The usefulness of a new visual field measurement algorithm “smart Strategy Alpha” using the Variational Bayes Linear Regression in comparing with the conventional algorithm in healthy subjects

Abstract

Purpose : We previously reported that a visual field (VF) measurement algorithm, which was named “smart Strategy”, using the Variational Bayes Linear Regression with KOWA AP-7000 (Kowa Company, Ltd., Japan) was useful for saving measurement duration while maintaining the same accuracy as the Swedish Interactive Thresholding Algorithm Standard with Humphrey Field Analyzer (Carl Zeiss Meditec, Dublin, USA) (Br J Ophthalmol in press). We developed a new time-saving algorithm “smart Strategy Alpha”, which was modified the thresholding algorithm from smart Strategy, and assessed its usefulness by comparing with a conventional algorithm “Quick 1” in healthy subjects.

Methods : 22 eyes of 12 healthy subjects were included in the current study. VF measurements were performed using smart Strategy Alpha and the conventional Quick 1 with KOWA AP-7000. The 24-2 tests were conducted in the VF measurements. Measurement duration and visual sensitivities were compared between these algorithms using the linear mixed model where the subjects were treated as random effects.

Results : smart Strategy Alpha measurement duration was significantly 55 % shorter than Quick 1 on average. Mean (standard deviation) measurement duration were 114 (28) seconds for smart Strategy Alpha, and 251 (29) seconds for Quick 1 (p<0.001). On the other hand, mean deviation (MD) values were no significant differences between these algorithms. Mean (standard deviation) MD values were 0.03 (0.92) dB for smart Strategy Alpha, and -0.18 (1.09) dB for Quick 1 (p=0.19).

Conclusions : smart Strategy Alpha saved measurement duration while maintaining the same accuracy as Quick 1 in healthy subjects.

This is a 2021 ARVO Annual Meeting abstract.