Lower Parietal Encoding Activation Is Associated with Sharper Information and Better Memory

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Neuroimaging

Background:

• Mean fMRI activation in the ventral posterior parietal cortex (vPPC) during memory encoding often inversely predicts memory success.
• This has led to interpretations of vPPC engagement as "off-task" processing.
• However, recent studies suggest distributed patterns of activity in vPPC may actively represent encoded information.

Purpose of the Study:

• To investigate the relationship between pattern-based content representations in vPPC, mean vPPC activation, and subsequent memory recall.
• To determine if distributed activity patterns in vPPC reflect memory encoding quality.

Main Methods:

• Analysis of fMRI data from two experiments involving memory encoding and recall of word-face or word-scene associations.
• Measurement of mean univariate activation within vPPC for each encoding trial.
• Quantification of the strength of face/scene information using pattern analysis within the same vPPC voxels.

Main Results:

• Mean vPPC activation negatively predicted subsequent remembering.
• The strength of pattern-based information within vPPC positively predicted later memory performance.
• A negative correlation was observed between mean vPPC activation and pattern-based information strength.
• Reduced univariate activation was most pronounced in voxels not strongly tuned to the encoded stimulus category.

Conclusions:

• vPPC activity patterns encode the content and quality of memory representations.
• Lower overall vPPC activation can correspond to stronger, more detailed pattern-based information, challenging the "off-task" interpretation.
• This highlights the importance of considering distributed neural patterns alongside mean activation in fMRI studies of memory.