Investigation of visual field employing Bayes and machine learning method

2014 Fiscal Year Final Research Report

• Project/Area Number: 25861618
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Ophthalmology
• Research Institution: The University of Tokyo
• Principal Investigator: MURATA Hiroshi 東京大学, 医学部附属病院, 助教 (80635748)
• Co-Investigator(Renkei-kenkyūsha): ASAOKA Ryo 東京大学, 医学部附属病院眼科, 臨床講師 (00362202)
• Project Period (FY): 2013-04-01 – 2015-03-31
• Keywords: 緑内障 / 視野 / ベイズ統計 / 変分ベイズ / 平均場近似

Outline of Final Research Achievements

Glaucoma is characterized by progressive visual field (VF) damage, which is irreversible, so it is important to predict VF progression in clinical use. In Humphrey Field Analyzer (HFA), which is commonly used in clinical settings, employs simple linear regression for analysis of VF data. However, it ignores the correlation between test points and so on.
We created a novel model for VF progression, and by means of learning the model using a lot of data we had, we were able to predict future VF more precisely than simple linear regression.

Free Research Field

緑内障