Minimum rim width better correlates with structural and functional measurements than horizontal rim width

Abstract

 

Purpose

 

Minimum rim width (MRW) - the minimum distance from Bruch’s membrane opening (BMO) to the internal limiting membrane (ILM) - has been shown to be more sensitive than a more traditional horizontal-rim-width (HRW) measure (the horizontal distance from the BMO to the ILM) for the diagnosis of glaucoma. The purpose of this work is to compare MRW and HRW measures when performing structure-structure and structure-function correlations at the level of individual 24-2 visual field test locations.

 

Methods

 

Forty-four optic nerve head SD-OCT volumes from 44 human patients that presented with varying stages of glaucoma were acquired using Cirrus HD-OCT (Carl Zeiss Meditec, Inc.) for purpose of computing MRW and HRW in each of 72 5-degree radial sectors using our automated approach (Fig. 1a). Subjects also underwent Humphrey 24-2 SITA perimetry and a 9-field-per-eye Spectralis SD-OCT imaging protocol. The individual fields were aligned en-face to form a wide field image, covering all the 24-2 testing locations. From the aligned set of images the retinal nerve fiber layer (RNFL) was automatically co-segmented and stitched together. The RNFL was then partitioned into 54 sectors corresponding to the 24-2 grid and the mean layer thickness was computed per sector (Fig. 1b). At each cell of the Humphrey visual field (HVF) grid, the best correlation of VF threshold and RNFL thickness with MRW and HRW across all 44 subjects among the 72 sectors was reported.

 

Results

 

Region-based correlation maps are shown in Fig. 2. The mean ± standard deviation Pearson correlation coefficient of HVF threshold with MRW and HRW were 0.495 ± 0.059 and 0.390 ± 0.069, respectively. Similarly, the mean ± standard deviation Pearson correlation coefficient of RNFL thickness with MRW and HRW were 0.574 ± 0.135 and 0.348 ± 0.059, respectively.

 

Conclusions

 

Overall, the MRW better correlates with structural and functional measurements than HRW, providing further evidence of the value of use of MRW as a structural parameter when compared to HRW.  

 

Fig. 1. BMO-MRW and BMO-HRW measurements (a). The HVF grid on top of NFL thickness map (b).

 

 

Fig. 2. The structure-structure and structure-function correlation results. The HVF threshold values are flipped vertically to be in the same orientation as NFL threshold values (with superior retina on top).