Editors Selection IGR 9-1

Computerized image analysis: SAP loss versus OCT thickness

Donald Hood

Comment by Donald Hood on:

16967 The relationship between nerve fiber layer and perimetry measurements, Harwerth RS; Vilupuru AS; Rangaswamy NV et al., Investigative Ophthalmology and Visual Science, 2007; 48: 763-773

Find related abstracts

The new technologies, such as optical coherence tomography (OCT), for in vivo measurement of retinal nerve fiber layer (RNFL) thickness provide a new venue for the old debate about the relationship between structural and functional damage in glaucoma. Understanding this relationship has both practical, as well as theoretical consequences. For example, knowing the precise functional relationship between two tests, if combined with an measure of their intersubject variability, allows one to predict their relative sensitivity.1,2 Harwerth et al. (121) investigated the relationship between RNFL thickness (measured with OCT) and behavioral sensitivity (measured with static automated perimetry (SAP)) in monkeys with experimental glaucoma. This is an important and unique data set. Unfortunately, this study does not show the actual data (i.e., a plot of SAP loss vs. OCT thickness). Instead, each measure is converted to an estimated loss in retinal ganglion cell (RGC) axons. Thus, it is important to understand the assumptions underlying these conversions. Without going into details, it appears that there are at least three key assumptions that are open to debate. First, they assume, based upon earlier work,e.g.,3,4 that the relationship between the log of the number of RGCs and log SAP sensitivity is linear. Others have argued that this relationship is linear on linear, not log, coordinates.1,2,5-8 Second, they assume that when all RGC axons are gone, the RNFL thickness is zero. In a similar study with human SAP and OCT data, we have argued that there is a residual OCT RNFL thickness remaining after essentially all RGC axons have died.9 Third, they assume that in the normal eye, the SAP sensitivity also varies with number of axons and RNFL thickness, an assumption that needs testing. While the fit of their model supplies evidence for their assumptions, their data need to be tested with alternative models. In any case, this is an important study that speaks to an important aspect of the structure vs. function debate.


  1. Hood DC, Greenstein VC, Odel JG, et al. Visual field defects and multifocal visual evoked potentials: evidence for a linear relationship. Arch Ophthalmol 2002; 120: 1672-1681.
  2. Hood DC, Greenstein, VC. The multifocal VEP and ganglion cell damage: applications and limitations for the study of glaucoma. Progress in Retinal and Eye Research 2003; 22: 201-251.
  3. Harwerth RS, Carter-Dawson L, Smith EL III, et al. Neural losses correlated with visual losses in clinical perimetry. Invest Ophthalmol Vis Sci 2004; 45: 3152-3160.
  4. Harwerth RS, Quigley HA. Visual field defects and retinal ganglion cell losses in patients with glaucoma. Arch Ophthalmol 2006; 124: 853-859.
  5. Garway-Heath DF. Comparison of structural and functional methods. In: Weinreb RN, Greve EL, eds. Glaucoma Diagnosis. Structure and Function. The Hague: Kugler Publications 2004:135-143.
  6. Garway-Heath DF, Holder GE, Fitzke FW, Hitchings RA. Relationship between electrophysiological, psychophysical, and anatomical measurements in glaucoma. Invest. Ophthalmol Vis Sci 2002; 43: 2213-2220.
  7. Bowd C, Zangwill LM, Medeiros FA, et al. Structure-function relationships using confocal scanning laser ophthalmoscopy, optical coherence tomography, and scanning laser polarimetry. Invest Ophthalmol Vis Sci 2006; 47: 2889-2895.
  8. Swanson WH, Felius J, Pan F. Perimetric defects and ganglion cell damage: Interpreting linear relations using a two-stage neural model. Invest. Ophthalmol Vis Sci 2004; 45: 466-472.
  9. Hood, DC, Anderson, SC, Wall, M, Kardon, RH. Structure versus function in glaucoma: An application of a linear model. Investigative Ophthalmology and Visual Science 2007 In press

Issue 9-1

Select Issue