SFN: Area PITd, a novel cortical area controlling visual attention?

I’ll admit that I listened to this talk only because it immediately preceded a talk on cholinergic modulation of visual and olfactory attention. However, I’ll be posting my notes from this talk. My reasons are two-fold: 1) the visual/olfaction talk was not as satisfying as I had hoped and 2) this talk appeared to reflect a solid scientific effort. Working on a mid-brain structure (the superior colliculus, thus the interest in attention), I am not as familiar as some about different cortical areas, and the process of identifying the involvement of those areas in particular processes. Those readers who are more familiar, please comment on your impressions of the research below. Without further ado…

This talk was entitled Cortical Area PITd, a ventral pathway area for the control of spatial visual attention?, by H. Stemman and W. Freiwald from Rockefeller University. As suggested by the title, the research identifies a role for a novel cortical region in the control of attention.

Starting off, the speaker (Stemmann) defined the functional properties that a region must exhibit in order to be classified as an attention control area. First, neurons within the region must encode visual information in a featureless manner. Second, the area should be involved in different kinds of attention. Finally, and most critically, electrical stimulation of the region must enhance the representation, and facilitate the recognition of objects in a spatially specific way.

The researchers started out using fMRI on monkeys to detect brain areas showing attention-related enhancement of activity. Their study identified an area in the inferior temporal cortex: dorsal posterior inferior temporal cortex. They note the magnitude of the attentional modulation in PITd was similar to modulation also seen in V1, V2, MT, FEF and LIP of their monkeys.

The researchers then recorded from PITd with multiunit electrodes while presenting the monkey with a variety of visual stimuli. They found no direction selectivity, but they did see enhanced firing activity when animal is directing its spatial attention to an area inside the receptive field of a recorded neuron. They recorded activity during a behavioral task, and were able to infer the trial outcome based on the spiking activity of the neuron (enhancement of firing rate was that strong). Furthermore, responses of PITd neurons accurately predicted task errors. This high degree of predictive ability was true for their entire cohort of 56 neurons.

Although they found no explicit direction selectivity, the researchers wondered whether the activity of PITd neurons was dependent stimulus motion (given that their stimulus involved fields of moving dots)? In order to confirm that the attentional effects/activity was not dependent of stimulus motion, they switched paradigms to a color discrimination task. They found that PITd neurons had no color tuning, but were still strongly modulated by attention.

From these results, they concluded that neurons located in PITd do not encode visual features of the stimulus, and therefore encode visual information in a featureless manner, and are involved in different kinds of attentional tasks (based on attentional modulation during two distinct task paradigms).

With confirmation that PITd satisfied two of the criteria for categorization as an area involved in controlling attention, the researchers next attempted to determine whether the region satisfied the third criteria. They asked whether electrical stimulation of PITd would cause changes in attention. Stimulation of PITd neurons when the visual target was inside the neurons’ receptive field result stimulated PITd when the target was inside the receptive field, saw an increase i

Electrical micro-stimulation inside receptive field of a PITd neuron made the monkey perform the visual task better. Micro-stimulation of PITd outside the receptive field caused the monkey to make significantly more errors as if attention was being directed to a spatial location outside the receptive field. Put another way, following micro-stimulation outside the receptive field, the monkey used the motion information encoded by the distracter to make their decision, not the cued location to which they should have been paying attention. These results suggested that by stimulating PITd, they could shift the spotlight of attention from to the cued location to the micro-stimulated location.

In summary, the presenter concludes that his research suggests that cortical area PITd, a ventral pathway area, may be involved in the control of visual attention.

A note that it is unclear (at least to me) what are the efferent and afferent connections to PITd, or how PITd fits into the established cortical attention circuit. However, the strength of the attentional modulation of PITd neurons, and their apparent ability to shift spatial attention is certainly impressive.