Limits...
Predicting individuals' learning success from patterns of pre-learning MRI activity.

Vo LT, Walther DB, Kramer AF, Erickson KI, Boot WR, Voss MW, Prakash RS, Lee H, Fabiani M, Gratton G, Simons DJ, Sutton BP, Wang MY - PLoS ONE (2011)

Bottom Line: These predictions explained more than half of the variance in learning success among individuals, suggesting that individual differences in neuroanatomy or persistent physiology predict whether and to what extent people will benefit from training in a complex task.Surprisingly, predictions from white matter were highly accurate, while voxels in the gray matter of the dorsal striatum did not contain any information about future training success.The ability to predict who will benefit from training by using neuroimaging data collected in the early training phase may have far-reaching implications for the assessment of candidates for specific training programs as well as the study of populations that show deficiencies in learning new skills.

View Article: PubMed Central - PubMed

Affiliation: Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, Illinois, United States of America.

ABSTRACT
Performance in most complex cognitive and psychomotor tasks improves with training, yet the extent of improvement varies among individuals. Is it possible to forecast the benefit that a person might reap from training? Several behavioral measures have been used to predict individual differences in task improvement, but their predictive power is limited. Here we show that individual differences in patterns of time-averaged T2*-weighted MRI images in the dorsal striatum recorded at the initial stage of training predict subsequent learning success in a complex video game with high accuracy. These predictions explained more than half of the variance in learning success among individuals, suggesting that individual differences in neuroanatomy or persistent physiology predict whether and to what extent people will benefit from training in a complex task. Surprisingly, predictions from white matter were highly accurate, while voxels in the gray matter of the dorsal striatum did not contain any information about future training success. Prediction accuracy was higher in the anterior than the posterior half of the dorsal striatum. The link between trainability and the time-averaged T2*-weighted signal in the dorsal striatum reaffirms the role of this part of the basal ganglia in learning and executive functions, such as task-switching and task coordination processes. The ability to predict who will benefit from training by using neuroimaging data collected in the early training phase may have far-reaching implications for the assessment of candidates for specific training programs as well as the study of populations that show deficiencies in learning new skills.

Show MeSH

Related in: MedlinePlus

Predicting score improvement from MRI activity in the dorsal striatum.(A) Correlation of measured score improvement with the spatial mean of the time-averaged T2*-weighted signal in the dorsal striatum. Mean activity of 34 subjects is significantly correlated with score improvement. (B) Correlation of measured score improvements with score improvement predicted from multi-voxel patterns of the T2*-weighted signal in the dorsal striatum. It shows an even higher correlation than in A). The dashed lines show the least-squares best linear fits in figures A and B. **p<0.01, ***p<0.001.
© Copyright Policy
Related In: Results  -  Collection


getmorefigures.php?uid=PMC3021541&req=5

pone-0016093-g002: Predicting score improvement from MRI activity in the dorsal striatum.(A) Correlation of measured score improvement with the spatial mean of the time-averaged T2*-weighted signal in the dorsal striatum. Mean activity of 34 subjects is significantly correlated with score improvement. (B) Correlation of measured score improvements with score improvement predicted from multi-voxel patterns of the T2*-weighted signal in the dorsal striatum. It shows an even higher correlation than in A). The dashed lines show the least-squares best linear fits in figures A and B. **p<0.01, ***p<0.001.

Mentions: For the spatial mean activity analysis, we averaged the intensity of all voxels inside an anatomically defined region. As a first test, we divided subjects into groups of good and poor learners based on a median split of their score improvements. We found significantly higher mean activity for good than poor learners in the dorsal striatum (p = 0.011), but not in the ventral striatum (p = 0.75, two-sample t tests with n1 = n2 = 17). To determine the relationship between subjects' numerical score improvements and mean activity within an ROI we computed their Pearson correlation. In the dorsal striatum, the correlation was significant (r = 0.47, p = 0.0053; see Figure 2A), but again not in the ventral striatum (r = −0.09).


Predicting individuals' learning success from patterns of pre-learning MRI activity.

Vo LT, Walther DB, Kramer AF, Erickson KI, Boot WR, Voss MW, Prakash RS, Lee H, Fabiani M, Gratton G, Simons DJ, Sutton BP, Wang MY - PLoS ONE (2011)

Predicting score improvement from MRI activity in the dorsal striatum.(A) Correlation of measured score improvement with the spatial mean of the time-averaged T2*-weighted signal in the dorsal striatum. Mean activity of 34 subjects is significantly correlated with score improvement. (B) Correlation of measured score improvements with score improvement predicted from multi-voxel patterns of the T2*-weighted signal in the dorsal striatum. It shows an even higher correlation than in A). The dashed lines show the least-squares best linear fits in figures A and B. **p<0.01, ***p<0.001.
© Copyright Policy
Related In: Results  -  Collection

Show All Figures
getmorefigures.php?uid=PMC3021541&req=5

pone-0016093-g002: Predicting score improvement from MRI activity in the dorsal striatum.(A) Correlation of measured score improvement with the spatial mean of the time-averaged T2*-weighted signal in the dorsal striatum. Mean activity of 34 subjects is significantly correlated with score improvement. (B) Correlation of measured score improvements with score improvement predicted from multi-voxel patterns of the T2*-weighted signal in the dorsal striatum. It shows an even higher correlation than in A). The dashed lines show the least-squares best linear fits in figures A and B. **p<0.01, ***p<0.001.
Mentions: For the spatial mean activity analysis, we averaged the intensity of all voxels inside an anatomically defined region. As a first test, we divided subjects into groups of good and poor learners based on a median split of their score improvements. We found significantly higher mean activity for good than poor learners in the dorsal striatum (p = 0.011), but not in the ventral striatum (p = 0.75, two-sample t tests with n1 = n2 = 17). To determine the relationship between subjects' numerical score improvements and mean activity within an ROI we computed their Pearson correlation. In the dorsal striatum, the correlation was significant (r = 0.47, p = 0.0053; see Figure 2A), but again not in the ventral striatum (r = −0.09).

Bottom Line: These predictions explained more than half of the variance in learning success among individuals, suggesting that individual differences in neuroanatomy or persistent physiology predict whether and to what extent people will benefit from training in a complex task.Surprisingly, predictions from white matter were highly accurate, while voxels in the gray matter of the dorsal striatum did not contain any information about future training success.The ability to predict who will benefit from training by using neuroimaging data collected in the early training phase may have far-reaching implications for the assessment of candidates for specific training programs as well as the study of populations that show deficiencies in learning new skills.

View Article: PubMed Central - PubMed

Affiliation: Beckman Institute, University of Illinois at Urbana-Champaign, Urbana, Illinois, United States of America.

ABSTRACT
Performance in most complex cognitive and psychomotor tasks improves with training, yet the extent of improvement varies among individuals. Is it possible to forecast the benefit that a person might reap from training? Several behavioral measures have been used to predict individual differences in task improvement, but their predictive power is limited. Here we show that individual differences in patterns of time-averaged T2*-weighted MRI images in the dorsal striatum recorded at the initial stage of training predict subsequent learning success in a complex video game with high accuracy. These predictions explained more than half of the variance in learning success among individuals, suggesting that individual differences in neuroanatomy or persistent physiology predict whether and to what extent people will benefit from training in a complex task. Surprisingly, predictions from white matter were highly accurate, while voxels in the gray matter of the dorsal striatum did not contain any information about future training success. Prediction accuracy was higher in the anterior than the posterior half of the dorsal striatum. The link between trainability and the time-averaged T2*-weighted signal in the dorsal striatum reaffirms the role of this part of the basal ganglia in learning and executive functions, such as task-switching and task coordination processes. The ability to predict who will benefit from training by using neuroimaging data collected in the early training phase may have far-reaching implications for the assessment of candidates for specific training programs as well as the study of populations that show deficiencies in learning new skills.

Show MeSH
Related in: MedlinePlus