EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Characterization of Reaction Time in Standard Automated Perimetry and Its Application to Reliability Assessment

Hiroki Nomoto, Satoru Ueno, Sachiko Okuyama, Chota Matsumoto, Tairo Kimura, Keiji Yoshikawa, Shiro Mizoue, Mami Nanno, Shunji Kusaka

Ophthalmology Science · 2026-11 · 101470

DOI: 10.1016/j.xops.2026.101470 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Ophthalmology Science, 2026-11. Verfasst von Hiroki Nomoto und 8 weiteren Autor:innen. Die Arbeit nennt 23 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Kindai University.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Purpose

To characterize reaction time (RT) behavior during standard automated perimetry (SAP), including its determinants, distributional structure, and patient-level consistency, and to explore its potential to provide complementary information on conventional false-negative (FN) tendency.

Design

Retrospective cross-sectional study.

Participants

1,239 eyes of 657 glaucoma patients were tested using the AIZE algorithm (imo perimeter, 24plus(1–2) pattern); 172,226 responses.

Methods

Per-eye mean RT was analyzed with rank-based generalized estimating equations (age, mean deviation [MD], pattern standard deviation [PSD], sex; patient as cluster). Each eye's RT distribution was decomposed into ex-Gaussian parameters (Gaussian core μ and σ; exponential tail τ), and the mean–standard deviation (SD) relation (Wagenmakers–Brown law) was evaluated. Fellow-eye agreement was quantified with intraclass correlation coefficients (ICCs). A delayed-response rate (DRR) was derived in half of patients (window search, 600–1,200 ms) and validated in the other half against a conventional FN rate ≥33%. Main Outcome Measures Standardized β, ICC, and area under the receiver operating characteristic curve (AUC).

Results

Mean RT (median 550.8 ms) was independently associated with age (+0.78 ms/year), male sex (–13.3 ms), MD, and PSD, and decreased with increasing stimulus intensity relative to local threshold; the within-test SD increased linearly with mean RT (slope 0.27). Male sex was selectively associated with μ, whereas age showed its strongest association with μ and MD with τ. Fellow-eye agreement was high for μ (ICC = 0.90) and mean RT (ICC = 0.91), exceeding that of MD (ICC = 0.63). The DRR (>800 ms), which correlated strongly with τ (ρ = 0.70), discriminated FN ≥33% in the validation cohort (AUC = 0.782, 95% CI 0.701–0.851), showed weaker MD dependence than the conventional FN rate (ρ = –0.40 vs. –0.54), and remained associated with FNs after adjustment for MD, age, and sex (P < 0.001). RT-based false-positive rates did not differ by sex (P = 0.59).

Conclusions

RT exhibits established psychophysical properties and substantial patient-level consistency, with distinct demographic and disease-related influences on its distributional components. Delayed-response behavior provides complementary information on conventional FN tendency using routinely recorded RT data.

Förderung

  • Kindai University

Zitieren

APA

Nomoto, H., Ueno, S., Okuyama, S., Matsumoto, C., Kimura, T., Yoshikawa, K., Mizoue, S., Nanno, M., & Kusaka, S. (2026). Characterization of Reaction Time in Standard Automated Perimetry and Its Application to Reliability Assessment. Ophthalmology Science, 101470. https://doi.org/10.1016/j.xops.2026.101470

Vancouver

Nomoto H, Ueno S, Okuyama S, Matsumoto C, Kimura T, Yoshikawa K, et al. Characterization of Reaction Time in Standard Automated Perimetry and Its Application to Reliability Assessment. Ophthalmology Science. 2026:101470. doi:10.1016/j.xops.2026.101470

BibTeX
@article{Nomoto2026,
  author = {Nomoto, Hiroki and Ueno, Satoru and Okuyama, Sachiko and Matsumoto, Chota and Kimura, Tairo and Yoshikawa, Keiji and Mizoue, Shiro and Nanno, Mami and Kusaka, Shunji},
  title = {Characterization of Reaction Time in Standard Automated Perimetry and Its Application to Reliability Assessment},
  journal = {Ophthalmology Science},
  year = {2026},
  pages = {101470},
  doi = {10.1016/j.xops.2026.101470}
}
Kostenlos starten →