Self-reports map the landscape of task states derived from brain imaging

Area of Science:

• Cognitive Neuroscience
• Psychology
• Machine Learning

Background:

• Psychological states impact happiness and productivity.
• Introspection's accuracy in quantifying psychological states has been debated.
• Prior studies show correlations between self-reports and objective cognitive/brain measures.

Purpose of the Study:

• To investigate if machine learning can map introspective descriptions of psychological states to objective brain activity.
• To determine if self-reported experiences can predict brain states.

Main Methods:

• Utilized machine learning to analyze self-reported psychological states from 194 participants performing various tasks.
• Combined introspective reports with functional magnetic resonance imaging (fMRI) data from prior studies.
• Developed a 'state-space' model to predict brain activity patterns from experience descriptions.

Main Results:

• Machine learning models reliably predicted brain activity patterns using only unseen descriptions of psychological experiences.
• Demonstrated that introspective reports share significant information with the objective task landscape derived from brain activity.
• The 'state-space' model successfully mapped subjective experiences to objective neural representations.

Conclusions:

• Introspective reports are a valuable source of information that can be quantitatively mapped to objective neural data.
• This approach advances the understanding of the relationship between subjective experience and brain function.
• Findings suggest a more objective quantification of psychological states is possible through integrating self-reports and neuroimaging.