Abstract
Introduction: The threshold contrast and the slope of the psychometric function can be determined by using computer-based stimulus presentations and an adaptive staircase, like the TuebingenCSTest (TueCST). The TueCST assesses contrast sensitivity with high repeatability and reliability, using a fixed slope. The slope of the psychometric function is the measure of change in performance with changing stimulus intensity, which reflects the variability of the estimated threshold contrast. The slope relies on several factors, like finger errors, eye artefacts due to e.g. blinks, observers' attention or motivation. The aim of the study was to estimate individual slopes for each participant in order to increase the accuracy of the threshold contrast assessment. Methods: In this study, the slope of the psychometric function was estimated with the TueCST, using 250 trials for different spatial frequencies in eight experienced (0.5, 1.5, 3, 6 and 12 cycle per degree – cpd) and eight naïve participants (1.5, 3, 6, 12, 20 and 30cpd). The stimulus was a Gabor Patch grating with four possible orientations to identify. Results: Repeated ANOVA returned a significant effect for spatial frequency in experienced participants (F(4,28)=30.9, p< 0.001, ω2=0.78) and naïve participants (F(5,35)=62.4, p< 0.001, ω2=0.88). Largest slopes were estimated for lower spatial frequencies (≤6cpd) and smallest slopes for higher spatial frequencies (≥12cpd) in both subgroups. In naïve participants, the threshold contrast was significantly lower for a first measurement with a fixed slope compared to the second measurement for slope estimation (p< 0.05). Comparing the slopes of naïve and experienced participants, there was no significant difference (p=0.7). Conclusion: When using the TueCST, the estimated slopes of the psychometric function showed a difference between higher and lower spatial frequencies. These results revealed that naïve participants were sufficiently trained after one measurement with the TueCSTest, resulting in indistinguishable slopes between naïve and experienced participants.
Meeting abstract presented at VSS 2018