A Local Linear Regression Algorithm for Partial Volume Correction in Brain Oxygen Extraction Fraction Estimation
Yasheng Chen1, Chunwei Ying2, Peter Kang1, Slim Fellah1, Amy Mirro3, Melanie Fields3, Kristin Guilliams1, Jin-Moo Lee1, Andria Ford1, and Hongyu An4
1Neurology, Washington University School of Medicine, St. Louis, MO, United States, 2Biomedical Engineering, Washington University in St. Louis, St. Louis, MO, United States, 3Pediatrics Hematology, Washington University School of Medicine, St. Louis, MO, United States, 4Washington University School of Medicine, St. Louis, MO, United States
We developed
a local linear regression based partial volume correction (PVC) method for OEF
estimation. The PVC OEF method reduced model fitting errors and signal
contamination of CSF, and it improved the association between OEF and WMH
lesion burden.
Figure 2. The group average rOEF maps
without (A)
and with (B)
PVC, and the
group average fitting
error maps
without (C)
and with (D) PVC.
Figure 3. Correlation between WM rOEF and
rVWMH with and without PVC (A, B), and correlation between watershed
rOEF
and rVWMH with and without PVC (C, D).