CCADD attended this year's highly anticipated biennial conference hosted by the Korean Society of Medical Informatics (KOSMI). Held at Keimyung University Dongsan Hospital in Daegu, South Korea, the conference spanned two insightful days, supplemented by a three-hour tutorial program.

Under the theme of "Revolution and Innovation in Smart Healthcare," the event delved into an array of captivating topics. Attendees had the privilege of exploring groundbreaking subjects, including the "Application of Large Language Models in Healthcare" and the "Utilization of ChatGPT and Natural Language Processing Models in Medical Research."

Throughout the conference, CCADD seized every moment to absorb valuable lessons. The gathering also served as a catalyst for expanding perspectives and engaging in constructive discussions at this pivotal juncture in the realm of medicine and healthcare. With powerful technologies like ChatGPT emerging, the field is undoubtedly experiencing a wind of change, and KOSMI's conference provided a unique platform to address the challenges and opportunities presented by these advancements.

 In a gathering of brilliant minds from around the world, Professor Howard and his team attended Conference on Health, Inference, and Learning (CHIL) 2023. The three-day event, held from June 22nd to 24th in Cambridge, Massachusetts, provided a platform for cutting-edge research and discussions on the intersection of artificial intelligence (AI) and healthcare.

Prof. Howard, along with his students Jung Hyun and David, as well as his former student, Siun, were among the select few from South Korea to be invited to this renowned conference. Their attendance was further bolstered by the acceptance of their research paper, titled "Revisiting Machine-Learning Based Drug Repurposing: Drug Indications Are Not a Right Prediction Target," which they presented during the first day's poster session. The team explained the significance and implications of their study, actively engaging with fellow participants and answering questions. More information about the paper can be found here.

The conference encompassed an extensive array of topics, all of which were essential for researchers navigating the rapidly evolving field of AI in healthcare. Attendees were treated to enlightening talks and exciting ideas that spanned innovation, equity, and generalizability in machine learning for health. Discussions included "innovation in machine learning for health," "sharing health data in an age of generative AI: risks, limitations and solutions," "equity implications of machine learning in health: how technology could help or hurt," and "generalizability in machine learning for health: critical for robustness or a distraction from specific validation."

Every moment of the conference proved to be an eye-opening experience. It served as an extraordinary opportunity to broaden their perspectives, push boundaries, and venture beyond their comfort zones.

The Korean Society of Medical Oncology or KSMO is the largest academic congregation in the filed of oncology, consisting not only of medical or clinical oncologists, but of basic and clinical researchers involved in the development of anti-cancer drugs, diagnostics, and vaccines. Prof. Howard Lee was invited to this year's general assembly and symposium for KSMO and discussed the potential and pitfalls of antibody-drug conjugates or ADCs. The first ADC, Mylotarg, was approved in 2000 by the US FDA, followed by its voluntary withdrawal in 2010 due to safety issue, and re-approval in 2017, which just tells how difficult it has been to develop a safe and effective ADC for clinical usage in cancer patients. However, eight ADCs have been approved by the US FDA since 2019 totaling 13, and this indicates how promising the development horizon of ADCs can be. Prof. Lee talked about the pros and cons of ADCs mainly from the perspective of drug development. For example, an ADC consists of three independent components: an antibody (mostly monoclonal antibody or mAb), a cytotoxic payload or warhead, and a linker that combines both antibody and payload. Due to its structural complexity and interactions between elements, which has not been fully elucidated yet, the development of an ADC has faced many technical difficulties. At the same time, however, ADCs also can become a much better platform than other competing ones in the field of oncology because of the uniqueness of its mechanism of action. 

Prof. Howard Lee was invited to a forum organized by the Korea Ministry of Food and Drug Safety (MFDS) to celebrate its 10th anniversary of upgrade from a Ministry of Health and Welfare institution to a Prime Ministerial Office institution 10 years ago. Prof. Lee chaired the panel discussion at the forum. More detailed information about the forum can be found at this.

Prof. Howard Lee has been recently awarded Mid-Career Research Grant from the National Research Foundation (NRF) of South Korea. The total funding amounted to approximately 300 million Korean won spread across three years of research (from year 2023 to 2025). Prof. Howard Lee and his team at CCADD will focus on improving the generalization performance of natural language processing models in extracting drug safety information from multi-institutional clinical notes. More specifically, the team will analyze the semi-structured clinical notes written in diverse clinical settings and investigate the feasibility of using them, along with multi-task generative models, as a way to improve model generalizability. Furthermore, the study will explore the applicability of state-of-the-art generative language models and their ability to augment text data to accomplishing the study objective.

Prof. Howard Lee was invited to an online forum organized by the Federation Peruana de Enfermedades Raras or FEPER on February 28, 2023. In his talk, Prof. Lee summarized the current drug lags for orphan drugs in South Korea, followed by the challenges and issues for delayed and less-satisfactory reimbursement of orphan drugs. Prof. Lee concluded his talk by proposing new policies and systems to increase the accessibility of orphan drugs in South Korea. They included:1) reduce drug lags, particularly approval lags by the regulatory agency, by continuously innovating and harmonizing regulation, 2) ensure orphan drugs be reimbursed as soon as possible, and 3) fill the gap between the legal status of orphan drugs and the reimbursement policy for drugs to treat rare diseases.

The 2023 Winter Internship at CCADD has begun. For this year’s internship, CCADD has Jaeho Hyun, who will study and experience the on-going research projects at CCADD.

Jaeho Hyun is a senior in biomedical engineering at Korea University. He studied mechanical system design engineering at Hongik University until the second year in college before he transferred to Korea University starting his junior year. His motivation for transfer was his interest in finding engineering solutions to solve problems in the medical field. Prior to this internship, he studied artificial intelligence (AI) technology in Seoul Information and communication technology (ICT) innovation square in the summer 2022. The program deepened his interest in applying technical knowledge to addressing real world problems pertaining to healthcare and medicine. During the internship at CCADD, he intends to expand his knowledge and explore various applications of real-world data analysis and AI technology in drug development sciences.

Jung-Hyun Won was invited to the KOREA AI SUMMIT 2022 (“AI for Good”), where she presented the collaboration research project between CCADD (Siun Kim, Jung-Hyun Won, David Seung U Lee, and Prof. Howard) and Microsoft Research Asia (MSRA) entitled “Revisiting and Reformulating Machine Learning-based Drug Repositioning”, orally and as a poster. The Korea AI Summit 2022, organized by AI Innovation Hub, Microsoft Research, IITP, and Korea University AI Research Institute, aimed to provide a platform for academic leaders from the prominent universities and institutes to share and explore the latest developments and technological trends in AI research.

Jung-Hyun Won stated that the team (i.e., CCADD and MSRA) has pointed out problems in the current task formulation of machine learning-based drug repositioning studies. According to Jung-Hyun Won, the team has found the spurious correlation between data collection pattern of drug characteristics and approved indications of drug. They also found that train-test split based on random sampling led to the overestimation of the performances of existing DR models. Jung-Hyun Won stressed that the team is currently reformulating drug repositioning problem as predicting investigational disease and generating eligibility criteria of clinical trials for a drug-disease pair.

Technology showcase can be watched the following youtube thumbnail.

Yesol Hong won the Excellent Research Award at the 30th Annual Conference of the Korean Society for Clinical Pharmacology & Therapeutics (KSCPT), December 1-3, 2022, held at Bexco, Busan, South Korea. The title of YeSol's presentation was "One-year follow-up of patients with COVID-19: Analysis of new-onset disease after COVID-19 diagnosis using Real-World Data". She discussed the results of an analysis of the occurrence of disease in patients with COVID-19 over a year using two different real-world data sets. One of her analysis results was a nation-wide study, which was significant in that it included the largest number of study subjects among the studies analyzing the post-COVID-19-condition in Korea. Jung Hyun Won equally contributed to this study with Professor Howard Lee serving as the principal investigator.

Additionally, Siun Kim, Yoona Choi, David Seung U Lee also presented their research through poster exhibition at the conference. 

Full presentations can be watched at the following YouTube links: