This summer, CCADD welcomes two undergraduate interns, Hyunji Park and Dayoon Lee, for the 2026 Summer Internship Program.

Hyunji Park is an undergraduate student in Bio Nano Engineering at Hanyang University, where she is also pursuing interdisciplinary studies in Pharmaceutical Science and Intelligent Robotics. Her academic interests lie at the intersection of life sciences, artificial intelligence, and digital healthcare, with a particular focus on applying data-driven approaches to drug development and clinical research. She has gained interdisciplinary research experience in molecular biology, proteomics, and biomedical data analysis through projects involving DNA cloning, urine exosome-based biomarker discovery for the early diagnosis of preeclampsia, and machine learning-based proteomic data analysis using OmicLearn. These experiences have strengthened her interest in integrating experimental biology with computational approaches and were recognized with research awards from Hanyang University and the Women's Bioscience Forum. Through the CCADD internship, she hopes to contribute to interdisciplinary research while expanding her understanding of AI-driven drug development and computational approaches for evaluating drug efficacy and safety. She aims to further develop her expertise in biomedical data analysis, translational research, and regulatory science, ultimately contributing to safer and more effective drug development.

Dayoon Lee is an undergraduate student at Seoul National University, where she studies Animal Biotechnology alongside an interdisciplinary major in Artificial Intelligence. Her academic interests lie at the intersection of biomedical science and computational research, particularly in understanding how data-driven methods can support safer and more effective drug development. She has gained interdisciplinary research experience in CAR-T cell engineering, toxicology, and medical AI, including a 3D CT-based Vision-Language Model project conducted in collaboration with clinicians. Through the CCADD internship, she hopes to combine her biological with computational approaches and explore how AI and biomedical data can contribute to evaluating drug efficacy and safety. She aims to strengthen her expertise in data-driven drug development and regulatory science through participation in CCADD’s ongoing research projects.

CCADD warmly welcomes the two interns and hopes that the program will provide them with meaningful opportunities to expand their research experience, deepen their understanding of drug development, and pursue their scientific interests with curiosity and independence.

Researchers, clinicians, and engineers working at the crossroads of artificial intelligence and medicine got together once again this summer, and CCADD (Prof. Howard, David, Seeun) was there to join them. The occasion was the Conference on Health, Inference, and Learning (CHIL) 2026, a three-day gathering hosted in Seattle, Washington from June 28th through 30th.

Among a handful of research groups that were extended an invitation to the conference, Prof. Howard traveled together with his team Dr. David and Seeun. What made the trip especially meaningful this year was the publication of a paper from the lab, "Generation of Biingual Synthetic Clinical Notes for Realistic Data Augmentation," which the team unveiled to the community at the conference. Standing beside their poster, they walked visitors through the motivation and findings of the work, answering questions from curious attendees.

This year marked the seventh gathering of CHIL, jointly hosted by Sage Bionetworks and Seattle Children's Research Institute and hosted at the Institute's B. Wayne Hughes Building. As always, the program spanned a broad cross-section of issues shaping AI's role in healthcare. The conference opened with a keynote from Dr. Chris Longhurst, CEO of Seattle Children's, who spoke on "Evidence-based Medicine in the AI Era" and made the case for grounding AI adoption in real clinical implementation rather than hype. Dr. Pang Wei Koh followed with an invited talk on reinforcement learning approaches for deep research, and Dr. Jessilyn Dunn later delivered a keynote on wearable technology for disease detection titled "The Digital Physiome." A track devoted entirely to mental health also drew particular attention, weaving together talks on evaluating harm in AI safety, personalizing psychotherapy through language models, and building governance frameworks informed by practitioners, which were all capped off with a lively panel discussion.

For the team, the days in Seattle offered far more than a chance to share their research. They came away with new perspectives, new connections, and a renewed sense of how much further this field still has to grow.

CCADD, led by Professor Howard Lee, attended the 2026 Spring Conference of the Korean Society of Medical Informatics (KOSMI). The conference was held over three days, from June 24 to June 26, 2026, at OSCO in Cheongju and co-hosted by KOSMI and the Korea Health Industry Development Institute (KHIDI). This year’s theme was “From Operating System to Operating Society: Next Mission of Medical Informatics.” The sessions focused on artificial intelligence (AI) and large language models (LLMs) in medical informatics. Topics ranged from trustworthy medical AI to lifelog analysis and digital healthcare.

David Seung U Lee, a postdoctoral researcher at CCADD, gave an oral presentation in Session 2 (LLM-based Clinical Decision Support) on “DDI-Expert: Explainable Prediction of Pharmacokinetic Drug Interaction in Humans using Mixture of Experts.” His work moves beyond the limits of conventional drug–drug interaction (DDI) prediction. Using a Mixture of Experts (MoE) architecture that combines domain adaptation with domain application, it predicts pharmacokinetic DDI outcomes while also explaining their underlying mechanisms.

Yujin Kim, a student at CCADD, presented in the e-poster session (Generative AI and Clinical Text) on “Information Extraction and Structuring from Drug Approval Review Reports Using Large Language Models.” Her study introduced a pipeline that accurately extracts and structures information from drug approval review reports, which often have complex, multi-table layouts, while keeping costs low and performance high. The work covered the full process from model adaptation to evaluation.

Both presentations were well received at the conference. David Seung U Lee received the Best Oral Presentation Award, and Yujin Kim received the Best Poster Award. The awards mark another step in CCADD’s continued efforts to advance medical informatics through LLMs and AI-based approaches. CCADD will continue its research toward reliable and practical AI applications in healthcare and drug development.

A research article entitled "Design and Evaluation of an Automated Pediatric Acute Lymphoblastic Leukemia Registry from Clinical Data Warehouses" has been published in BMC Medical Informatics and Decision Making (2026). BMC Medical Informatics and Decision Making is a leading international journal that publishes high-impact research on the development, implementation, and evaluation of medical informatics, clinical decision-making systems, and health information technologies. Our study presents a comprehensive framework for constructing a high-quality, scalable pediatric acute lymphoblastic leukemia (ALL) registry through automated extraction of electronic medical record (EMR)-based real-world data (RWD) from clinical data warehouses (CDWs).

The study developed PeARL (Pediatric ALL Registry using eLectronic medical record-based RWD) by integrating standardized mapping, multivariate transformations, and rule-based natural language processing (NLP) to capture clinically relevant variables from two university-affiliated tertiary-care hospitals in South Korea: Seoul National University Hospital (SNUH) and the Catholic Medical Center (CMC). A total of 1,609 pediatric patients with ALL (663 from SNUH and 946 from CMC) were included. Data quality was systematically evaluated using 228 rules across five dimensions (completeness, validity, accuracy, uniqueness, consistency).

The results showed that the automated data extraction proportion reached 89.7% at SNUH and 75.0% at CMC, with most elements processed through single-field transformations. After applying the quality management process, the overall error rate decreased from 1.858% to 0.001% at SNUH and from 0.129% to 0.001% at CMC, corresponding to an estimated reliability of 99%. Moreover, the registry demonstrated robust multicenter applicability through standardized table specifications and cross-site review, despite structural differences between institutional CDWs.

Overall, the study highlights the potential of a clinically guided, standardized framework as a reproducible and scalable solution for registry construction. By combining structured mapping, multivariate transformation, and rule-based NLP, the proposed approach offers a practical pathway for multicenter research and regulatory-grade real-world evidence generation.

A new research article, entitled “Assessing the representativeness of single-center EMR data on ten cancer types: A comparative analysis with national statistics from South Korea (2011–2021)”, authored by Won Jung-Hyun and Howard Lee, was published in the International Journal of Medical Informatics(IJMI). IJMI is a leading international journal that publishes high-impact research on the development, implementation, and real-world evaluation of health information systems and clinical informatics technologies.

This study evaluated whether single-center electronic medical record (EMR) data from Seoul National University Hospital (SNUH) could represent a real-world data source for oncology research in South Korea. To address this, the study analyzed annual incidence and prevalence patterns for ten major cancer types (e.g., breast, gallbladder/biliary tract, gastric, kidney, liver, lung, pancreatic, prostate, thyroid cancers, and leukemia) from 2011 to 2021 using national cancer statistics from the Korean Statistical Information Service (KOSIS) as reference. The study also examined age and gender distributions between two datasets. The results showed that the SNUH EMR data included 8.2% of national incident cancer cases and 10.7% of prevalent cases. The study also found no significant differences in age and gender distributions across all cancer types between the SNUH EMR and national data. In addition, strong correlations in annual trends were observed for several major cancers, including breast, lung, and pancreatic cancers. Overall, this study showed that single-center EMR data can be a valuable, representative clinical resource for oncology research in South Korea. At the same time, the study emphasizes that the underrepresentation of certain cancer types and external temporal factors, such as changes in clinical guidelines or major healthcare disruptions, should be carefully considered when generalizing findings from a single institution to broader populations.

Won J-H, Lee H. Assessing the representativeness of single-center EMR data on ten cancer types: A comparative analysis with national statistics from South Korea (2011–2021). International Journal of Medical Informatics (2026): 106401

CCADD members attended the 2026 Graduate School of Convergence Science and Technology Convergence Academic Festival, held on March 26, 2026, at the Faculty Hall of Seoul National University. This academic festival aimed to promote the exchange of research achievements among the Department of Applied Bioengineering, the Department of Molecular Medicine and Biopharmaceutical Sciences, and the Department of Health Science and Technology, while also providing a forum for presentation and discussion on bio-convergence research. The event brought together faculty members, researchers, and students from diverse fields to share recent research progress. The festival began with opening remarks delivered by Prof. Howard Lee, setting the stage for academic exchange among the participating departments.

More importantly, CCADD members presented one oral presentation and four poster presentations at the festival.

David gave an oral presentation titled “Mechanistic Reasoning Framework for Predicting Multi-Mechanism Pharmacokinetic Drug Interactions”. His work introduced a framework for predicting complex pharmacokinetic drug interactions involving multiple mechanisms.

Yujin’s poster “Annotation-Level Validation and Error Taxonomy for LLM-Assisted Korean–English Pharmaceutical Regulatory Corpus Construction”, focused on validation and error patterns in building a Korean–English pharmaceutical regulatory corpus using LLMs.

In a poster titled “LLM-Based Extraction of Adverse Drug Reactions from Real-World Clinical Notes”, Seoyoon presented an LLM-based approach for extracting adverse drug reaction information from real-world clinical notes.

Suhyun’s poster “Improving Causality Assessment of ADRs in Large Language Models through Rationale-Based Knowledge Distillation”, explored a method for improving the accuracy of causality assessment in LLMs through rationale-based knowledge distillation.

Seeun delivered a poster presentation titled “Generation of Multi-Lingual Synthetic Clinical Notes for Realistic Data Augmentation”, which proposed a framework for generating multilingual synthetic clinical notes to support realistic data augmentation.

Through this academic festival, CCADD members shared their recent research and exchanged ideas with researchers from related fields. The event provided an opportunity to present ongoing work, receive feedback, and engage in discussion on current topics in bio-convergence research.

Join us in congratulating David Seung U Lee on his successful graduation with a PhD in Medical Science!

On February 25, David attended the 80th Commencement Ceremony of Seoul National University (SNU), where he received his PhD degree under the supervision of Prof. Howard Lee. He was also awarded with the Best Dissertation Award in recognition of his outstanding research achievements. His dissertation was entitled “Mechanistic Understanding and Prediction of Drug–Drug Interactions in Humans using LLM-Based Multimodal Explainable Framework”.

Dr. Lee is the first PhD graduate of CCADD, who received his degree in the Department of Molecular Medicine and Biopharmaceutical Sciences at the Graduate School of Convergence Science and Technology, Seoul National University. He will continue his academic career at CCADD as a postdoctoral researcher, focusing on clinical natural language processing and artificial intelligence–based drug development.