Intravoxel Incoherent Motion Quantification Dependent on Measurement SNR and Tissue Perfusion: A Simulation Study

The intravoxel incoherent motion (IVIM) model extracts both functional and structural information of a tissue using motion-sensitizing gradients. The Objective of the present work is to investigate the impact of signal to noise ratio (SNR) and physiologic conditions on the validity of IVIM parameters. This study is a simulation study, modeling IVIM at a voxel, and also done 10,000 times for every single simulation. Complex noises with various standard deviations were added to signal in-silico to investigate SNR effects on output validity. Besides, some blood perfusion situations for different tissues were considered based on their physiological range to explore the impacts of blood fraction at each voxel on the validity of the IVIM outputs. Coefficient variation (CV) and bias of the estimations were computed to assess the validity of the IVIM parameters. This study has shown that the validity of IVIM output parameters highly depends on measurement SNR and physiologic characteristics of the studied organ.   IVIM imaging could be useful if imaging parameters are correctly selected for each specific organ, considering hardware limitations.