Automated Perimetry: Visual Field Deficits in Glaucoma and Beyond

Evaluating the visual pathway can be integral to diagnosing and managing numerous conditions.

By Kristie Draskovic, OD, and John J. McSoley, OD

Release Date:

March 2016

Expiration Date:

March 1, 2019

Goal Statement:

Examining the visual fields using automated perimetry can help clinicians evaluate lesions that affect the visual pathway, establish baselines and screen for certain medication-induced optic neuropathies. It can also help monitor progression or recurrence of diseases, guide treatment decisions and aid in localizing lesions. This article will help clinicians better understand and interpret visual fields in glaucomatous and non-glaucomatous disease processes.

Faculty/Editorial Board:

Kristie Draskovic, OD, and John J. McSoley, OD

Credit Statement:

This course is COPE approved for 2 hours of CE credit. COPE ID is 48599-GL. Please check your state licensing board to see if this approval counts toward your CE requirement for relicensure.

Joint-Sponsorship Statement:

This continuing education course is joint-sponsored by the Pennsylvania College of Optometry.

Disclosure Statement:

Drs. Draskovic and McSoley have no financial relationships to disclose.

Examining the visual field is essential when considering potential vision loss from conditions affecting the visual pathway. Standard automated perimetry is a useful tool for identifying and following many neurological conditions, as well as glaucoma and glaucoma suspects. While several devices are currently available, the following discussion will consider the use of the Humphrey Field Analyzer (HFA, Zeiss) as an example. 

Anatomy and Physiology
Due to the anatomy of the visual pathway, clinicians can detect areas of concern with many conditions that cause visual impairment (Figure 1). Because light traveling from the temporal visual field falls on the nasal retina, retinal lesions are seen in the exact opposite quadrant of the visual field. All pre-chiasmal lesions, including those on the retina and optic nerve, will give rise to defects isolated to the affected eye.1

Since nasal fibers are responsible for the temporal visual field, a lesion at the optic chiasm will result in a visual field defect that affect the temporal field of both eyes, and gives rise to the classic bitemporal hemianopsia, which respects the vertical midline.1 Lesions of the optic chiasm can be the result of pituitary adenomas, suprasellar meningiomas, craniopharyngiomas or aneurysms. In some cases, visual field defects may reverse after treating the cause.2

Fig. 1. A simplified schematic of the visual pathway is shown.
Click image to enlarge. Fig. 1. A simplified schematic of the visual pathway is shown. Lesions which interrupt the visual pathway lead to visual field defects. (1) A complete right optic nerve lesion results in complete loss of the right visual field. (2) A lesion in the midline of the optic chiasm leads to a bitemporal hemianopsia. (3) A lesion of the uncrossed fibers of the right optic nerve at the optic chiasm leads to a nasal hemianopsia of the right eye. (4) A complete lesion of the right optic tract, lateral geniculate nucleus, or optic radiations results in a complete left homonymous hemianopsia. (5) A lesion of the right upper optic radiations results in a left inferior quadranopsia. (6) A lesion at the right lower optic radiations causes a left superior quadranopsia. (7) A lesion of both the superior and inferior right optic radiations causes a left homonymous hemianopsia. Illustration by Miquel Perello Nieto.

Fibers responsible for the visual field to the right of the midline are found on the left side of the brain, and vice versa. As a result of this crossover, all post-chiasmal lesions, including lesions of the optic tract and optic radiations, cause a homonymous hemianopsia.1 These defects are on the same side of the visual field in each eye and respect the vertical midline. When the defects are only seen superior or inferior, it is referred to as a quadranopsia. In the case of incomplete hemianopsia defects, anterior lesions are usually more incongruous, whereas posterior lesions will be more congruous between the two eyes.1

Glaucomatous visual field loss represents damage to the axons traveling along the retinal nerve fiber layer and usually follows an arcuate pattern to the optic nerve. Damage to these axons will give rise to localized visual field defects, most commonly arcuate scotomas, nasal steps and paracentral scotomas. A normal field of vision extends farthest temporally to 90 degrees, 70 degrees both superiorly and inferiorly and 60 degrees nasally and from fixation. The most valuable information for neurological deficits and glaucoma management is obtained within 30 degrees from fixation.2

Test Strategy
Automated static perimetry—and threshold perimetry specifically—presents a stimulus of fixed size but of variable intensity. The sensitivity of the different test locations is recorded based on the patient's responses to these stimuli. A size III stimulus (which is 4mm2 when projected on a 30cm bowl) is commonly used in clinical practice.

There are different threshold test patterns available. The 30-2 test pattern tests 76 locations within 30 degrees from fixation, while a 24-2 test pattern tests 54 locations removing for the ring of points at 30 degrees (except the two points which straddle the horizontal meridian nasally). These most peripheral points are more prone to variability, and 24-2 shortens test time by removing these peripheral points from the test. Spacing between test point locations is six degrees. 

In addition, the 10-2 test pattern targets 64 points within 10 degrees from fixation separated by two degrees. This option is preferable when central or paracentral defects arise with the 24-2 and 30-2 or when the field becomes so constricted the peripheral points are not clinically useful. Clinicians can also choose to increase the stimulus size to V, which may afford higher sensitivity values and a larger dynamic range through which to follow patients for change. However, currently there is no SITA type test algorithm, nor is there an available comparison to a normative database or progression analysis.

Case 1. Pituitary Macroadenoma

A patient presented for consultation in the setting of atypical, progressive normal tension glaucoma.
The T1 sagittal without gadolinium and T1 coronal views with gadolinium.
The patient's 24-2 visual field demonstrates bitemporal visual field loss.
Above, the patient's 24-2 visual field demonstrates bitemporal visual field loss.

In the MRIs at left, the T1 sagittal without gadolinium and T1 coronal views with gadolinium reveal a large sellar mass with slight enhancement and compressive effect on chiasm. This finding is consistent with pituitary macroadenoma.

Click images to enlarge.

After the patient underwent transsphenoidal surgery to remove the sellar mass.
The ganglion cell complex reveals nasal thinning in both eyes.
After the patient underwent transsphenoidal surgery to remove the sellar mass, the visual field, above, shows improvement of the visual field defect in the right eye and resolution of the defect in the left eye. Click image to enlarge. 

Above, the ganglion cell complex reveals nasal thinning in both eyes, which correlates to the temporal visual field loss. Click image to enlarge.

Recent studies show that 10-2 visual fields may be useful for identifying more than advanced glaucoma.3,4 One study revealed that some glaucoma patients had significant central cluster defects seen on 10-2 testing patterns despite normal central 30-2 visual field.4 The study suggests that, due to poor spatial sampling, glaucomatous central field deterioration could not be picked up by the 30-2 test grid alone, and denser estimation of the central 10 degrees was required.4 It also suggests that modifying the conventional visual field test pattern might improve detection of early glaucomatous defects in the central 10 degrees.4 

In daily practice, doctors often perform 10-2 fields when small central or paracentral scotomas appear on 30-2 and 24-2 testing, or when acuity is suspected or threatened to be depressed as a result of field loss. Also, a 10-2 test pattern may be a useful adjunct when there is high inter-test variability of paracentral points on a 24-2 pattern. The growing attention to ganglion cell function in the macular region could also increase attention to the area of the visual field tested by the 10-2 pattern.

Indications of reliability of visual field performance include fixation losses and both false positive and false negative responses. 

Table 1. Optic Neuropathies Commonly Detected with Visual Field Testing

Idiopathic intracranial hypertension Early: enlarged blind spot
Late: generalized constriction5,6
(can improve with treatment)
0ptic neuritis - Diffuse visual field loss (in almost half of cases)
- Other: altitudinal defect, central or cecocentral scotomas, arcuate or double arcuate defects and hemianopic defects5,7
Non-arteritic anterior ischemic optic neuropathy - Altitudinal defects that respect the horizontal midline are most common
- Other: central scotomas, arcuate defects and quadranopsias5,8
Posterior ischemic optic neuropathy - Central field defect5,9
Hereditary optic neuropathies
 - Leber's hereditary optic neuropathy
 - Dominant optic atrophy
- Cecocentral and central visual field loss5
Optic nerve head drusen - May mimic a glaucomatous pattern
Thyroid ophthalmopathy - Large variability
- May partially or fully resolve after treatment2
Medication-induced toxic optic neuropathy - Ethambutol (for tuberculosis treatment) toxicity may cause central scotomas and, less commonly, peripheral constriction and altitudinal defects5
- Vigabatrin (an anti-epileptic drug), may cause field defects that begin as bilateral nasal defects and later progress to concentric field defects while the central field remains intact12

Fixation can be monitored by periodic presentation of a stimulus to the physiologic blindspot (Heijl-Krakau method) or by monitoring the position of the corneal light reflex. On the HFA, upward deflection indicates change in position or fixation; downward deflection indicates the corneal light reflex cannot be located, such as during a change in head or eyelid position. 

False positives occur when a patient responds at a time when there is no associated stimulus or when a response is physiologically not possible. Patients with high false positives are often described as "trigger happy." This can reveal a visual field that appears more sensitive or more normal than expected or can lead to abnormally high threshold sensitivity values. 

False negatives occur when a patient fails to respond to a stimulus brighter than one already seen or when the response is not consistent with the pattern of responses in that region. The false negative value may be an indication of reliability or a reflection of the disease process. Abnormal regions of the visual field are associated with more intra-test and inter-test variability. As a result, regions of the visual field with low sensitivity are not included in the calculation of the false negative value. 

The threshold sensitivity is raw data with values recorded in decibels. The numbers indicate the degree of attenuation from maximum possible stimulus. Values of < 0 response indicate that the patient did not respond to the brightest stimulus available. The grayscale uses shaded symbols to describe the level of sensitivity, with the darker areas on the map corresponding to more reduced sensitivity. 

The total deviation represents the deviation from expected values based on the age-matched normal database. The pattern deviation corrects for overall sensitivity of field by removing generalized depressions (e.g., from cataracts) to identify any areas of localized abnormalities. Both total and pattern deviation have associated probability (p) values calculated based on distribution within a normal population. The range of normal values is wider peripherally than centrally. P symbols indicate the frequency of the tested value occurring in a normal age-matched population. Deviations are shown on the map if the tested threshold is worse than the bottom 5% of normal for that age.2 For example, if p < 0.5%, then fewer than 0.5% of reliable normal fields had a sensitivity value at a given point less than or equal to that recorded threshold.

Case 2. Multiple Sclerosis

A patient diagnosed with multiple sclerosis (MS) presented with vision loss.

The visual field demonstrates an incomplete right homonymous hemianopia.
Above, in the MRI, T2 flair axial view reveals a focal hyperintense lesion just posterior to the left lateral geniculate nucleus in the thalamus, which is responsible for the visual field defect seen. The lesion seen, as well other multiple hyperintense lesions, are typical characteristics of demyelinative disease such as MS.
The ganglion cell analysis reveals nasal thinning in the right eye and temporal thinning in the left eye.
Above, the visual field demonstrates an incomplete right homonymous hemianopia. Click image to enlarge.

At left, the ganglion cell analysis reveals nasal thinning in the right eye and temporal thinning in the left eye, which is consistent with the visual field findings. Click image to enlarge.
Pattern standard deviation (PSD) is a weighted standard deviation from the value expected based on age-matched normal values and the mean deviation (MD) index. The PSD reveals the distance between different points within the field. PSD is sensitive for both early and focal glaucoma defects. A low PSD value can be seen with normal visual fields, a visual field that is uniformly depressed or a completely blind field. An associated p value is the likelihood of the value occurring.

The MD is a weighted average of values in the total deviation numerical plot. MD may be influenced by a diffuse decrease in overall sensitivity or by a localized defect. An MD of 0 indicates a normal value, while a negative value represents deviation or loss from the normal database. Mean deviation weighs central points more heavily. 

Visual field index (VFI) is another age-corrected assessment, expressed as a percentage, where perimetrically normal is 100% and perimetrically blind is 0%. When calculating VFI, central points are weighted stronger than peripheral points. It provides a score for an individual visual field and is used in progression analysis.

The glaucoma hemifield test (GHT) compares the relative sensitivity of pattern deviation values of five zones in the superior and inferior hemifields. Each of these zones is compared with its mirror zone in the opposite hemifield, and both zones are compared with the normative database. GHT can detect glaucomatous visual field loss with high sensitivity and high specificity through five possible categories: outside normal limits, borderline, general reduction of sensitivity, abnormally high sensitivity and within normal limits. 

Establishing a reliable baseline visual field is crucial in glaucoma management and future monitoring for possible progression. Clinicians should obtain at least two reproducible baseline visual fields to detect smaller increments of change. Recognizing progression, particularly after escalation of therapy at times of progression, will require the clinician to establish a new baseline. 

 Case 3. Paracentral Glaucoma Defect

An example of variable points in superior paracentral scotoma on 24-2 with a focal deep defect on 10-2.

Visual fields done prior to the current medical or surgical treatment, or during an extended period of being lost to follow up off treatment, should not be used for the purpose of determining progression. Clinicians should use the new baseline visual fields performed after these treatment changes, and any visual field thereafter in judging progression.  

A reliable and representative baseline is also important for progression analysis programs. Visual fields tend to have higher variability with more eccentric location and mid-range sensitivities, and long-term fluctuation makes judging progression more challenging. Mid-level sensitivity points tend to be variable, whereas test values in upper and lower sensitivity ranges are more stable. Establishing a reliable, reproducible baseline visual field can help remove some of these challenges. 

Non-glaucomatous Optic Neuropathies
Visual fields can help evaluate lesions that affect the visual pathway, establish baselines and act as screening tools for certain drugs associated with optic neuropathies. They can also monitor progression or recurrence of diseases while helping guide treatment decisions and even aid in localizing lesions.5 The most common visual field manifestation of a non-glaucomatous optic neuropathy is a central defect. The larger the scotoma, the greater the likelihood that visual acuity is reduced (Table 1). 

Retinal Disease
Visual field defects caused by retinal disease tend to have well-defined, sharp borders that appear deeper than most glaucoma defects with less variability.2 Common conditions affecting the macular area, such as age-related macular degeneration and central serous chorioretinopathy, can give rise to central scotomas which can be seen on 10-2 as well as some 24-2 and 30-2 testing, depending on the extent of damage. 

Retinitis pigmentosa gives rise to peripheral visual field defects, which may lead to tunnel vision with only a central island in advanced disease. 

Arterial occlusions typically show absolute areas of defect whereas venous occlusions can be more shallow and diffuse.2 These often coexist with glaucoma, making interpretation more challenging. 

Chorioretinitis, or associated scars, can mask glaucomatous defects with an arcuate or wedge-like defect. A retinal detachment will give rise to a relative defect while retinoschisis will have an absolute defect. These conditions are often much more peripheral than what would appear on conventional central testing using 24-2 or 30-2. 

Glaucoma Defects
A defect can be a generalized or localized depression of sensitivity compared to the normalized database. General depressions are seen most commonly with cataracts, but can also arise from uncorrected refractive error or miosis. Localized defects can be further described by size, depth and location to help with diagnosis. A relative defect occurs when sensitivity is less than normal, or may be reduced relative to other areas of the field, but vision remains. An absolute defect is when the stimulus is presented at maximum brightness and not seen. 

Glaucomatous visual field loss may first occur in the nasal or in the arcuate region (Bjerrum area). These defects may extend from the blind spot, around the macular region, ending abruptly at the horizontal meridian nasally. Early glaucomatous defects are often localized relative scotomas (Table 2). 

Considerable test-retest variability is the hallmark of visual field areas affected by glaucomatous visual field loss. Variable sensitivity reductions occurring in the same area, but not always in the same test point locations, commonly precede consistent glaucomatous field defects.2 Variability, as is seen with false negatives, may indicate reliability, but larger amounts of variability are often seen due to the disease itself.

Table 2. Criteria for Glaucomatous Visual Field Defect with High Specificity and Sensitivity13,14 

  • Cluster of three or more non-edge points on the pattern deviation with p < 5% with one point at p < 1% over two consecutive fields
  • GHT outside normal limits on two consecutive fields
  • PSD < 5% over two consecutive fields
  • Asymmetry of MD >1.50dB that is repeatable and in the setting of clinical suspicion15,16

When visual fields are unreliable, they should be repeated to establish baselines, confirm a defect or confirm suspected progression. Usually, visual fields are not repeated on the same day due to patient fatigue, which can affect reliability. It is rarely urgent to have the visual field repeated immediately, as glaucoma is generally a slowly progressive optic neuropathy, and treatment decisions are driven by rate of change.2 Each practice should standardize a preferred test strategy and pattern and repeat the same test to enable more accurate comparisons on follow up testing.10

Evaluating change allows the practitioner to determine if the condition is stable, progressing or improving. Progression in a visual field may be due to a diffuse decrease in sensitivity, existing defects may get deeper or expand or new defects may arise. Depressed areas most commonly progress before new areas of visual field are affected. An initial increase in visual field variability is sometimes seen before a change or progression becomes constant. It is important to differentiate if evidence shows long-term fluctuation or a worsening trend. Once clinicians establish that the visual field is worse, they must decide if the change is due to glaucoma or another disease entity. Statistically and clinically significant changes on the visual field allow the practitioner to make the necessary changes to the patient's treatment and management plan. 

Practitioners must also evaluate how the rate of progression may alter the patient's quality of life.2 For example, clinicians must be more cautious treating younger patients with faster progression and monocular patients (Table 3). 

While it is difficult to predict which patients will progress slowly vs. rapidly, once a rate of progression is evaluated over time, the patient's management should be altered accordingly. The younger patient who shows signs of rapid progression of visual field defects, for example, will need a more aggressive management plan than an elderly patient with slow progression of an early field defect. It is reasonable that patients' treatment and management be personalized based on their specific clinical picture, including their field loss and rate of progression. Research shows if rate of progression is determined and no treatment change is initiated, past rates of progression can be predictive of future rates.10,11 At the same time, rates of decline can be altered by escalating therapy. Once an intervention is made, a new baseline should be established. 

Table 3. Suggestions for Judging Progression

  • Three points in an abnormal region decrease by 10dB*
  • Two new points, near a defect reduced by 10dB*


  • Two points in central 15 degrees or three outside 15 degrees down by 10dB*


  • Statistical comparisons (p<5%)*

*On two or more consecutive fields14,17

Evaluating change over time is critical in glaucoma management. With the guided progression analysis (GPA) software found on HFA, two baseline visual field tests are identified. Subsequent visual field tests are then compared with the averaged baseline using the pattern deviation values. When follow-up values decline to a degree larger than the variability of an age and defect matched population of stable glaucoma patients, the point is identified. If the change persists on consecutive repeat testing, points are marked as possibly (two consecutive) or likely (three or more consecutive fields).  

GPA uses both trend and event analysis to aid practitioners in identifying and quantifying progression of the visual field. Event analysis looks for a statistically significant change of a point or group of points, while trend analysis quantifies the direction of change over time, or the rate of change, including future projections. Clinicians can direct their attention to the potential underestimation of diffuse loss, subtle artifacts that may be associated with pattern deviation and the need for a sufficient number of high quality visual fields (minimum five) for optimal analysis. However, GPA is less influenced by cataract than other analysis tools. Each test strategy uses its own normative database.2 An additional goal is to identify a rate of progression and separate out the patients who are progressing rapidly and need increasingly aggressive therapy. 

Progression of Non-glaucomatous Optic Neuropathies
Because GPA was established specifically for glaucoma management, clinicians must take a different approach when evaluating for possible progression in non-glaucomatous conditions. Regression analysis of VFI or mean deviation, as well as series overview report, can be helpful when evaluating other conditions.2

Follow Up 
Use the same test strategy and pattern to allow easier comparison and monitoring for progression. Following established glaucoma patients with perimetry is essential in determining if the current management is adequate or if treatment changes are needed based on the stability or progression of the visual field. 

The frequency of follow up will depend on the extent of the disease and the clinical course. Patients who demonstrate stability could reasonably increase their follow up interval.  

There are several artifacts that may arise, imposing further challenges to interpreting visual fields. Clinicians must discern true visual field defects from pathology that correlate to the clinical picture vs. artifacts that may arise. At times when it is not clear if the visual field defect is real or an artifact, the visual field should be repeated in an appropriate timeline based on your level of suspicion and the clinical picture. Artifacts to look out for include:

Eyelid and brow ptosis. This can cause a dense superior defect along the superior edge points. Patients with visually significant ptosis may benefit from having their eyelids taped for testing. 

Rim artifacts. The positioning of the trial lens holder can lead to a rim artifact if it is too far from the patient's eye, creating a full or partial ring scotoma. 

Incorrect refractive error. This can lead to a generalized depression in sensitivity that may mimic that of a cataract. High refractive error can create a magnification or minification effect and require proper vertex distance calculations. Using the wrong lens power sign or not using the new refractive error after a patient has cataract surgery are two common errors. 

Patient fatigue. This may manifest with longer test times, high false negative value or preferentially abnormal peripheral sensitivity. Peripheral points are tested later in the course of the test and may be markedly reduced in cases of fatigue or waning attention, resulting in a darker, lobular ring or cloverleaf pattern on the grayscale.

A learning curve exists for both patients and those administering testing. Awareness of these commonly encountered artifacts allows for improved visual field interpretation and recognition of limitations of the reliability and quality of test results.

Why We Test
Understanding the visual pathway can give a provider valuable insight to localizing lesions. Each practice should standardize a preferred test strategy and pattern and repeat the same test on follow-up visits to enable more accurate comparisons throughout follow-up testing.18

Glaucoma management should be focused on preventing the loss of visual fields to the degree that it affects a patient's quality of life. It is important to evaluate the entire clinical picture to make sure variables correlate and coincide. Tests should be repeated if uncertainty exists to establish more reliable baseline measurements and confirm possible progression. 

Drs. Draskovic and McSoley are staff optometrists at the Bascom Palmer Eye Institute.

1. Lens A, Langley T, Nemeth SC, Shea C. Visual Pathway. In: Ocular Anatomy and Physiology. Thorofare, NJ: SLACK; 1999:90-5. 

2. Heijl A, Patella VM, Bengtsson B. The Field Analyzer Primer: Effective Perimetry. 4th ed. Carl Zeiss Meditec; 2012. 

3. Park HY, Hwang BE, Shin HY, Park CK. Clinical clues to predict the presence of parafoveal scotoma on Humphrey 10-2 visual field using a Humphrey 24-2 visual field. Am J Ophthalmol. 2016;161:150-9. 

4. Ehrlich AC, Raza AS, Ritch R, et al. Modifying the conventional visual field test pattern to improve the detection of early glaucomatous defects in the central 10°. Transl Vis Sci Technol. 2014 Oct;3(6):6. 

5. Kedar S, Ghate D, Corbett JJ. Visual Fields in Neuro-Ophthalmology. Indian Journal of Ophthalmology. 2011;59(2):103–9. 

6. Wall M, George D. Idiopathic intracranial hypertension. A prospective study of 50 patients. Brain. 1991;114:155-80.

7. Keltner JL, Johnson CA, Spurr JO, Beck RW. Baseline visual field profile of optic neuritis. The experience of the optic neuritis treatment trial. Optic Neuritis Study Group. Arch Ophthalmol. 1993;111:231-4.

8. Hayreh SS, Zimmerman B. Visual field abnormalities in nonarteritic anterior ischemic optic neuropathy: their pattern and prevalence at initial examination. Arch Ophthalmol. 2005;123:1554–62. 

9. Hayreh SS. Posterior ischaemic optic neuropathy: clinical features, pathogenesis, and management. Eye (Lond). 2004;18:1188–206.

10. Bengtsson B. Prediction of glaucomatous visual field loss by extrapolation of linear trends. Arch Ophthalmol. 2009;127(12):1610-5. 

11. Heijl A. Reduction of intraocular pressure and glaucoma progression. Arch Ophthalmol. 2002;120(10):1268-79.

12. Willmore LJ, Abelson MB, Ben-Menachem E, et al. Vigabatrin: 2008 update. Epilepsia. 2009;50:163–73.

13. Katz J, Sommer A, Gaasterland DE, Anderson DR. A comparison of analytic algorithms for detecting glaucomatous visual field loss. Arch Ophthalmol. 1991;109(12):1017-25. 

14. Hodapp E, Parrish RK II, Anderson DR. Clinical decisions in glaucoma. St. Louis: The CV Mosby Co;1993:52-61. 

15. Feuer WJ, Anderson DR. Static threshold asymmetry in early glaucomatous visual field loss. Ophthalmology. 1989;96:1285-97.

16. Brenton RS, Phelps CD, Rojas P, Woolson RF. Interocular differences of the visual field in normal subjects. Invest Ophthalmol Vis Sci. 1986;27:799-805.

17. Anderson DR, Chauhan B, Johnson C, et al. Criteria for progression of glaucoma in clinical management and in outcome studies. Am J Ophthalmol. 2000;130(6):827-29.

18. Weinreb RN. Progression of Glaucoma: The 8th Consensus Report of the World Glaucoma Association. Amsterdam: Kugler Publications; 2011.