Figure DDD.2-1 Schematic Representation of the Human Eye
Figure DDD.2-2 Sample Report from an Automated Visual Field Machine
The diagnosis and management of Glaucoma, a disease of the optic nerve, is the primary use of visual field testing. In this regard, automated visual fields are used to assess quantitatively the function of the optic nerve with the intent of detecting defects caused by glaucoma.
The first step in analyzing a visual field report is to confirm that it came from the correct patient. Demographic information including the patient’s name, gender, date of birth, and perhaps medical record number are therefore essential data to collect. The patient’s age is also important in the analysis of the visual field (see below) as optic nerve function changes with age. Finally, it is important to document the patient’s refractive error as this needs to be corrected properly for the test to be valid.
Second, the clinician needs to assess the reliability of the test. This can be determined in a number of ways. One of these is by monitoring patient fixation during the test. To be meaningful, a visual field test assumes that the subject was looking at a fixed point throughout the test and was responding to stimuli in the periphery. Currently available techniques for monitoring this fixation include blind spot mapping, pupil tracking, and observation by the technician conducting the test. Blind spot mapping starts by identifying the small region of the visual field corresponding to the optic nerve head. Since the patient cannot detect stimuli in this area, any positive response to a stimulus placed there later in the test indicates that the patient has lost fixation and the blind spot has “moved”. Both pupil tracking and direct observation by the technician are now easily carried out using a camera focused on the patient’s eye.
Figure DDD.2-3 Information Related to Test Reliability
Another means of assessing the reliability of the test is to count both false positive and false negative responses. False positives occur when the subject presses the button either in response to no stimulus or in response to a stimulus with intensity significantly below one they had not detected previously. False negatives are recorded when the patient fails to respond to stimulus significantly more intense than one they had previously seen. Taken together, fixation losses, false positives, and false negatives provide an indication of the quality of the test.
The next phase of visual field interpretation is to assess for the presence of disease. The first aspect of the visual field data used here are the raw sensitivity values. These are usually expressed as a function of the amount of attenuation that could be applied to the maximum possible stimulus such that the patient could still see it when displayed. Since a value is available at each point tested in the visual field, these values can be represented either as raw values or as a graphical map.
Figure DDD.2-4 Sample Output from an Automated VF Machine Including Raw Sensitivity Values (Left, Larger Numbers are Better) and an Interpolated Gray-Scale Image
Because the raw intensity values can be affected by a number of factors including age and other non-optic nerve problems including refractive error or any opacity along the visual axis (cornea, lens, vitreous), it is helpful to also evaluate some corrected values. One set of corrected intensity values is usually some indication of the difference of each tested point from its expected value based on patient age. Another set of corrected intensity values, referred to as “Pattern deviation or “Corrected comparison” are normalized for age and also have a value subtracted from the deviation at each test point, which is estimated to be due to diffuse visual field loss This latter set is useful for focal rather than diffuse defects in visual function. In the case of glaucoma and most other optic nerve disease, clinicians are more interested in focal defects so this second set of normalized data is useful.
Figure DDD.2-5 Examples of Age Corrected Deviation from Normative Values (upper left) and Mean Defect Corrected Deviation from Normative Data (upper right)
For all normalized visual field sensitivity data, it is useful to know how a particular value compares to a group of normal patients. Vendors of automated visual field machines therefore go to great lengths to collect data on such “normal” subjects to allow subsequent analysis. Furthermore, the various sets of values mentioned above can be summarized further using calculations like a mean and standard deviation. These values give some idea about the average amount of field loss (mean) and the focality of that loss (standard deviation).
A final step in the clinical assessment of a visual field test is to review any disease-specific tests that are performed on the data. One such test is the Glaucoma Hemifield Test which has been designed to identify field loss consistent with glaucoma. These tests are frequently vendor-specific.
In addition to primary diseases of the optic nerve, like glaucoma, visual fields are useful for assessing damage to the visual pathway occurring between the optic chiasm and occipital cortex. There is the same need for demographic information, for assessment of reliability, and for the various raw and normalized sensitivity values. At this time, there are no well-established automated tests for the presence of neurological defects.
Figure DDD.2-6 Example of Visual Field Loss Due to Damage to the Occipital Cortex Because of a Stroke
The Diffuse Defect is an estimate of the portion of a patient’s visual field loss that is diffuse, or spread evenly across all portions of the visual field, in dB. In this graphical display, deviation from the average normal value for each test point is ranked on the x axis from 1 to 59, with 59 being the test point that has the greatest deviation from normal. Deviations from normal at each test point are represented on the y axis, in dB. The patient’s actual test point deviations are represented by the thin blue line. Age corrected normal values are represented by the light blue band. The patient’s deviation from normal at the test point ranked 25% among his or her own deviations is then estimated to be his or her diffuse visual field loss, represented by the dark blue band. This provides a graphical estimate of the remaining visual field loss for this patient, which is then presumed to consist of local visual field defects, which are more significant in management of glaucoma than diffuse defects.
Figure DDD.2-7 Example of Diffuse Defect
The Local Defect is an estimate of the portion of a patient’s visual field loss that is local, or not spread evenly across all portions of the visual field. The x and y axis in this graphical display have the same meaning as in the diffuse defect. In this graphical display the top line/blue band represent age corrected normal values. This line is shifted downward by the amount estimated to be due to diffuse visual field loss for this patient, according to the calculation in Figure DDD.2-7 (Diffuse Defect). The difference between the patient’s test value at each point in the ranking on the horizontal axis and the point on the lower curve at the 50% point is represented by the dark blue section of the graph. This accentuates the degree of local visual field defect, which is more significant in management of glaucoma than diffuse defects. The Local Defect is an index that highly correlates with square root of the loss variance (sLV) but is less susceptible to false positives. In addition to the usage in white/white perimetry it is especially helpful as early identifier for abnormal results in perimetry methods with higher inter subject variability such as blue/yellow (SWAP) or flicker perimetry. An example of Local Defect is shown in Figure DDD.2-8 and is expressed in dark blue in dB and is normalized to be comparable between different test patterns.
Figure DDD.2-8 Example of Local Defect