Advancing 3D Generative Foundation Model for Multi-scale Biomedical Images. This project aims to enhance 2D multi-scale biomedical image analysis by leveraging 3D information through advanced generat

Description

Advancing 3D Generative Foundation Model for Multi-scale Biomedical Images. This project aims to enhance 2D multi-scale biomedical image analysis by leveraging 3D information through advanced generative artificial intelligence (AI) algorithms. While 3D biomedical imaging offers detailed insights, its high-cost limits accessibility, making 2D imaging more common. We propose developing a 3D generative foundation model for multi-scale biomedical images (3DGBio), trained on extensive multi-scale 2D-3D images, to generate 3D images from 2D counterparts. Our goals include creating a 2D-3D generative algorithm, fine-tuning it for specific biological scales. This approach will make advanced 3D insights more accessible and practical for various research applications and lead to potential long-term health sector benefits.. Scheme: Discovery Projects. Field: 4603 - Computer Vision and Multimedia Computation. Lead: Prof Jinman Kim