MRI and machine learning will help predict emotions

In People's Minds Researchers from the Center for Neuroscience Research (CNR) of the Institute of Basic Sciences (IBN), in collaboration with Dartmouth College, were able to predict subjective feelings in people's thoughts using functional magnetic resonance imaging (MRI) and machine learning algorithms. The study, published in the journal Proceedings of the National Academy of Sciences, opens new horizons in understanding the human brain. DiscussMRI and machine learning will help predict emotions


Research showed that it is possible to develop models for predicting emotional state during spontaneous thoughts by combining personal narratives with fMRI data. Study participants, in single interviews, created personalized narratives reflecting their past experiences and emotions. While they read the stories inside the MRI, their brain activity was recorded.

The resulting data was used to train predictive models that decoded the “emotional dimensions” of thoughts in real time. The researchers identified the importance of different networks in predicting subjective feelings and their connections to different parts of the brain.

The models also successfully predicted emotions not only during story reading, but also during free thinking or in a resting state. This discovery opens new perspectives in understanding internal states and helps in assessing mental well-being.