Henson 9000: Technical profile

Threshold tests are used when there is a need to accurately measure the depth of any visual field loss.

The first threshold algorithm (Full Threshold) was developed in the 1970s. It was reasonably accurate but took in excess of 10 minutes per eye to perform. A later development (FastPac, Fast Threshold) shortened the test time but was less accurate and as such was not widely adopted. In the 1980/90s a series of papers were published by vision scientists on the more efficient Bayesian approach to obtaining thresholds. One of the algorithms promoted by these researchers (King-Smith 1994) was called ZEST (Zippy Estimate by Sequential Testing). In 1997 a Swedish group of ophthalmologists utilised this work to develop a new perimetric algorithm called SITA (Swedish Interactive Threshold Algorithm). The ZEST algorithm has also been used to develop ZATA (Zippy Adaptive Threshold Algorithm) the threshold test used on the Henson 9000.

An important characteristic of these new algorithms is what is known as their terminating criteria. This specifies how accurate the threshold measure should be. As a general rule the more accurate you want a threshold test to be the longer it will take to perform. The Swedish (SITA) group set their terminating criteria to give an accuracy similar to the 1970s Full Threshold technique (termed SITA Standard), later producing a second version that was faster, but less accurate (SITA Fast). Since the development of SITA a lot more has been learnt about how patients with glaucoma respond to stimuli and the importance of keeping test times down to a minimum. These findings have been used to develop ZATA, the fastest, most accurate threshold test available today.

While ZATA is based upon the same principles as SITA it differs in some important ways.

1. When possible it uses prior data as a starting level. In instances where patients have been tested previously, ZATA will build on this prior data for subsequent tests, rather than starting a new test from age normative data, as is common in other perimeters. This helps negate false positive responses early in a test by patients with established loss and serves to reduce patient anxiety, too, by reducing the number of unseen presentations. It can also speed up test times – benefiting patients and enhancing practice workflow.
2. ZATA does not just use single terminating criteria. ZATA varies the terminating criteria to give more accurate thresholds at damaged and neighbouring locations. This reduces test times, both for patients with advanced visual fi eld loss and for those with no loss at all.
3. ZATA uses looser terminating criteria in severely damaged locations (<10 dB) ZATA does not attempt an accurate measurement of thresholds below 10 dB where variability is high and attempts at accurate measures do not yield any useful additional data.
4. ZATA allows both 24-2 and 30-2 stimulus patterns in a single test. Via a simple ‘extend’ facility operators can extend the 24-2 test pattern to a 30-2 test pattern during or at the end of each test so avoiding the need to retest the 24-2 locations if later opting for a 30-2 test.
5. ZATA uses the standard printout for universal compatibility. The standard print format is used to aid interpretation and comparison with data from other perimeters. The Henson 9000 software also allows users to switch between multiple views – threshold, grayscale, or defect values – at the end of each test.

You have all the tools you need to detect and monitor functional loss with our range of supra-threshold and threshold tests, as well as monitoring the whole visual field with the Esterman binocular driving tests (Groups 1 & 2).

Our Smart Supra results are designed to be compatible with threshold results, whilst our ZATA results include all recognized global indices (mean defect, standard deviation defect, hemifield), plus threshold and grayscale, total deviation and pattern deviation values.

ZATA progression analysis tools display global indices (mean defect and pattern standard deviation) against the age of the patient and the mean defect rate of change displayed with confidence levels.

Threshold data can also be presented on Glaucoma Staging System II. Using both the mean defect and pattern standard deviation is better than using either index in isolation.

The ZATA threshold test includes foveal threshold measurement facility as standard.

• 100% sensitivity = every screening test would capture those with a visual field defect, i.e. result = fail
• 100% specificity = every screening test would not fail those with a healthy visual field, i.e. result = pass

Ideally, a screening test would be 100% sensitive and 100% specific, but in reality there will always be anomalies – i.e. ‘false negatives’ (those with the condition who pass the test) and ‘false positives’ (those without the condition who fail the test).

The sensitivity/specificity of a visual field screening test can be altered by changing the fail criteria, e.g. if the fail criteria of a supra-threshold visual field test were to be changed from a single missed point to a cluster of three missed points, then the sensitivity would go down and the specificity would go up.

So, which is the more important? To answer this question you have to answer another. What are the relative costs of a false positive and false negative? When screening for glaucoma in an unselected population, where the prevalence of the disease is low (<2%), it is generally accepted that the specificity should be ≥95%. If it falls much below this figure then we end up failing a relatively large number of people that do not have the disease. If the study population is a high risk one (higher percentage of true cases) then the sensitivity should be increased at the cost of lowered specificity.

There are 2 ways in which you can screen the visual field with the Henson 9000.

You can use the fast and accurate supra-threshold strategies (Smart Supra on Version 4.0 software), with a choice of single or multiple stimulus presentations for additional flexibility, or the rigorous and detailed ZATA threshold test.

Screening with the supra-threshold strategies is certainly the fastest option but the testing times of ZATA, compared to other threshold algorithms, make it feasible to offer a threshold strategy by default as a form of enhanced screening test.

The length of a ZATA test will be determined by the extent of a patient’s field loss.

• A normal patient with no field loss will complete the test in approximately 2.5 minutes (with average response times)
• A patient with significant field loss tested from previously stored data will likely complete the test in under 4 minutes due to the ZATA test’s variable termination criteria.
• A patient with moderate field loss, or one who has no prior data stored, will take approximately 4-5 minutes per eye to complete the test.

ZATA was developed to counter all of the problems associated with testing patients with field loss on a full threshold type test. It runs the 24/2 test pattern but uses supra- as well as full threshold testing with different terminating criteria. Furthermore, it uses data from surrounding points as well as prior data (if available).

All of these strategies combine to make ZATA an easier test for patients to perform while offering the same level of accuracy as a standard full threshold would in monitoring glaucoma progression.

The ZATA algorithm is able to make use of prior patient data. If the Henson database contains a record for a previous threshold test then ZATA uploads the previous threshold measures to use as starting values for the new test. This reduces test times and improves accuracy.

Nor does ZATA use a fixed terminating criteria. In areas of severe loss, where thresholds estimates are very variable, and at locations near to normal values, it uses looser terminating criteria, reserving more accurate measures for locations where there is existing mild damage. This again speeds up test times by focusing on locations where change is expected to occur.

In other ways ZATA and SITA are similar. They both use the same test patterns (24/30-2). The threshold values are compatible as are the global indices.

All Hensons are static only perimeters. Isopters require kinetic testing, i.e. a stimulus that moves across the visual field. Kinetic perimetry is only available on selected HFA and Octopus perimeters.

The Henson 9000 performs static determination within the central 30° and qualitative automated threshold perimetry, including the 30-2 and 24-2 patterns of test locations. It is very fast, although exact test times are dependent upon the number of test locations and the type of visual field loss.

In a recent as yet unpublished project with 17 patients, the test time for a 24-2 ZATA Standard test with a superior arcuate defect was 3:01 mins, compared to 7:45 for a Full Threshold test with the same patient sample.

The Henson Esterman test is approved for testing Group 1 and Group 2 drivers.

Yes, it can be operated from any PC running the MS Windows™ Professional operating system (version 7, or above).

The control device must also have two spare USB ports for interfacing with the Henson 9000.

Yes. The software is supplied on a USB memory stick and is a single file installer. Simply click on it and it will install all parts of the software.

Both Henson machines are supplied with a quick start guide that takes you through the whole process step-by-step.

This relates to a problem with the settings for your backup.

Firstly, check the location of the backup allows files to be created. If it has been set to a read-only drive, or one that is protected by Windows, then the record will save in the database successfully but the backup will fail – hence the error message you are receiving.

Such read-only locations may be: the root of the C drive; a fundus camera (these often show up in Windows as USB flash drives); a Windows system folder or a mapped drive on a computer that is not switched on.

To ensure a successful backup set the location to a mapped network drive on your PC or a removable hard drive/USB flash drive.

Yes.

The Henson software prints the test result to the default installed printer on the laptop/computer being used as the control device. Printing is done via the default installed PDF viewer (Adobe Reader in the case of Henson 8000). Essentially, as long as the installed printer is Windows-compatible it will work without a problem.