Back

Diffusion-weighted Imaging And Retinal Oximetry Predict Functional Outcome After The First Episode Of Optic Neuritis

Hok, P.; Valosek, J.; Svrcinova, T.; Odstrcil, F.; Rybarikova, M.; Kral, M.; Nijhar, K. S.; Arkhipova, A.; Jasenska, M.; Mares, J.; Sin, M.

2025-04-04 ophthalmology
10.1101/2025.04.03.25324536 medRxiv
Show abstract

PurposeTo evaluate diffusion weighted imaging (DWI) with advanced diffusion models, optical coherence tomography (OCT), and automatic retinal oximetry as potential biomarkers for visual deficits after optic neuritis (ON). MethodsTwenty-five patients with acute unilateral ON underwent brain DWI, OCT, and oximetry at the treatment initiation, and 3 and 6 months later. Additionally, 56 healthy volunteers with normal vision were examined once. Clinical outcomes were best-corrected visual acuity (BCVA) and contrast sensitivity. Predictors included retinal nerve fiber layer (RNFL) thickness, arteriolar and venular oxygen saturation (AS, VS), and arterio-venous difference (AVD) in the affected eye as well as DWI parameters from both optic radiations. ResultsMultivariate analysis of variance (MANOVA) for baseline DWI indicated higher secondary partial volume fraction (f2) in patients, while both groups exhibited significant left-right differences for all DWI parameters. Longitudinal analysis in 17 patients with follow-up data revealed a multivariate effect of time when adjusted for affected side and time since onset, however, no DWI parameters changed significantly on a univariate level. Stratified unadjusted model indicated lower overall fractional anisotropy (FA) in patients with incomplete recovery. On uncorrected level, baseline FA and oximetry (VS) were correlated with follow-up BCVA at month 6, while axonal diffusivity (AD) predicted follow-up VS and AVD. In turn, baseline VS and AVD predicted follow-up RNFL thickness. ConclusionsDWI and retinal oximetry are promising early predictors of visual function outcome after ON. Translational RelevanceDWI and retinal oximetry are potentially suitable for patient stratification in studies investigating novel therapeutic interventions.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.