Machine Learning Allows Empathy To Be Predicted in Resting Brain Interview with:

Leonardo Christov-Moore, Ph.D. Postdoctoral Scholar Brain and Creativity Institute University of Southern California

Dr. Christov-Moore

Leonardo Christov-Moore, Ph.D.
Postdoctoral Scholar
Brain and Creativity Institute
University of Southern California What is the background for this study?

Response: We’ve known for some time that empathy has both bottom-up, affective, somatomotor components, that let us quickly feel and internally simulate other peoples’ internal states, and more cognitive, top-down components through which we make conscious inferences about others’ beliefs, intentions and internal states. And there is interesting work suggesting that in many cases, these components work together. In our work, we took this idea further to propose that they exist in constant interaction, with the bottom-up systems  providing information that informs the top-down processes (aiding in our inference), which in turn provide modulation and control to the bottom-up processes (modulating the extent to which we “resonate” with others based on context, affiliation, etc.).

Specifically, we found that you could predict many aspects of prosocial decision-making ( a top-down task) from bottom-up and top-down systems’ interaction during simple bottom-up empathy tasks (passively observing someone experience emotion or pain). This led us to hypothesize that peoples’ levels of empathic concern for others are dictated by stable patterns of interaction between these systems.

In the current study, we made a strong test of this hypothesis: if these empathy-predicting patterns of interaction are stable across task demands, we should be able to observe them (and predict empathic concern from them) even when the brain is not doing anything ostensibly related to empathy! So that’s what we did. What are the main findings? Is empathy centered in a particular brain locus or set of neurons?

Response: We found that you could use machine learning (which is able to make predictions based on subtle patterns in data) to predict subjects’ levels of empathic concern for others, just from patterns of interaction between sensorimotor and emotional brain areas, and more cognitive, prefrontal/temporal “control” areas, in the resting brain. This suggests that empathy relies on complex interactions between many brain systems that are constantly dynamically interacting, rather than a single group or type of neurons. It’s a network level phenomenon. What should readers take away from your report?

1) That empathy arises from many systems in interaction, that aren’t cleanly separable. There isn’t an “empathy brain area”. This may  be why so many different psychiatric and neurological disorders exhibit deficits in empathy and social cognition. It’s a complex process that requires the brain to work in coordination.

2) That in time we may be able predict aspects of daily behavior and personality without having to use conventional self-reports or other measures. This makes it possible to diagnose and assess brain function in people who may (for other reasons) be unable to perform these tasks or fill out the questionnaires. What recommendations do you have for future research as a result of this work? 

Response: We should continue exploring the diagnostic potential of the brain at rest, and we should study how cognitive processes emerge from brain organization, rather than start with psychological constructs and work our way down to brain architecture. Is there anything else you would like to add?

Response: Empathy isn’t just a key to compassion and ethical behavior, it’s also important for mental health!


Leonardo Christov-Moore, Nicco Reggente, Pamela K. Douglas, Jamie D. Feusner, Marco Iacoboni. Predicting Empathy From Resting State Brain Connectivity: A Multivariate Approach. Frontiers in Integrative Neuroscience, 2020; 14 DOI: 10.3389/fnint.2020.00003



The information on is provided for educational purposes only, and is in no way intended to diagnose, cure, or treat any medical or other condition. Always seek the advice of your physician or other qualified health and ask your doctor any questions you may have regarding a medical condition. In addition to all other limitations and disclaimers in this agreement, service provider and its third party providers disclaim any liability or loss in connection with the content provided on this website.


Last Updated on February 25, 2020 by Marie Benz MD FAAD