Purpose:
To develop methods for topographically comparing local visualfield (VF) defects to local retinal ganglion cell (RGC) andretinal nerve fiber layer (RNFL) thicknesses.
Methods:
Macular and disc 3D cube scans (Topcon, Inc), as well as 24-2and/or 10-2 VFs (Carl Zeiss, Inc), were obtained from healthycontrols and patients with disease of the RGC/optic nerve. Usingan automated segmentation algorithm,[1] RNFL and RGC+inner plexiformlayer (RGC+) thicknesses were obtained. The 6mm x 6mm 3D densityplots for the macula and optic disc were co-registered for boththe RGC+ and RNFL (upper fig: RGC+ for a patient). To accountfor the displacement of the RGC bodies, the location of VF testpoints were adjusted;[2,3] see large squares (24-2 VF) and smallcircles (10-2 VF) in upper fig. Probability plots of abnormallythin RNFL and RGC+ regions, as compared to 128 healthy controleyes, were combined with the abnormal VF points using a commonprobability scale (lower fig.).
Results:
We found: 1. Accounting for the displacement of the RGCs isimportant for VF test points within about ±6° offixation; 2. Arcuate RNFL/RGC defects need not appear as arcuateson VF; 3. Wider field fdOCT scans are needed to optimize fdOCTto VF comparisons; 4. VF points are not optimally positionedfor detecting arcuate defects; 5. The joint probability plotscan help identifying subtle defects.
Conclusions:
Improved methods for comparing local field and fdOCT defectswill be useful in understanding and confirming defects due todiseases of RGC/optic nerve, and for improving VF and OCT tests.1. Yang Q, et al., Opt. Exp. 2010; 2. Hood, Rasa et al, IOVS,2010; 3. Drasdo, Millican et al, VR, 2007.
Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • visual fields • imaging/image analysis: clinical