Response of the multiple-demand network during simple stimulus discriminations

Area of Science:

• Cognitive Neuroscience
• Neuroimaging
• Human Perception

Background:

• The multiple-demand (MD) network is crucial for cognitive control, responding to various task difficulty aspects like rule complexity and memory load.
• Existing research suggests MD activity is linked to top-down control, but its role when sensory input quality limits performance is less understood.
• Prior studies indicated minimal impact of sensory manipulations on MD activity, questioning the extent of top-down compensation for sensory limitations.

Purpose of the Study:

• To investigate the relationship between sensory task difficulty and multiple-demand (MD) network activity.
• To examine how manipulating motion coherence and distractor salience affects MD responses.
• To compare the impact of sensory difficulty versus rule complexity on MD network engagement.

Main Methods:

• Participants judged motion direction with difficulty manipulated across six levels via motion coherence or irrelevant dot salience.
• Task difficulty was presented in either blocked (predictable) or randomized (unpredictable) conditions.
• Functional magnetic resonance imaging (fMRI) measured brain activity, focusing on the MD network.

Main Results:

• MD network activity showed minimal changes with increasing sensory difficulty, particularly for the motion coherence manipulation.
• Randomized difficulty levels did not significantly alter the small effects of sensory manipulations on MD activity.
• In contrast, increased rule complexity consistently elevated MD activity, as expected.

Conclusions:

• For simple sensory discriminations, the multiple-demand network may not effectively compensate for reduced stimulus information through increased top-down control.
• These findings highlight a potential limit in the capacity of top-down control to overcome sensory input quality degradation.
• The dissociation between sensory difficulty and rule complexity effects suggests distinct mechanisms underlying MD network engagement.