dufan Wu, Matthew Tivnan, Zhennong Chen, Yuang Wang, Jie Luo, Jianan Cui, Quanzheng Li, Huafeng Liu, Ning Guo
    • Conference: The 2024 Annual Meeting of Society of Nuclear Medicine and Molecular Imaging, Toronto, ON, Canada, 8-11 June, 2024
    • Description: This research focuses on generating realistic 3D tumor images using a conditional latent diffusion model. The generated images are intended to enhance training datasets for various diagnostic and therapeutic applications.
  4. Unsupervised Low-Dose PET Image Reconstruction Based on Pre-trained Denoising Diffusion Probabilistic Prior
    • Authors: Rui Hu, Dufan Wu, Matthew Tivnan, Ning Guo, Siyeop Yoon, Zhennong Chen, Yuang Wang, Jie Luo, Song Xue, Jianan Cui, Huafeng Liu, Quanzheng Li
    • Conference: The 2024 Annual Meeting of Society of Nuclear Medicine and Molecular Imaging, Toronto, ON, Canada, 8-11 June, 2024
    • Description: This paper presents an unsupervised method for low-dose PET image reconstruction utilizing a pre-trained denoising diffusion probabilistic model. The approach aims to reduce radiation exposure while maintaining image quality.
  5. Noise Controlled CT Super-Resolution with Conditional Diffusion Model
    • Authors: Yuang Wang, Siyeop Yoon, Rui Hu, Baihui Yu, Duhgoon Lee, Rajiv Gupta, Li Zhang, Zhiqiang Chen, and Dufan Wu
    • Conference: 8th International Conference on Image Formation in X-Ray Computed Tomography, Bamberg, Germany, 5-9 August, 2024
    • Description: This study introduces a conditional diffusion model for noise-controlled super-resolution of CT images. The model aims to enhance the resolution of CT scans while effectively managing noise, leading to clearer CT imaging.
Siyeop Yoon, 3 papers accepted at the MICCAI 2024 main conference!
Tagged on: