Limits...
Lateralization of Temporal Lobe Epilepsy Based on Resting-State Functional Magnetic Resonance Imaging and Machine Learning.

Yang Z, Choupan J, Reutens D, Hocking J - Front Neurol (2015)

Bottom Line: A comprehensive feature space was constructed to include network properties within local brain regions, between brain regions, and across the whole network.A leave-one-patient-out cross validation was carried out on 12 patients and a prediction accuracy of 83% was achieved.The importance of selected features was analyzed to demonstrate the contribution of resting-state connectivity attributes at voxel, region, and network levels to TLE lateralization.

View Article: PubMed Central - PubMed

Affiliation: School of Information Technology and Electrical Engineering, The University of Queensland , Brisbane, QLD , Australia ; Centre for Advanced Imaging, The University of Queensland , Brisbane, QLD , Australia.

ABSTRACT
Lateralization of temporal lobe epilepsy (TLE) is critical for successful outcome of surgery to relieve seizures. TLE affects brain regions beyond the temporal lobes and has been associated with aberrant brain networks, based on evidence from functional magnetic resonance imaging. We present here a machine learning-based method for determining the laterality of TLE, using features extracted from resting-state functional connectivity of the brain. A comprehensive feature space was constructed to include network properties within local brain regions, between brain regions, and across the whole network. Feature selection was performed based on random forest and a support vector machine was employed to train a linear model to predict the laterality of TLE on unseen patients. A leave-one-patient-out cross validation was carried out on 12 patients and a prediction accuracy of 83% was achieved. The importance of selected features was analyzed to demonstrate the contribution of resting-state connectivity attributes at voxel, region, and network levels to TLE lateralization.

No MeSH data available.


Related in: MedlinePlus

Clusters with group difference in voxelwise properties of resting-state connectivity plotted in blue using xjView. (A) ReHo with four clusters and (B) fALFF with one cluster. No cluster was found in ALFF.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4553409&req=5

Figure 1: Clusters with group difference in voxelwise properties of resting-state connectivity plotted in blue using xjView. (A) ReHo with four clusters and (B) fALFF with one cluster. No cluster was found in ALFF.

Mentions: The results of group comparison showed no region with group-wise difference in ALFF. The clusters with significant group difference in fALFF and ReHo are plotted in Figure 1 using xjView (http://www.alivelearn.net/xjview8/). The AAL ROIs containing these clusters are listed in Table 3. The AAL ROIs, MNI coordinates and scores of relative importance of the top five ranked selected features of ALFF, fALFF, and ReHo are in Table 4. Note that only 1 out of the 15 top ranked features was in AAL ROIs that had group difference, which was ReHo of a voxel in right middle frontal gyrus.Ho.


Lateralization of Temporal Lobe Epilepsy Based on Resting-State Functional Magnetic Resonance Imaging and Machine Learning.

Yang Z, Choupan J, Reutens D, Hocking J - Front Neurol (2015)

Clusters with group difference in voxelwise properties of resting-state connectivity plotted in blue using xjView. (A) ReHo with four clusters and (B) fALFF with one cluster. No cluster was found in ALFF.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4553409&req=5

Figure 1: Clusters with group difference in voxelwise properties of resting-state connectivity plotted in blue using xjView. (A) ReHo with four clusters and (B) fALFF with one cluster. No cluster was found in ALFF.
Mentions: The results of group comparison showed no region with group-wise difference in ALFF. The clusters with significant group difference in fALFF and ReHo are plotted in Figure 1 using xjView (http://www.alivelearn.net/xjview8/). The AAL ROIs containing these clusters are listed in Table 3. The AAL ROIs, MNI coordinates and scores of relative importance of the top five ranked selected features of ALFF, fALFF, and ReHo are in Table 4. Note that only 1 out of the 15 top ranked features was in AAL ROIs that had group difference, which was ReHo of a voxel in right middle frontal gyrus.Ho.

Bottom Line: A comprehensive feature space was constructed to include network properties within local brain regions, between brain regions, and across the whole network.A leave-one-patient-out cross validation was carried out on 12 patients and a prediction accuracy of 83% was achieved.The importance of selected features was analyzed to demonstrate the contribution of resting-state connectivity attributes at voxel, region, and network levels to TLE lateralization.

View Article: PubMed Central - PubMed

Affiliation: School of Information Technology and Electrical Engineering, The University of Queensland , Brisbane, QLD , Australia ; Centre for Advanced Imaging, The University of Queensland , Brisbane, QLD , Australia.

ABSTRACT
Lateralization of temporal lobe epilepsy (TLE) is critical for successful outcome of surgery to relieve seizures. TLE affects brain regions beyond the temporal lobes and has been associated with aberrant brain networks, based on evidence from functional magnetic resonance imaging. We present here a machine learning-based method for determining the laterality of TLE, using features extracted from resting-state functional connectivity of the brain. A comprehensive feature space was constructed to include network properties within local brain regions, between brain regions, and across the whole network. Feature selection was performed based on random forest and a support vector machine was employed to train a linear model to predict the laterality of TLE on unseen patients. A leave-one-patient-out cross validation was carried out on 12 patients and a prediction accuracy of 83% was achieved. The importance of selected features was analyzed to demonstrate the contribution of resting-state connectivity attributes at voxel, region, and network levels to TLE lateralization.

No MeSH data available.


Related in: MedlinePlus