Abstract
Purpose: :
To examine the effect of a spectral-domain (SD) OCT guided reference plane for laser scanning-tomography of the optic disc on diagnostic accuracy in patients with early and advanced glaucomas.
Methods: :
194 eyes of 89 normals, 39 preperimetric and 66 perimetric glaucomas from the Erlanger glaucoma registry (NTC00494923) were imaged simultanously with SD-OCT (Spectralis) and HRT II (both Heidelberg Engineering, Germany). Temporal edge of Elschnig’s scleral rim was visualized with SD-OCT, the distance to the surface measured perpendicularly to disc’s surface (A) and to scleral plane (B), and both values applied to the 0° position of HRT’s contour line creating a new reference plane (rp). Values for HRT’s standard variables (rim and cup area/volume, third moment, globally and in four sectors) with rp A and rp B were compared to HRT’s standard (st) rp 50µm below contour line’s 350-356° or -10° to -4°. Sensitivity (S) at 90% specificity was calculated and Receiver Operating Characteristics (ROC) curves and area under ROC (AUC) were used to determine diagnostic precision using SPSS Version 14.
Results: :
Refractive error (Dpt) was between -5.0 and +5.0, visual acuity above 20/32. Perimetric mean defect (dB, mean, sd) was: controls 0.38 ± 0,8; preperimetric -0.05 ± 1.2; perimetric glaucomas 7.7 ± 6.4. Disc area was not significantly different between diagnostic groups. Reference height was significantly higher with rp A and B than with st.rp(p<0.001). Third moment was not influenced by a new rp. Rim/cup area/volume were significantly different with rp A and B than with st.rp. Rim volume global and in four sectors in preperimetric glaucoma showed higher AUC and sensitivity values with the new rp: rim volume superior AUC 0.66 vs. 0.72, S 38% vs. 46% (p<0.01). Perimetric glaucomas did not show significantly higher AUC and sensitivity values with rp A and B except for rim volume inferior: AUC 0.89 vs 0.91, S 70% vs S 79% (p<0.01). Rp A and B’s accuracy was not significantly different from each other.
Conclusions: :
SD-OCT corrected rp significantly improved the discrimination between normal and glaucomatous discs using HRT’s rim volume. There was no or just a tendency for the other variables in both diagnostic groups. Segmentation of SD-OCT’s disc’s volume scans with Elschnig’s scleral rim as reference plane might be an important tool to increase the significance of HRT II measurements in long term follow-ups.
Keywords: optic disc • imaging/image analysis: clinical • imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound)