Datathon Tokyo 2023 

After a three-year hiatus due to the COVID-19 pandemic, the Datathon Japan 2023 was finally held in Tokyo, offering a unique opportunity for healthcare professionals, data scientists, and researchers to converge and explore the profound intersection of big data and healthcare. Attendees engaged in a series of hands-on workshops, lectures, and collaborative problem-solving sessions over a three-day period (September 1st-3rd), aiming to address real-life clinical challenges proposed by physicians. With access to extensive health databases from three continents, the guidance of expert mentors, and IT infrastructure support from institutions like MIT and NUS, Datathon-Japan 2023 was a resounding success, fostering innovative solutions and discoveries in the field of data analytics and AI in healthcare.

The first day of the Datathon consisted of a series of engaging lectures and hands-on workshops that showcased the profound impact of artificial intelligence and data science in the field of healthcare. The event kickstarted with eager welcome messages from Dr. Satoru Hashimoto, the Representative Organizer, and Dr. Arinobu Tojo, the Vice President of the Tokyo Medical and Dental University, setting an enthusiastic tone for the day.

 Throughout the morning, participants were treated to a diverse array of lectures, each delivered by renowned experts in their respective fields. Lectures ranged from theoretical topics such as Dr. Leo Anthony Celi's discussion on "Building Capacity in Artificial Intelligence through Datathons" and Dr. Terri Menser’s talk on “AI Ethics and Principles” to more practical applications of AI such as Dr. Luis Nakayama's exploration of "Artificial Intelligence in Ophthalmology " and Dr. Daniel Hashimoto’s presentation on “AI in surgery”. Each presentation not only provided valuable insights into the potential applications of AI and data science in healthcare but also raised crucial ethical and logistical considerations for their real-world implementation.

Following the last lecture by Dr. Keiichi Itatani on medical image data handling from cardiovascular surgery, attendees had the opportunity to delve into interactive hands-on workshops. Participants were able to choose which workshop they wanted to attend, and each catered to learning a specific skill or thought process related to AI in healthcare; workshops included Python programming, analyzing administrative data from Japanese acute care hospitals, and learning about the essential trust and reliability statistics in deep learning models for healthcare. With such a robust schedule of events, the first day of the Datathon conference set a high bar for the days to come, promising to be an enriching experience for all participants as everyone continued to explore the limitless possibilities of AI and data science in the medical field.

The following two days of the Datathon consisted of workshops where each team was assigned to develop a machine learning model that can recalibrate pulse oximeter data in order to prevent failure of detecting hidden hypoxemia. At the end of the final day, teams presented their models and findings, and were given feedback on their model and outcomes.

To find out more about the Datathon and future opportunities, please see here: https://datathon-japan.jp/2023tokyo/