Limits...
Fine tuning breath-hold-based cerebrovascular reactivity analysis models.

van Niftrik CH, Piccirelli M, Bozinov O, Pangalu A, Valavanis A, Regli L, Fierstra J - Brain Behav (2016)

Bottom Line: All models showed significant differences in CVR and coherence between the affected-hemodynamic impaired-and unaffected hemisphere.Voxel-wise phase determination significantly increases CVR (0.60 ± 0.18 vs. 0.82 ± 0.27; P < 0.05).Our CVR analysis demonstrates an improved CVR and coherence after implementation of voxel-wise phase and frequency adjustment.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurosurgery University Hospital Zurich University of Zurich Clinical Neuroscience Center Frauenklinikstrasse 10 8091 Zurich Switzerland.

ABSTRACT

Introduction: We elaborate on existing analysis methods for breath-hold (BH)-derived cerebrovascular reactivity (CVR) measurements and describe novel insights and models toward more exact CVR interpretation.

Methods: Five blood-oxygen-level-dependent (BOLD) fMRI datasets of neurovascular patients with unilateral hemispheric hemodynamic impairment were used to test various BH CVR analysis methods. Temporal lag (phase), percent BOLD signal change (CVR), and explained variance (coherence) maps were calculated using three different sine models and two novel "Optimal Signal" model-free methods based on the unaffected hemisphere and the sagittal sinus fMRI signal time series, respectively.

Results: All models showed significant differences in CVR and coherence between the affected-hemodynamic impaired-and unaffected hemisphere. Voxel-wise phase determination significantly increases CVR (0.60 ± 0.18 vs. 0.82 ± 0.27; P < 0.05). Incorporating different durations of breath hold and resting period in one sine model (two-task) did increase coherence in the unaffected hemisphere, as well as eliminating negative phase commonly obtained by one-task frequency models. The novel model-free "optimal signal" methods both explained the BOLD MR data similar to the two task sine model.

Conclusions: Our CVR analysis demonstrates an improved CVR and coherence after implementation of voxel-wise phase and frequency adjustment. The novel "optimal signal" methods provide a robust and feasible alternative to the sine models, as both are model-free and independent of compliance. Here, the sagittal sinus model may be advantageous, as it is independent of hemispheric CVR impairment.

No MeSH data available.


Related in: MedlinePlus

Coherence histograms of five analysis models. To evaluate the effectiveness of each model, the mean percent coherence per bins of 10% coherence was determined and presented in two histograms. (A) The histogram portrays the number of voxels explained by the three sine models. From left to right: global‐delay sine (white bars), voxel‐wise‐delay sine (gray bars), and frequency‐Adjusted sine (black bars). The global‐delay sine explains almost a factor of three more voxels with coherence below 10% and is decreasingly present in bins with higher coherence. Compared to the voxel‐wise sine, the frequency‐adjusted sine is only more present in the bins from 70% and higher. (B) The histogram portrays the number of voxels explained by the two “Optimal Signal” models divided in bins of 10% coherence. The unaffected hemisphere (gray bars) has a high number of voxels explained in the bins between 70 and 100%, whereas the sagittal sinus (black bars) explains more voxels in de lower bins.
© Copyright Policy - creativeCommonsBy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4834934&req=5

brb3426-fig-0003: Coherence histograms of five analysis models. To evaluate the effectiveness of each model, the mean percent coherence per bins of 10% coherence was determined and presented in two histograms. (A) The histogram portrays the number of voxels explained by the three sine models. From left to right: global‐delay sine (white bars), voxel‐wise‐delay sine (gray bars), and frequency‐Adjusted sine (black bars). The global‐delay sine explains almost a factor of three more voxels with coherence below 10% and is decreasingly present in bins with higher coherence. Compared to the voxel‐wise sine, the frequency‐adjusted sine is only more present in the bins from 70% and higher. (B) The histogram portrays the number of voxels explained by the two “Optimal Signal” models divided in bins of 10% coherence. The unaffected hemisphere (gray bars) has a high number of voxels explained in the bins between 70 and 100%, whereas the sagittal sinus (black bars) explains more voxels in de lower bins.

Mentions: After application of the coherence threshold of 0.05, the models with a voxel‐wise shift showed a significant increase in the percent of voxels surpassing the threshold, compared to the global sinus (P < 0.05). This global coherence difference divided in bins of 10% can be observed in Figure 3. An obvious difference can be seen between the global‐delay sine and voxel‐wise sine models. The coherence of the affected and unaffected hemisphere is presented in Figure 4. A significant difference in coherence in the unaffected hemisphere between the voxel‐wise‐delay sine and the global‐delay sine was found (P < 0.05). The whole brain and affected hemisphere coherence of the global‐delay sine were not significantly lower than the voxel‐wise‐delay sine, but after a one sample t‐test analysis on the subtracted difference both show a significant increase (whole brain: P = 0.02, unaffected hemisphere: P = 0.03) For the voxel‐wise‐delay sine, a linear regression was performed to find the correlation between average phase and CVR over both hemispheres, shown in Figure 5A. Taken the average phase and CVR of both hemispheres of each subject into account, phase shows a good correlation with CVR (n = 10, R2 = 0.721, P < 0.001).


Fine tuning breath-hold-based cerebrovascular reactivity analysis models.

van Niftrik CH, Piccirelli M, Bozinov O, Pangalu A, Valavanis A, Regli L, Fierstra J - Brain Behav (2016)

Coherence histograms of five analysis models. To evaluate the effectiveness of each model, the mean percent coherence per bins of 10% coherence was determined and presented in two histograms. (A) The histogram portrays the number of voxels explained by the three sine models. From left to right: global‐delay sine (white bars), voxel‐wise‐delay sine (gray bars), and frequency‐Adjusted sine (black bars). The global‐delay sine explains almost a factor of three more voxels with coherence below 10% and is decreasingly present in bins with higher coherence. Compared to the voxel‐wise sine, the frequency‐adjusted sine is only more present in the bins from 70% and higher. (B) The histogram portrays the number of voxels explained by the two “Optimal Signal” models divided in bins of 10% coherence. The unaffected hemisphere (gray bars) has a high number of voxels explained in the bins between 70 and 100%, whereas the sagittal sinus (black bars) explains more voxels in de lower bins.
© Copyright Policy - creativeCommonsBy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4834934&req=5

brb3426-fig-0003: Coherence histograms of five analysis models. To evaluate the effectiveness of each model, the mean percent coherence per bins of 10% coherence was determined and presented in two histograms. (A) The histogram portrays the number of voxels explained by the three sine models. From left to right: global‐delay sine (white bars), voxel‐wise‐delay sine (gray bars), and frequency‐Adjusted sine (black bars). The global‐delay sine explains almost a factor of three more voxels with coherence below 10% and is decreasingly present in bins with higher coherence. Compared to the voxel‐wise sine, the frequency‐adjusted sine is only more present in the bins from 70% and higher. (B) The histogram portrays the number of voxels explained by the two “Optimal Signal” models divided in bins of 10% coherence. The unaffected hemisphere (gray bars) has a high number of voxels explained in the bins between 70 and 100%, whereas the sagittal sinus (black bars) explains more voxels in de lower bins.
Mentions: After application of the coherence threshold of 0.05, the models with a voxel‐wise shift showed a significant increase in the percent of voxels surpassing the threshold, compared to the global sinus (P < 0.05). This global coherence difference divided in bins of 10% can be observed in Figure 3. An obvious difference can be seen between the global‐delay sine and voxel‐wise sine models. The coherence of the affected and unaffected hemisphere is presented in Figure 4. A significant difference in coherence in the unaffected hemisphere between the voxel‐wise‐delay sine and the global‐delay sine was found (P < 0.05). The whole brain and affected hemisphere coherence of the global‐delay sine were not significantly lower than the voxel‐wise‐delay sine, but after a one sample t‐test analysis on the subtracted difference both show a significant increase (whole brain: P = 0.02, unaffected hemisphere: P = 0.03) For the voxel‐wise‐delay sine, a linear regression was performed to find the correlation between average phase and CVR over both hemispheres, shown in Figure 5A. Taken the average phase and CVR of both hemispheres of each subject into account, phase shows a good correlation with CVR (n = 10, R2 = 0.721, P < 0.001).

Bottom Line: All models showed significant differences in CVR and coherence between the affected-hemodynamic impaired-and unaffected hemisphere.Voxel-wise phase determination significantly increases CVR (0.60 ± 0.18 vs. 0.82 ± 0.27; P < 0.05).Our CVR analysis demonstrates an improved CVR and coherence after implementation of voxel-wise phase and frequency adjustment.

View Article: PubMed Central - PubMed

Affiliation: Department of Neurosurgery University Hospital Zurich University of Zurich Clinical Neuroscience Center Frauenklinikstrasse 10 8091 Zurich Switzerland.

ABSTRACT

Introduction: We elaborate on existing analysis methods for breath-hold (BH)-derived cerebrovascular reactivity (CVR) measurements and describe novel insights and models toward more exact CVR interpretation.

Methods: Five blood-oxygen-level-dependent (BOLD) fMRI datasets of neurovascular patients with unilateral hemispheric hemodynamic impairment were used to test various BH CVR analysis methods. Temporal lag (phase), percent BOLD signal change (CVR), and explained variance (coherence) maps were calculated using three different sine models and two novel "Optimal Signal" model-free methods based on the unaffected hemisphere and the sagittal sinus fMRI signal time series, respectively.

Results: All models showed significant differences in CVR and coherence between the affected-hemodynamic impaired-and unaffected hemisphere. Voxel-wise phase determination significantly increases CVR (0.60 ± 0.18 vs. 0.82 ± 0.27; P < 0.05). Incorporating different durations of breath hold and resting period in one sine model (two-task) did increase coherence in the unaffected hemisphere, as well as eliminating negative phase commonly obtained by one-task frequency models. The novel model-free "optimal signal" methods both explained the BOLD MR data similar to the two task sine model.

Conclusions: Our CVR analysis demonstrates an improved CVR and coherence after implementation of voxel-wise phase and frequency adjustment. The novel "optimal signal" methods provide a robust and feasible alternative to the sine models, as both are model-free and independent of compliance. Here, the sagittal sinus model may be advantageous, as it is independent of hemispheric CVR impairment.

No MeSH data available.


Related in: MedlinePlus