The Macular Search Paradigm
Project Completed
This was a cooperation with Donald C. Fletcher, MD, the director of the Low Vision Service at California Pacific Medical Center, San Francisco.
The most important limitation in conventional perimetry and other methods of measuring vision topographically is the fact that the subject/patient has to steadily fixate a known location. This means that every eye movement during the measurement adds a mistake to the retinal coordinates of the perceived stimulus. As this laboratory deals mostly with elderly subjects, especially those with maculopathies, we can never forget this fundamental problem.
The research project aimed at overcoming this problem by freeing the patients from the fixation restriction. The topographic component is here preserved only by the fact that targets appear in different known but unpredictable locations. The task is: Find and identify the target. The patient can make any eye movement necessary to solve it. Since eye movements change spatial coordinates, we measured the time it takes the patient to solve the task. Only correct answers were accepted.
First, we measure the size threshold for each patient. The threshold is then doubled, which determines the size of target he/she has to identify. When a trial starts, the patient casually looks at the middle of the screen and then starts the searching eye movements. The target can appear in any of 32 locations, all of which are used only once. Targets are Landolt Cs (Where is the gap in the ring?)
Summary:
Participants: 115 consecutive low vision patients with varying diagnoses and 15 control subjects of comparable median age.
Methods: Psychophysical measurement of performance in a task that required finding and identifying visual targets. They appeared consecutively on a computer screen in 32 locations of the central visual field. The task required no physical action from the subjects who were free to make eye movements as needed. Target size was always double the size threshold. Best attainable reading performance at any size was measured routinely (MNread).
Results: Measurements yielded a wide variety of performance levels, with a factor of 14-16 between best and worst performers. Only weak correlations were found between performance and age or visual acuity. No statistically significant correlations were found with age or diagnosis. We found the highest correlation between median latency and best attainable reading speed at any size (MNread).
Conclusions: The “search-and-identify” paradigm and reading share an important mechanism that determines performance in both tasks. We hypothesize that the factor enabling patients to perform well in both paradigms is oculomotor control and/or eye movement strategy. The search-and-identify paradigm is closer to real-life visual tasks than standard clinical tests. It makes assessment of visual impairment independent of language, level of literacy and reading habits.
Publication:
MacKeben M & Fletcher DC. Target search and identification performance in low vision patients. Submitted and in review