Spatio-temporal Contrast Sensitivity in the Cardinal Directions of the Colour Space. A Review

Contrast sensitivity function in deep networks

The contrast sensitivity function (CSF) is a fundamental signature of the visual system that has been measured extensively in several species. It is defined by the visibility threshold for sinusoidal gratings at all spatial frequencies. Here, we investigated the CSF in deep neural networks using the same 2AFC contrast detection paradigm as in human psychophysics. We examined 240 networks pretrained on several tasks. To obtain their corresponding CSFs, we trained a linear classifier on top of the extracted features from frozen pretrained networks. The linear classifier is exclusively trained on a contrast discrimination task with natural images. It has to find which of the two input images has higher contrast. The network's CSF is measured by detecting which one of two images contains a sinusoidal grating of varying orientation and spatial frequency. Our results demonstrate characteristics of the human CSF are manifested in deep networks both in the luminance channel (a band-limited inverted U-shaped function) and in the chromatic channels (two low-pass functions of similar properties). The exact shape of the networks' CSF appears to be task-dependent. The human CSF is better captured by networks trained on low-level visual tasks such as image-denoising or autoencoding. However, human-like CSF also emerges in mid- and high-level tasks such as edge detection and object recognition. Our analysis shows that human-like CSF appears in all architectures but at different depths of processing, some at early layers, while others in intermediate and final layers. Overall, these results suggest that (i) deep networks model the human CSF faithfully, making them suitable candidates for applications of image quality and compression, (ii) efficient/purposeful processing of the natural world drives the CSF shape, and (iii) visual representation from all levels of visual hierarchy contribute to the tuning curve of the CSF, in turn implying a function which we intuitively think of as modulated by low-level visual features may arise as a consequence of pooling from a larger set of neurons at all levels of the visual system.

Automated grating contrast-sensitivity: The easy test for hidden visual loss in recovered optic neuritis patient

PURPOSE:

Residual visual loss is an important predictor of optic neuritis relapse and progression. This study aimed to investigate the hidden residual visual loss in patients with optic neuritis using automated contrast-sensitivity (CS) function testing.

MATERIALS AND METHODS:

This cross-sectional study investigated 29 recovered optic neuritis patients (age: 27.69 ± 13.32 years, range: 13–51). Twenty age-matched controls with normal visual acuity (VA, in LogMAR) were recruited, for comparison with patients' VA and CS function after stable recovery from optic neuritis (6 months of follow-up). CS tests used a novel software that displays a single set of Gabor patches (2 cycles per degree at 10° ×10° of visual angle) with contrasts grating from 100% to 0.1%.

RESULTS:

There were 13 adolescent patients (63.6%: retrobulbar neuritis [RN]; 36.4%: papillitis), 14 adult patients (50%: RN; 42.9%: papillitis), and only 2 older patients (all with neuroretinitis). There was improvement of VA in the patient group at first diagnosis and follow-up (VA initial vs. final: 1.438 ± 1.134 vs. 0.235 ± 0.272, P < 0.001). This VA improvement was similar to control group (P = 0.052). In CS, there were significant differences in patient versus control groups (69.069% ± 70.235% vs. 27.215% ± 25.27%, P = 0.025). Linear regression showed that initial VA and CS function could not predict final VA (P = 0.183 and P = 0.138, respectively).

CONCLUSIONS:

Patients with optic neuritis showed decreased CS compared to control group which indicated the residual visual loss. Automated CS testing is useful in detecting residual visual loss in patients who recovered from optic neuritis.

Spatio-chromatic contrast sensitivity under mesopic and photopic light levels

Contrast sensitivity functions (CSFs) characterize the sensitivity of the human visual system at different spatial scales, but little is known as to how contrast sensitivity for achromatic and chromatic stimuli changes from a mesopic to a highly photopic range reflecting outdoor illumination levels. The purpose of our study was to further characterize the CSF by measuring both achromatic and chromatic sensitivities for background luminance levels from 0.02 cd/m² to 7,000 cd/m². Stimuli consisted of Gabor patches of different spatial frequencies and angular sizes, varying from 0.125 to 6 cpd, which were displayed on a custom high dynamic range (HDR) display with luminance levels up to 15,000 cd/m². Contrast sensitivity was measured in three directions in color space, an achromatic direction, an isoluminant “red-green” direction, and an S-cone isolating “yellow-violet” direction, selected to isolate the luminance, L/M-cone opponent, and S-cone opponent pathways, respectively, of the early postreceptoral processing stages. Within each session, observers were fully adapted to the fixed background luminance (0.02, 2, 20, 200, 2,000, or 7,000 cd/m²). Our main finding is that the background luminance has a differential effect on achromatic contrast sensitivity compared to chromatic contrast sensitivity. The achromatic contrast sensitivity increases with higher background luminance up to 200 cd/m² and then shows a sharp decline when background luminance is increased further. In contrast, the chromatic sensitivity curves do not show a significant sensitivity drop at higher luminance levels. We present a computational luminance-dependent model that predicts the CSF for achromatic and chromatic stimuli of arbitrary size.

Visuocortical changes during a freezing-like state in humans

An adaptive response to threat requires optimized detection of critical sensory cues. This optimization is thought to be aided by freezing - an evolutionarily preserved defensive state of immobility characterized by parasympathetically mediated fear bradycardia and regulated by the amygdala-periaqueductal grey (PAG) circuit. Behavioral observations in humans and animals have suggested that freezing is also a state of enhanced visual sensitivity, particularly for coarse visual information, but the underlying neural mechanisms remain unclear. We induced a freezing-like state in healthy volunteers using threat of electrical shock and measured threat-related changes in both stimulus-independent (baseline) and stimulus-evoked visuocortical activity to low-vs. high-spatial frequency gratings, using functional MRI. As measuring immobility is not feasible in MRI environments, we used fear bradycardia and amygdala-PAG coupling in inferring a freezing-like state. An independent functional localizer and retinotopic mapping were used to assess the retinotopic specificity of visuocortical modulations. We found a threat-induced increase in baseline (stimulus-independent) visuocortical activity that was retinotopically nonspecific, which was accompanied by increased connectivity with the amygdala. A positive correlation between visuocortical activity and fear bradycardia (while controlling for sympathetic activation), and a concomitant increase in amygdala-PAG connectivity, confirmed the specificity of these findings for the parasympathetically dominated freezing-like state. Visuocortical responses to gratings were retinotopically specific, but did not differ between threat and safe conditions across participants. However, individuals who exhibited better discrimination of low-spatial frequency stimuli showed reduced stimulus-evoked V1 responses under threat. Our findings suggest that a defensive state of freezing involves an integration of preparatory defensive and perceptual changes which may be regulated by a common mechanism involving the amygdala.

Contrast sensitivity in relapsing-remitting multiple sclerosis assessed by sine-wave gratings and angular frequency stimuli

Previous studies have shown that multiple sclerosis (MS) affects the visual system, mainly by reducing contrast sensitivity (CS), a function that can be assessed by measuring contrast sensitivity function (CSF). To this end, we measured both the CSF for sine-wave gratings and angular frequency stimuli with 20 participants aged between 21 and 44 years, of both genders, with normal or corrected to normal visual acuity. Of these 20 participants, there were 10 volunteers with clinically defined MS of the relapsing-remitting clinical form, with no history of optic neuritis (ON), as well as 10 healthy volunteers who served as the control group (CG). We used a forced-choice detection paradigm. The results showed reduced CS to both classes of stimuli. Differences were found for sine-wave gratings at spatial frequencies of 0.5, 1.25, and 2.5 cycles per degree (cpd) (P < 0.002) and for angular frequency stimuli of 4, 24, and 48 cycles/360° (P < 0.05). On the one hand, comparing the maxima of the respective CSFs, the CS to angular frequency stimuli (24 cycles/360°) was 1.61-fold higher than that of the CS to vertical sine-wave gratings (4.0 cpd) in the CG; for the MS group, these values were 1.55-fold higher. On the other hand, CS in the MS group attained only 75% for 24 cycles/360° and 78% for 4.0 cpd of the 100% CS estimates found for the CG at the peak frequencies. These findings suggest that MS affects the visual system, mostly at its maximum contrast sensitivities. Also, since angular frequencies and sine-wave gratings operate at distinct levels of contrast in the visual system, MS seems to affect CS at both high and low levels of contrast.