Keisuke Ejima, postdoctoral fellow in epidemiology and biostatistics, has received a two-year grant to research body mass index and mortality using artificial intelligence.
The fellowship was awarded based on Ejima’s study proposal titled “Estimate the Association between BMI Trajectory and Mortality Using Artificial Intelligence in the Establishment of Mathematical Obesity Epidemiology.”
“Although a single-time-point body mass index is the most commonly used measurement to assess excess adiposity, the utility in providing clinically meaningful information in the context of obesity treatment is limited. One of the reasons behind it is BMI or weight history has rarely been considered. For example, weight loss can be caused by chronic illness related to obesity. Even though these people lost weight due to the illness and are now in the range of ‘normal weight,’ their mortality risk is much higher than people with stable normal weight,” says Ejima.
The focus of this Japan Society for the Promotion of Science (JSPS)-funded study is to identify weight trajectory patterns and investigate their association with mortality risk. An artificial intelligence model will be employed to classify weight trajectory data into typical patterns. Subsequently, the association between those patterns and mortality risk is examined. Additionally, developing a mathematical model that captures weight trajectory by parameterizing biologically relevant factors, such as energy intake and energy expenditure, is planned. This model will also assist in identifying factors that differentiate weight trajectory patterns.
Ejima will be primarily mentored by Dean Allison but plans to work with the other researchers in the School of Public Health-Bloomington. “Obesity is a complex disease and has been studied from various perspectives, such as public health, medicine, biology and sociology. The Bloomington campus, inside and outside of the School of Public Health, is a unique environment that covers a huge variety of research topics, which makes this campus one of the best places to pursue research of this complex disease,” says Ejima.
“This fellowship program is designed to let Japanese researchers gain research experience in research institutes in other countries. I believe this experience helps me not only to gain knowledge but also to be an international researcher. I appreciate both JSPS and the Indiana University School of Public Health-Bloomington for supporting and offering this opportunity to study in Indiana University, and I’m sure I will learn more than what I expected.”
Having trained originally as a mathematical engineer, Ejima became interested in nutrition/obesity research after the completion of his Ph.D. Recent research includes mathematical model for intergenerational propagation of obesity, obesity and mortality (disability) association, and statistical methods for nutrition/obesity study.
The JSPS was established in 1932 to advance a range of programs essential to the development of talented researchers and to further vital scientific investigation. With only 200 researchers selected from 1,000 applicants each year, Ejima joins a select group of highly capable Japanese researchers given the opportunity to conduct long-term research at a university or institution outside their country to gain a wide international perspective.