Translational utility is the ability of certain biomedical imaging features to capture useful subject-level characteristics in clinical settings, yielding sensible descriptions and/or predictions for individualized treatment trajectory. An important step in achieving translational utility is to demonstrate the association between imaging features and individual characteristics, such as sex, age, and other relevant assessments, on a large out-of-sample unaffected population (no diagnosed illnesses). This initial step then provides a strong normative basis for comparison with patient populations in clinical settings. Given the complexity of the human brain, it is likely that a multi-view approach based on features from several imaging modalities will yield the greatest utility and be least prone to biases and confounds such as site/scanner effects. Detailed information.
- Start date: April 22, 2020
- Entry and team merger deadline: June 20, 2020
- End date (preliminary private leaderboard revealed): June 29, 2020
- Code and leaderboard result verification (winners announced): July 26, 2020
Contacts: For further information, please contact us at rsilva@gsu.edu and vcalhoun@gsu.edu.
Visit the challenge website coming soon at www.kaggle.com.
