Purchase this article with an account.
Hiroshi Murata, Ryo Asaoka; Validating variational Bayes linear regression model with 10-2. Invest. Ophthalmol. Vis. Sci. 2018;59(9):5118. doi: https://doi.org/.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
We proposed a variational Bayes linear regression (VBLR) model for visual fields (VFs) with 24-2 algorithm (IOVS 2014). The purpose of this study is to validate the model likewise with 10-2.
This retrospective study included VF series from 149 eyes of 110 glaucoma patients as test data, and VF series from 1097 eyes of 811 patients as training data. Using training data, Variational Bayes linear regression was created to predict VF progression.The performance of VBLR was compared against ordinary least-squares linear regression (OLSLR) by predicting VFs in the test dataset. The total deviation (TD) values of test patients’ 11th VFs were predicted using TD values from their 2nd to 10th VFs (VF2-10), the root mean squared error (RMSE) associated with each approach was then calculated. Similarly, mean TD (mTD) of test patients’ 11th VFs was predicted using VBLR and OLSLR, and the absolute prediction errors compared.
RMSE resulting from VBLR averaged 4.7 ± 2.5 (standard deviation) dB and 6.8 ± 3.3 dB for prediction based on the 2nd to 10th VFs (VF2-10) and the 2nd to 4th VFs (VF2-4), respectively. The RMSE resulting from OLSLR was 5.5 ± 2.7 dB (VF2-10) and 29.6 ± 15.8 dB (VF2-4).
VBLR more accurately predicts future VF progression in glaucoma patients compared with conventional OLSLR in 10-2 VFs.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.
This PDF is available to Subscribers Only