Interestingly, LGK-974 chemical structure the responses to the ambiguous images identified as picture B (the one eliciting responses) were not significantly different, both in terms of magnitude and latency, from the ones obtained when showing picture B without morphing. The responses to picture A (the one not eliciting responses) were also not significantly

different, in terms of magnitude and latency, from the ones to the ambiguous pictures recognized as A when considering the whole response strength (Figure 3). However, in this case there was a tendency for higher responses to the ambiguous pictures that did reach statistical significance when considering the time-resolved average responses (Figure 4). Thus, the lack of statistical difference with the whole response strength may be attributed to variability in the neurons’ responses. In fact, in about 20% of the cases, a linear classifier could distinguish above chance the presentation CP-673451 in vitro of the original and ambiguous pictures leading to the same

perceptual decision. Such differences could, in principle, be attributed to a higher cognitive load when deciding the identity of a morphed compared to a nonmorphed image, which could have involved different degrees of attention. It is also possible that, even if eventually making a single decision in each trial, subjects may have had (at least in some cases) an alternating percept between both identities when seeing the morphed pictures. We also observed a smaller difference between the responses to the original and the morphed presentations for the images eliciting responses (picture B) compared to the difference for the images not eliciting responses (picture A), which could in principle be attributed to a firing rate saturation—i.e., there was little modulation in the responses to picture B

because the neurons were already close to their maximum firing rates. Ambiguous percepts have a long history of being used to dissociate neural responses underlying the subjective perception by the subject from the sensory representation of the first visual stimuli (Logothetis and Schall, 1989, Leopold and Logothetis, 1996, Logothetis et al., 1996, Logothetis, 1998 and Kanwisher, 2001). In this respect, a classic paradigm is binocular rivalry, where two distinct images presented at each eye compete with each other and give rise to a fluctuating perception of one or the other. Single-cell recordings along the ventral visual pathway in monkeys have shown an increase in the number of neurons following the subjective perception in higher visual areas (Logothetis, 1998). Higher visual areas project to the MTL (Saleem and Tanaka, 1996, Suzuki, 1996 and Lavenex and Amaral, 2000), where modulations of the neurons’ firing with subjective perception were also found using a binocular rivalry paradigm (Kreiman et al.