Home >

news Help

Publication Information


Title
Japanese: 
English:Cancer Prevention Using Machine Learning, Nudge Theory and Social Impact Bond 
Author
Japanese: 三澤大太郎, 福吉潤, 仙石愼太郎.  
English: Daitaro Misawa, Jun Fukuyoshi, Shintaro Sengoku.  
Language English 
Journal/Book name
Japanese: 
English:International Journal of Environmental Research and Public Health 
Volume, Number, Page Vol. 17    No. 3    790
Published date Jan. 28, 2020 
Publisher
Japanese: 
English:MDPI 
Conference name
Japanese: 
English: 
Conference site
Japanese: 
English: 
Official URL https://www.mdpi.com/1660-4601/17/3/790
 
DOI https://doi.org/10.3390/ijerph17030790
Abstract There have been prior attempts to utilize machine learning to address issues in the medical field, particularly in diagnoses using medical images and developing therapeutic regimens. However, few cases have demonstrated the usefulness of machine learning for enhancing health consciousness of patients or the public in general, which is necessary to cause behavioral changes. This paper describes a novel case wherein the uptake rate for colorectal cancer examinations has significantly increased due to the application of machine learning and nudge theory. The paper also discusses the effectiveness of social impact bonds (SIBs) as a scheme for realizing these applications. During a healthcare SIB project conducted in the city of Hachioji, Tokyo, machine learning, based on historical data obtained from designated periodical health examinations, digitalized medical insurance receipts, and medical examination records for colorectal cancer, was used to deduce segments for whom the examination was recommended. The result revealed that out of the 12,162 people for whom the examination was recommended, 3264 (26.8%) received it, which exceeded the upper expectation limit of the initial plan (19.0%). We conclude that this was a successful case that stimulated discussion on potential further applications of this approach to wider regions and more diseases.

©2007 Institute of Science Tokyo All rights reserved.