Meeting Banner
Abstract #2692

Separation of Signal and Noise in Dynamic MRI Data Using the Kolmogorov-Smirnov Test

David S. Smith1, Stephanie Barnes1, Thomas E. Yankeelov1

1Vanderbilt University, Nashville, TN, United States

We present preliminary efforts that indicate that the Kolmogorov-Smirnov statistical test may be an extremely useful method for automatically separating signal from noise in dynamic imaging data, especially when aliased power should be captured but noise should be ignored. We compare to Otsu's method and demonstrate an improved automatic classification of signal and noise in in vivo tumor-bearing mouse data.

Keywords

accuracy acquisition aliased almost applications approximately arbitrarily artifacts assumed audience automatically away black bolus breast cancer capture captured catheterized cells characteristics classifying collected collection compressed considered contained containing contrast curve cutoff cybernetics density describing determine distinguish distribution distributions domain dynamic dynamics edge efforts empirical employed entire equality error especially even exact experiment extremely fact female generated generation gradient green ignored ignoring include incorrectly indeed indicate indicates injected injection insensitive inside intensity intraclass know labeled limit lumped lumping machine magnetization magnitude marked mask masks matrix measured minimizing minutes missed moments mouse need noise nude ones original outside overlaid panel power precision preferably preliminary produced properties protocol pure quality random reached reconstructions reflect required requires researchers resistant root sample sampling scanner scientists sensing separating separation simplicity smith solid square statistical subcutaneously superior suspect systems taken target threshold thresholds trans tumor turn uncorrelated underlying unintentionally unknown upper uptake useful usually variance volume whether zero