Limitations of imaging techniques
Although these imaging technologies provide vital assessments of structural damage, they have innate limitations. It should be noted that these imaging devices augment clinical assessment, but are not diagnostic on their own. Also, measurement data and classifications should be interpreted in the context of all available clinical data in order to gain an overall view of their accuracy.
CSLO
Confocal scanning laser ophthalmoscopy (CSLO) has a limited ability to detect changes prior to the onset of glaucoma.
The technological limitations of CSLO include:
- Insufficient normative database for population-based screening
- Operator error in placement of the disc margin and the lack of a reproducible standard reference plane that can impact outcome variables
- Measures can be influenced by acute changes in IOP and even cardiac cycle
SLP
- Atypical scans may occur in the elderly, eyes with myopia or a pale fundus
- Corneal compensation may be unreliable in the presence of macular disease
- Ability to detect change needs validation
Scanning laser polarimetry (SLP) and VCC may fail to produce accurate scans in certain patient populations, and has a limited ability to detect changes over time.
The technological limitations of SLP include:
- Inaccuracy of assessment without variable corneal compensation (VCC) as it compensates for variability in the magnitude and axis of corneal birefringence. Results without VCC may be inaccurate, with less correlation with visual function and RNFL thickness assessments.
- Scans can be seen as atypical in patients with macular degeneration. Disruption of the Henle’s layer by macular disease can result in failure of the strategies for neutralization of corneal birefringence required for accurate SLP with VCC.
- Lack of validity for change-detection strategies
- Although such strategies do exist, the required statistical units of change probability are absent. This limits SLP’s ability to differentiate progression from test variability.
- Lack of histological validation of SLP
- Imaging artifact
- Backward compatibility
- Recent versions of GDx are not compatible with previous versions, making it difficult to track long-term progression in patients while upgrading to the latest technology.
OCT
- Experienced operator required for quantitative data acquisition
- Signal strength may impact nerve fiber layer thickness measurement
- RNFL detection algorithm may fail
- Ability to detect change needs validation
- Often requires pupillary dilation
OCT’s ability to detect disease progression is unknown, and it is unable to show change probability.
The technological limitations of OCT include:
- An experienced operator is required for quantitative data acquisition
- Lack of long-term or longitudinal data that show OCT can detect disease progression
- The change analysis software used with optical coherence tomography was only recently introduced. Furthermore, this software does not have any statistical units of change probability, which means that it cannot differentiate biological change from test-retest variability.
- Reproducibility
-
This has been an issue with OCT when comparing points/quadrant scans.
- Need for pupillary dilation to achieve optimal image quality
Normative Data
All the devices require normative reference data. The normative data allow the establishment of normal ranges with respect to various subject characteristics, such as optic disc size (HRT), age, and race. Algorithms compare individual subjects to the normative data.
HRT III, GDx VCC and OCT III all contain age-stratified normative data. The assessment of photographs has no normative database for comparison. Of note, greater subject age is associated with reduced neuroretinal rim area and RNFL thickness measurements.
Instrument Databases
| Device |
Number of Subjects |
Stratified by Ethnic Representation |
| HRT III |
948 |
Yes* |
| GDx VCC |
540 |
No |
| OCT III |
328 |
No |
*African American or European ancestry
|