William A. Grissom1, Lorne W. Hofstetter2,
Viola Rieke3, Yoav Medan4, Kim Butts Pauly3,
Cynthia E. Davis2
1GE Global Research,
Munich, Germany; 2GE Global Research, Niskayuna, NY, United
States; 3Radiology, Stanford University, Stanford, CA, United
States; 4Insightec, Ltd, Tirat Carmel, Israel
Proton resonance frequency-shift thermometry is a promising technique to monitor thermal therapies in organs such as the breast and prostate, but is complicated by the presence of fat, organ motion, and the small size of these organs relative to typical hot spot sizes. While thermometry can be performed by suppressing fat, it can also be exploited as a reference to improve the separation of phase shifts caused by time-varying main field changes from temperature-induced phase shifts. We introduce an extension to the hybrid multibaseline subtraction and referenceless thermometry method that exploits the fat image as a reference to yield accurate temperature estimates that are robust to main field changes, large hot spot sizes, and motion in organs such as the prostate and breast that are surrounded by or contain fat.