On the development of a clinical test of facial recognition based on a spatial frequency model of visual information processing

This study developed a clinical test of visual processing based on spatial frequency theory. The change from light-to-dark-to-light over space is referred to as a spatial frequency and is measured in cycles per degree of visual angle. The theory holds that the visual system processes information by decomposing images into component spatial frequencies. As it applies to cerebral lateralization, the theory posits that the right cerebral hemisphere is differentially sensitive to relatively low spatial frequencies and that the left hemisphere is differentially sensitive to relatively high spatial frequencies. This study constructed a forced-choice, face-matching test that took into account the spatial frequency components of stimuli. Briefly, 30 high-frequency decision items and 30 low-frequency decision items (high and low 89 frequency scales) were constructed using undergraduate volunteers. On each item subjects were required to select from among a target-foil pair of spatial-frequency screened (high or low) black and white faces to match a full-spectrum face. The test was administered to 24 brain injured adults drawn from a Veteran's Administration Hospital and a large urban medical center. Half of the subjects had focal cerebral lesions localized to the left-hemisphere, post-rolandic area by radiographic information. The remaining half had lesions similarly located in the right cerebral hemisphere. It was hypothesized that side of cerebral lesion would interact with spatial frequency scale (high and low) such that left-hemisphere lesioned patients would perform worse on the high frequency scale than on the low frequency scale, and vice-versa for right-hemisphere lesioned patients. Presence of receptive language deficit was likewise hypothesized to interact with spatial frequency. Presence of left-visual neglect was expected not to correlate with test performance.;Results failed to confirm effects of lesion side and language deficit on spatial frequency. As expected, visual neglect had no effect on performance on the test. Extensive analysis of subject and lesion characteristics failed to uncover any significant influence of these variables. The need to conduct similar studies in patients with diffuse (but lateralized) brain damage was emphasized.