# Early Detection of Dementia Through Spectralis Optical Coherence Tomography in a Taiwanese Cohort

## TL;DR

This study explores how retinal scans using Spectralis OCT can help detect early signs of dementia in a Taiwanese population.

## Contribution

The study evaluates OCT parameters for detecting mild cognitive impairment and mild dementia in an Asian population.

## Key findings

- BMO-MRW nasal inferior region best distinguishes MCI from healthy controls (AUC = 0.720).
- Combining five OCT parameters improves diagnostic accuracy (AUC = 0.861).
- BMO-MRW temporal superior region effectively differentiates mild dementia from healthy controls (AUC = 0.764).

## Abstract

Background and Objectives: Dementia is an essential neurodegenerative disease with pathologic changes in the central nervous system, but also the retina. To evaluate the diagnostic performance of Spectralis optical coherence tomography (OCT) parameters for mild cognitive impairment (MCI) and mild dementia in an Asian population from Taiwan. Methods: This retrospective cross-sectional study evaluated 43 patients with MCI (mean deviation [MD]: −5.05 ± 4.25 dB), 13 patients with mild dementia (MD: −9.03 ± 6.66 dB), and 32 healthy controls (MD: −2.50 ± 2.12 dB). OCT was performed on both eyes of each subject. The diagnostic sensitivity in identifying individuals with cognitive impairment of the Spectralis OCT parameters—such as those of the optic nerve head and macula—was compared across these groups. The area under the receiver operating characteristic curve (AUC) for each parameter was calculated to assess its sensitivity in differentiating between healthy eyes and those of individuals with MCI or mild dementia. Results: Among the parameters evaluated, the Bruch’s membrane opening minimum rim width (BMO-MRW) nasal inferior region (ACU = 0.720) was the optimal parameter for distinguishing individuals with MCI from healthy controls. However, the highest AUC of 0.861 was achieved through a combination of five parameters. In distinguishing individuals with mild dementia from healthy controls, the BMO-MRW temporal superior region (ACU = 0.764) was the optimal parameter, with an AUC of 0.940 after adjusting for age and MD. Moreover, the condition of the macular nerve fiber layer outer inferior parameter moderately predicted disease progression (AUC = 0.713). Conclusions: Our preliminary data demonstrate that Spectralis OCT shows potential in detecting MCI and mild dementia as well as for assessing disease progression in a Taiwanese population. Additional large-scale longitudinal and multiracial studies are essential to validate these findings.