Psychophysics behind Visual Field Testing

Question 1

What is the clinical procedure called used to map out Visual Field?

Answer 1

• Clinical procedure used to map out VF known as perimetry
• Perimetry uses light detection task to measure extent & sensitivity of VF
  Larger targets, in general, are more likely to be detected due to better spatial summation of stimulus

Question 2

Describe Target Size in Visual Field testing?

Answer 2

• Target sizes of stimuli used in VF testing follow convention of Goldmann perimetry
• 5 different stimulus sizes are defined by Roman numerals I to V
• Each stimulus area increases 4-fold ranging from 0.25mm^2 for a size one stimulus to 64mm^2 for a size V stimulus
• For Humphrey perimeter, a test stimulus of 4mm^2 which corresponds to Goldmann size III stimulus is most commonly used for standard testing whereas size V is sometimes used for individuals with advanced field loss

Question 3

Describe the Humphrey Visual Field - give decibel values too?

Answer 3

• Luminance unit for background & stimulus in visual field testing is called the Apostilib (asb)
• This is a type of luminance measurement reflected from a surface
• Background luminance of the testing bowl is set at 31.5asb which is bright enough to saturate rod photoreceptors
• Means VF testing is carried out in photopic range within which Weber’s law holds
• Test light intensities runs over 5 orders of magnitude from 10,000 to 0.1asb (quite large)
• According to Weber’s law, we are more sensitive to a relative change in luminance rather than absolute level of luminance
• Decibel (dB) which is 0.1Bel is used to denote 2 numbers that are proportional at ratio of 100.1:1
  o 3dB corresponds approximately to a factor of 2
• Every reduction in light intensity in log order corresponds to 10dB meaning the perimeter can measure sensitivities over a 50dB range
• When range of stimuli intensities provided by a Humphrey machine is 0.1 to 10,000asb  range corresponds to 50 to 0 dB respectively
• Higher dB value of threshold sensitivity for given locus in the VF, the dimmer in luminance, the more sensitive the visual system must be at that point for stimulus to be detected
• Threshold values recorded for each location reflect extent to which light can be dimmed and still detected
  o 30dB indicates stimulus can be dimmed 1000-fold to 10asb from max luminance of 10,000asb but still be detected

Question 4

What is the visual sensitivity map?

Answer 4

• Effective areas of VF present as points in VF printout w/ lower dB values than those in surrounding normal areas
• Numerical values that appear on Humphrey VF are not in unit of luminance but values represent sensitivity in dB

Question 5

What are the 4 methods of threshold determination in visual field testing?

Answer 5

• Yes-no procedure: seen/ not seen
• Full threshold strategy 1:1 up-down staircase
• FASTPAC: methods of limits
• Swedishe Interactive Threshold Algorithm (SITA)

Question 6

Visual Field Testing: Threshold Determination: Describe Yes-no procedure: seen/ not seen?

Answer 6

o Stimulus is presented for 200ms to maximise temporal summation while minimising eye movement that may result in fixation loss
o Depending upon algorithm used stimuli of various intensities can be presented at each of the tested locations
o Goal of the algorithms is to measure intensity of dimmest stimulus that can be detected 50% of time under photopic condition

Question 7

Visual Field Testing: Threshold Determination: Describe Full threshold strategy 1:1 up-down staircase?

Answer 7

o In 1st trial, stimulus brighter than age-corrected normative value is presented at one location
o If stimuli is seen, then stimuli are presented with intensity repeatedly reduced after each trial until stimulus is not seen
o Staircases are randomly interleaved between locations so that multiple locations in VF are tested at same time instead of completing a staircase in one location at a time
o Stimulus intensity changes in steps of 4dB until first reversal occurs & subsequently in steps of 2dB
o With Humphrey analyser, stimulus intensity of last-seen presentation is taken as final threshold estimate after a 2nd reversal at a given location occurs
o Depending upon number of test locations & level of abnormality of px, test can easily take >15mins – time consuming & demanding

Question 8

Visual Field Testing: Threshold Determination: FASTPAC: methods of limits?

Answer 8

o Developed to cut down testing time
o Uses 3dB step size & estimates threshold after only one reversal except for 4 primary seed locations in which standard full threshold testing procedure is carried out
o If thresholds of locations differ from expected values by 4dB or more, then thresholds of those locations are re-estimated
o Considerably faster than conventional standard threshold strategy but can underestimate mean deviation, pattern standard deviation & corrected pattern standard deviation

Question 9

Visual Field Testing: Threshold Determination: Swedish Interactive Threshold Algorithm (SITA)?

Answer 9

o Considered as a modified adaptive staircase where algorithm starts data collection by presenting initial set of stimuli intensities based on a model probability distribution of age-matched, normal sensitivity at 4 primary locations
o Thresholds measured at these locations will be used as a starting intensity at neighbouring locations
o V similar to Psi method – model probability space is assumed & model approaches observer’s distribution on the fly by updating posterior distribution based on response made by observer in each trial
o Testing time reduced in half compared to full procedure w/o sacrificing accuracy & precision
o SITA standard -> SITA fast -> SITA faster