Perimetry
Perimetry maps how well each part of the visual field sees. It is central to diagnosing and following glaucoma, to localising lesions of the visual pathway and to deciding who may drive. This lesson explains kinetic and static testing, takes a Humphrey printout apart zone by zone, and covers the Amsler grid and the Esterman test.
The field as a hill of vision
The visual field is everything an eye can see while it looks steadily at one point. Its outer limits are set by the face: about 60 degrees upwards and towards the nose, about 70 degrees downwards and up to about 100 degrees temporally. The two fields overlap in the centre.
Sensitivity is not even across the field. It is highest at fixation, where cones are most densely packed, and falls away towards the edge. Traquair described this as an island of vision in a sea of blindness, and it is usually drawn as a hill of vision: height is sensitivity, the peak is the fovea and the shoreline is the edge of the field. The blind spot is a pit in the hill about 15 degrees temporal to fixation, where the optic disc has no photoreceptors.
A scotoma is an area of reduced sensitivity surrounded by better vision. It is absolute if nothing is seen there and relative if only dim targets are missed. Disease can also wear down the whole hill (general depression, as with cataract or a small pupil) or erode its edge (constriction). Perimetry measures the shape of the hill.
Kinetic and static perimetry
There are two ways to survey a hill. You can walk in from the sea at a fixed height until you meet the slope, or you can stand at fixed points and measure the height at each.
Kinetic perimetry is the first. A target of fixed size and brightness is moved from the non-seeing periphery towards the centre until the patient reports it. The points where it is first seen are joined to make a contour line called an isopter. Repeating with smaller or dimmer targets draws inner isopters, giving a contour map. The Goldmann perimeter is a white bowl in which an examiner moves a projected light by hand. Targets are named by size (Roman numerals I to V, each four times the area of the last) and brightness, as in III4e.
Static perimetry is the second. The target stays still at preset locations and its brightness is varied to find the dimmest light seen at each one. The Humphrey Field Analyser does this automatically and is the standard instrument. NICE NG81 specifies standard automated perimetry for diagnosing and monitoring glaucoma.
| Kinetic (Goldmann) | Static (Humphrey) | |
|---|---|---|
| Target | Moves; fixed brightness | Stationary; brightness varies |
| Result | Isopters, like contour lines | A sensitivity value at each test point |
| Operator | Skilled examiner throughout | Automated, with supervision |
| Area covered | Whole field to the far periphery | Usually the central 24 or 30 degrees |
| Strengths | Peripheral and neurological defects; patients who cannot manage automated tests | Sensitive to shallow early defects; numerical, repeatable, compared with a normal database |
| Weaknesses | Depends on the examiner; may miss shallow scotomas between isopters | Demands concentration; has a learning effect |
How the Humphrey test works
The patient sits with one eye covered, wearing a near correction, and fixes on a central light inside a dimly lit bowl. White lights of a standard size (Goldmann size III) flash briefly at different points, and the patient presses a button for each one seen. The threshold at a point is the dimmest light seen half the times it is shown.
Sensitivity is reported in decibels (dB). The scale is logarithmic and counts how far the light has been dimmed from the machine's maximum. A high number therefore means a dim light was seen, which is good sensitivity. 0 dB means even the brightest light was missed. Each 10 dB is a tenfold dimming.
The usual test pattern for glaucoma is the 24-2, which tests 54 points in a grid 6 degrees apart out to 24 degrees, with two extra points nasally where early glaucoma often shows. The 30-2 tests 76 points out to 30 degrees. Because the points are 6 degrees apart, a small scotoma close to fixation can fall between them. The 10-2 pattern packs its points into the central 10 degrees and is used for defects near fixation and for macular disease.
Measuring a full threshold at every point takes a long time and tires the patient. SITA (Swedish Interactive Threshold Algorithm) uses a statistical model of normal and glaucomatous fields to home in quickly, cutting the test to a few minutes per eye.
Reading the printout: is it reliable?
Read a Humphrey printout in a fixed order. First confirm the name, date of birth, eye and test pattern. A wrong date of birth compares the patient with the wrong age group. Then look at the reliability indices at the top left. If the test is unreliable, nothing below it can be trusted.
| Index | How it is measured | What a high value means |
|---|---|---|
| Fixation losses | Lights are shown in the blind spot; a response means the eye has moved | The patient is looking around; over 20% is taken as unreliable |
| False positives | The patient presses when no light was shown | A trigger-happy patient; the field looks better than it is, with white patches on the grey scale |
| False negatives | The patient misses a bright light at a point already seen at a dimmer level | Fatigue or inattention; also occurs in real, advanced field loss |
The traditional limits are 20% for fixation losses and 33% for false positives and false negatives. False positives damage the result most, and with the SITA strategies a rate above 15% is flagged as unreliable. Published cut-offs vary. Newer machines track gaze directly and print the record along the bottom of the page.
Reading the printout: the maps and indices
The printout then shows the same data in several forms. Each answers a different question.
- Numerical plot: the raw threshold in dB at every point.
- Grey scale: the same numbers as shading, dark for low sensitivity. It gives a quick impression of the shape of a defect but is not adjusted for age, so it should not be used alone.
- Total deviation: the difference at each point between the patient and the average for healthy people of the same age. A probability plot below marks each point with a symbol that darkens as the result becomes rarer in normal people.
- Pattern deviation: the total deviation after subtracting any overall depression of the hill. It removes the effect of cataract or a small pupil and leaves localised loss.
Comparing the two deviation plots is the key step. If total deviation is abnormal everywhere but pattern deviation is clean, the loss is general and the usual cause is cataract. If the defect is the same in both, the loss is purely focal. If both are abnormal but pattern deviation shows less, there is focal loss under a general depression, for example glaucoma with cataract.
| Index | Meaning |
|---|---|
| Mean deviation (MD) | The average of the total deviation values: how far the whole field lies below normal for age. More negative is worse. It falls with both general and focal loss. |
| Pattern standard deviation (PSD) | How uneven the hill is after removing general depression. It is near zero for a normal or uniformly depressed field and rises with focal defects. It can fall again in very advanced loss, when the whole field is uniformly poor. |
| Visual field index (VFI) | The field as a percentage of normal for age, weighted towards the centre and less affected by cataract. It is used to plot the rate of progression. |
The glaucoma hemifield test uses the anatomy of glaucoma. Nerve fibres arch above and below the fovea and do not cross the horizontal midline in the temporal retina, so glaucoma damages the upper and lower halves of the field unequally. The test compares five zones in the upper field with their mirror images in the lower field. It reports one of five results: within normal limits, borderline, outside normal limits, general reduction of sensitivity, or abnormally high sensitivity. The last usually means a trigger-happy patient.
Glaucoma produces a recognisable set of defects: a paracentral scotoma, a nasal step (a difference across the nasal horizontal midline), an arcuate scotoma, and in late disease a small central island. A field defect is only attributed to glaucoma when it matches the damage seen at the optic disc. Field patterns from lesions of the visual pathway are covered in the neuro-ophthalmology chapter.
The Amsler grid
The Amsler grid is a 10 cm square of small squares with a central dot, held at reading distance, about 33 cm, and viewed with one eye wearing reading glasses. Each 5 mm square covers about 1 degree, so the chart tests the central 10 degrees on each side of fixation. The patient looks at the dot and reports whether any lines are wavy, blurred or missing.
Wavy lines are metamorphopsia, which occurs when fluid or a membrane displaces the photoreceptors so that the image falls on the wrong cells. A missing patch is a central scotoma. The grid is cheap and can be used at home, and its main use is by people with age-related macular degeneration to detect the onset of the wet form early. It is not sensitive, because the brain fills in small gaps, so a normal grid does not exclude macular disease.
The Esterman test and driving
Driving depends on the field of both eyes together, since one eye covers for defects in the other. The Esterman test is a binocular programme on the Humphrey machine. Both eyes are open, the patient wears their usual driving glasses, and bright lights of a single intensity are shown across a wide field. Each point is simply recorded as seen or missed. It is the test the DVLA requires when a field assessment is needed.
For a car or motorcycle licence (Group 1) the DVLA standard is a horizontal field of at least 120 degrees, extending at least 50 degrees to the left and to the right, with no significant defect within 20 degrees of fixation above or below the horizontal meridian. In the central 20 degrees, scattered single missed points or one cluster of up to three adjoining points is acceptable, and a cluster of four or more is not. The chart is accepted only if the false positive score is no more than 20%. Lorry and bus drivers (Group 2) need a wider field of 160 degrees.
Summary
- The visual field is a hill of vision that peaks at the fovea, with the blind spot about 15 degrees temporal to fixation.
- Kinetic perimetry (Goldmann) moves a target of fixed brightness to plot isopters; static perimetry (Humphrey) varies brightness at fixed points to measure thresholds.
- Sensitivity is in decibels: a higher number means a dimmer light was seen, and 0 dB means the brightest light was missed.
- The 24-2 pattern tests 54 points 6 degrees apart; the 10-2 tests the central 10 degrees.
- Check reliability first: fixation losses over 20%, or high false positives or false negatives, make the field untrustworthy.
- Total deviation compares with age-matched normals; pattern deviation removes general depression such as that from cataract.
- Mean deviation measures overall loss and pattern standard deviation measures focal loss.
- The glaucoma hemifield test compares mirror-image zones above and below the horizontal midline.
- The Amsler grid tests the central 10 degrees each side of fixation for metamorphopsia and scotoma in macular disease.
- The DVLA uses the binocular Esterman test: 120 degrees horizontally and no significant defect in the central 20 degrees for Group 1.
Check yourself
Quick recall on this lesson. These are easier than the question bank.
Written for revision. It is not a source of clinical advice, and guidance changes.