In the study we developed a three-level framework for interpreting neuroimaging-based machine-learning models consisting of the model-level, feature-level and biology-level assessments. We an example workflow with data and codes available at our Github page.
You can also check out our video abstract!
Kohoutová, L., Heo, J., Cha, S., Lee, S., Moon, T., Wager, T. D., & Woo, C. W. (2020). Toward a unified framework for interpreting machine-learning models in neuroimaging. Nature protocols, 15(4), 1399-1435.