NLM DIR Seminar Schedule
UPCOMING SEMINARS
-
July 15, 2025 Noam Rotenberg
Cell phenotypes in the biomedical literature: a systematic analysis and the NLM CellLink text mining corpus
RECENT SEMINARS
-
July 3, 2025 Matthew Diller
Using Ontologies to Make Knowledge Computable -
July 1, 2025 Yoshitaka Inoue
Graph-Aware Interpretable Drug Response Prediction and LLM-Driven Multi-Agent Drug-Target Interaction Prediction -
June 10, 2025 Aleksandra Foerster
Interactions at pre-bonding distances and bond formation for open p-shell atoms: a step toward biomolecular interaction modeling using electrostatics -
June 3, 2025 MG Hirsch
Interactions among subclones and immunity controls melanoma progression -
May 29, 2025 Harutyun Sahakyan
In silico evolution of globular protein folds from random sequences
Scheduled Seminars on June 15, 2023
Contact NLMDIRSeminarScheduling@mail.nih.gov with questions about this seminar.
Abstract:
Abstract: Despite the remarkable performance of medical image analysis applications enabled by the recent advances in machine learning techniques, current learning-based models may suffer from poor generalizability since such models are often designed for solving specific problems, or they might not be able to handle the potentially scarce and noisy data in clinical practice. Therefore, properly designed models that incorporate prior knowledge and constraints and robust training schemes are demanded to fill the gap and better aid clinical medicine's diagnosis and prognosis.
In this talk, I will introduce my research on generalizable deep learning models for medical image analysis. In particular, I will mainly illustrate methods I developed on two research threads: 1) how to design more robust medical image segmentation models; 2) how to mitigate the lack of annotated data for training medical image analysis models. Several new models will be discussed, including adversarially learned image segmentation models and selective deep generative models for synthetic image augmentation. While I aim to work towards more robust, reliable, and accessible AI methods, the broader impact of my research is to promote AI-empowered applications in healthcare, clinical medicine, and other areas to benefit a more general population.
Bio: Dr. Yuan Xue is currently a postdoctoral research fellow at Johns Hopkins University working with Prof. Jerry Prince. He will join the Ohio State University as an assistant professor starting in the Fall of 2023. He received his Ph.D. in Information Sciences and Technology from Penn State University under the supervision of Prof. Sharon X. Huang. His research interests lie in computer vision and deep learning, especially with applications in biomedical image analysis using generative models and data efficient learning. He has published more than thirty papers at high-impact venues in the area of computer vision, artificial intelligence, and medical image analysis. He received the MICCAI 2019 best presentation award and the MedIA-MICCAI 2020 best paper award runner-up. He has been organizing the Data Augmentation, Labeling, and Imperfections (DALI) workshops at MICCAI since 2021.
Dr. Xue was invited to present his work by Drs. Zhiyun Xue and Sameer Antani, CHRB, LHNCBC, NLM IRP.