Abstract #2078

Functional Liver Imaging Using Partial Least Squares (PLS) Analysis

Alyaa Hussein Elzibak^1,2, Jeff Fortuna^2,3, John F. MacGregor³, Colm Boylan⁴, Michael D. Noseworthy^2,5

¹Medical Physics and Applied Radiation Sciences, McMaster University, Hamilton, Ontario, Canada; ²Brain-Body Institute, St. Joseph's Healthcare, Hamilton, Ontario, Canada; ³Chemical Engineering, McMaster University, Hamilton, Ontario, Canada; ⁴Diagnostic Imaging, St. Joseph's Healthcare, Hamilton, Ontario, Canada; ⁵Electrical and Computer Engineering, School of Biomedical Engineering, McMaster University, Hamilton, Ontario, Canada

We developed a liver challenge procedure that relies on modulating the BOLD signal with hyperoxia and meal intake. Temporal BOLD data analysis using partial least squares (PLS) was superior to a traditional general linear model (GLM). We show how the method can differentiate healthy from diseased liver tissue.