Study
PLCO
(Learn more about this study)
Project ID
PLCOI-880
Initial CDAS Request Approval
Apr 25, 2022
Title
Annotation-Ecient Medical Image Analysis
Summary
There is an intense interest in adopting computer-aided diagnosis (CAD) systems empowered by artificial intelligence and deep learning (AI/DL) in biomedical imaging. However, developing such CAD systems is impeded by a significant barrier: the lack of large annotated datasets, because annotating medical images is not only tedious and time-consuming, but it also demands costly, specialty-oriented knowledge and skills, which are not easily accessible. To overcome this barrier, my lab's research objective is to move beyond the limitations of prior CAD systems with an innovative, cost-effective methodology. Specifically, our long-term goal is to exploit novel methods to minimize the requirement for manual labeling efforts to achieve a rapid, precise computer-aided diagnosis of multiple diseases appearing in medical images. Therefore, with the help of PLCO and other publicly available datasets, we propose the following diversed research aims.
Aims
(1) Develop organ-oriented and modality-oriented pre-trained models from unlabeled medical images.
(2) Explore the power of transformer-based models in developing effective and efficient medical imaging systems.
(3) Construct a training strategy that encourages Vision Transformer to understand the semantic features of X-ray images and pair them with its radiologist report.
(4) Developing a generic network with a mixture of various datasets, and exploiting generic knowledge directly from unannotated images
(5) Build (dense) pixel-wise embeddings for chest X-rays from unlabeled images.
(6) Create a comprehensive benchmark suite to evaluate transfer learning from supervised and self-supervised methods for medical imaging.
Collaborators
Jae Shin, Arizona State University
Jiaxuan Pang, Arizona State University
Mohammad Reza Hosseinzadeh Taher, Arizona State University
Fatemeh Haghighi, Arizona State University
Dongao Ma, Arizona State University
Nahid Islam, Arizona State University
Utkarsh Nath, Arizona State University
