Integration of multiple MRI modalities improves prediction of cognitive ability

Integration of multiple MRI modalities improves prediction of cognitive ability

Predicting cognitive ability from brain imaging has long been a central goal in cognitive neuroscience. While machine learning has modestly improved predictions using brain MRI data, most studies rely on a single MRI modality. Narun Pat and colleagues have integrated multiple MRI modalities through a technique called stacking.

The method combines structural MRI (e.g. cortical thickness), rest state and task-based functional connectivity, as well as task-evoked blood oxygen level-dependent (BOLD) contrasts to create a more robust neural marker of cognitive function. The authors analyzed data from 2,131 participants aged 22 to 100 from three large MRI datasets in the United States and New Zealand. Across the three data sets, stacking consistently and significantly improved predictions of cognitive test scores collected off the scanner. To assess whether stacking could capture stable cognitive traits, the authors applied the method to the Dunedin Multidisciplinary Health and Development Study. Using brain imaging at age 45, the model predicted cognitive scores in childhood (ages 7, 9, and 11) with a correlation of 0.52, recognizing a significant level of predictive accuracy. Stacking also faced a major challenge in MRI-based models: test-retest reliability—the stability of individual rankings over time. The improved consistency suggests that stacking allows MRI data to capture more persistent individual differences in cognitive ability than single MRI modality models.

Finally, the researchers evaluated the generalizability of stacking by training on one data set and testing on a separate, independent data set. Due to differences in task protocols, the authors were unable to include several important MRI modalities. Nevertheless, the model achieved a Pearson correlation of 0.25. Although this was lower than the performance within the data set, the correlation still demonstrated a meaningful degree of cross-sample applicability. According to the authors, the study sets a valuable benchmark for how stacking can strengthen the use of brain MRI as a reliable and robust neural marker of cognitive function.

Sources: