Hanbin Ko

Hanbin Ko

Medical AI Researcher · Ph.D. Student
Seoul National University — Biomedical Engineering
Integrated Major in Innovative Medical Science

Hi, I'm Hanbin

I'm a Ph.D. student at Seoul National University, advised by Prof. Chang Min Park. I build multimodal AI for radiology: vision-language models and text encoders that connect chest X-rays and CT scans with the reports radiologists write. Recently I've worked on CLIP-style pretraining that understands negation and uncertainty, LLM-based report encoders, and learning interval change between follow-up X-rays.

Before my Ph.D., I spent four years as a Medical AI Researcher at VUNO Inc., building clinical AI for radiology and digital pathology.

Multimodal AI Vision-Language Models Representation Learning Chest X-Ray Chest CT Musculoskeletal Radiology Reports Foundation Models

News

2026 📰 LLM encoders for CXR image-text retrieval published in IEEE JBHI.
2026 🎉 Temporal Inversion accepted to CVPR 2026.
2026 🧾 Co-authored "Automated Chest X-ray Report Generation Remains Unsolved" at PSB 2026.
2025 🎉 CREPE, a fast CXR report evaluation metric, accepted to EMNLP 2025 (Main).
2025.02 🎉 Bringing CLIP to the Clinic accepted to CVPR 2025.
2024.09 🎓 Started my Ph.D. at Seoul National University.

Publications

Blue edge = first author
2026
CVPR 2026
Temporal Inversion for Learning Interval Change in Chest X-Rays
Hanbin Ko, Kyungmin Jeon, Doowoong Choi, Chang Min Park
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026
IEEE JBHI 2026
Exploring the Capabilities of Large Language Model Encoders for Image-Text Retrieval in Chest X-rays
Hanbin Ko, R. Yang, Gihun Cho, Inhyeok Baek, Donguk Kim, Joonbeom Koo, Changi Kim, Dongheon Lee, Chang Min Park
IEEE Journal of Biomedical and Health Informatics, 2026
Under Review 2026
Chest2Vec: A Multipurpose Text Encoder Conditioned by Instructions for Chest Radiograph and Computed Tomography Reports
Hanbin Ko, R. Yang, Dongheon Lee, Chang Min Park
PSB 2026
Automated Chest X-ray Report Generation Remains Unsolved
Xiaoman Zhang, Julian Nicolas Acosta, Xiaoli Yang, … , Inhyeok Baek, Hanbin Ko, … , Pranav Rajpurkar
Pacific Symposium on Biocomputing (PSB), 2026
Preprint
Deep Learning–Based Osteoporotic Fracture Risk Assessment from Routine Chest Radiographs: A Multi-Institutional Cohort Study with External Validation
Y. Kim, Hanbin Ko, Dongheon Lee, J.H. Kim, Chang Min Park, S.H. Kong
SSRN preprint
2025
EMNLP 2025 · Main
CREPE: Rapid Chest X-ray Report Evaluation by Predicting Multi-category Error Counts
Gihun Cho, S. Jang, Hanbin Ko, Inhyeok Baek, Chang Min Park
Conference on Empirical Methods in Natural Language Processing (EMNLP), 2025
CVPR 2025
Bringing CLIP to the Clinic: Dynamic Soft Labels and Negation-Aware Learning for Medical Analysis
Hanbin Ko, Chang Min Park
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2025
2020 – 2022
MIDL 2022
PILLET-GAN: Pixel-Level Lesion Traversal Generative Adversarial Network for Pneumonia Localization
H.W. Kim, Hanbin Ko, J.J. Kim
International Conference on Medical Imaging with Deep Learning (MIDL), 2022
ACPR 2021
Joint Dermatological Lesion Classification and Confidence Modeling with Uncertainty Estimation
G.H. Lee, Hanbin Ko, S.W. Lee
Asian Conference on Pattern Recognition (ACPR), 2021
COLING 2020
CoNAN: A Complementary Neighboring-Based Attention Network for Referring Expression Generation
J. Kim, Hanbin Ko, J. Wu
International Conference on Computational Linguistics (COLING), 2020

Current Projects

🫁
Robust Medical Foundation Model
Foundation model for chest & musculoskeletal X-rays and CTs with robust clinical applicability.
🔗
Medical Vision-Language Pretraining
Large-scale pretraining tailored for radiology to enhance imaging-language synergy.
📝
Radiology Report Encoders
Instruction-conditioned text embeddings for chest X-ray and CT reports (Chest2Vec).
📐
Temporal Change Detection
Learning interval changes across sequential medical images for longitudinal analysis.

Education

2024.09 — Present
Ph.D. Student · Seoul National University
Biomedical Engineering + Integrated Major in Innovative Medical Science
2018 — 2020
M.S. · Korea University
Brain and Cognitive Engineering
2013 — 2018
B.S. · Korea University
Biomedical Engineering

Experience

2020 — 2024
Medical AI Researcher · VUNO Inc.
Developed clinical AI solutions for radiology and digital pathology.