Perimetry knows a wide range of "image types": grayscale, defect map, indices, cumulativ curve... They are always based on 1 of 3 potential basic values. The basic value is decisive for the meaning of a visual field representation.
Threshold values are the untreated measured values of the examination (raw data). The importance is more traditional: In the 1970 decade, threshold values were the only representation option for the young computer perimetry
Threshold values depend on many factors:
- Position of the measured point (normal value is lower in the periphery than in the center, nasal lower than temporal)
- Age of patient
- Instrument type
- Stimulus size
- Background luminance and other parameters
- Strategy (measurement algorithm)
- Damage
- individual variation
- Measurement error
Defect depth values are the deviation of the measured values from the age-corrected normal value for the instrument type at the specific measurement point. In case of double measurement, the average measured value is taken. The formula is:
measured value – normal value
Therefore, defect depth values are already corrected for measurement position, age, instrument, parameter and strategy. The values only depend on:
- Damage
- individual variation (usually less than 1 db shift)
- Measurement error
Pattern deviation is derived from total deviation (defect depth) by shifting all values with a shift value. The formula is:
total deviation (defect depth) - shift value
The shift value is calculated in a way to get the better values in the field being around zero damage pattern is better visible by the shifting, because the diffuse damage is removed from total damage.
NOTE: Pattern deviation values are mostly lower than total deviation values. It is possible to underestimate the defect depth and the damage
NOTE in a follow-up analysis: the shift value is different for each examination. Pattern deviation values may differ even when the actual defect depth is equal - and vice versa
NOTE: Different instrument types use different calculation methods