Traditional Posters: Body Imaging

Claudia Lenz¹, Gregor Sommer², Klaus Scheffler¹, Leopold Winter², Markus Klarhöfer<sup>1

1</sup>Radiological Physics, University of Basel Hospital, Basel, Switzerland; ²Department of Radiology, University of Basel Hospital, Basel, Switzerland

In biologic tissues, microscopic motion of water not only includes molecular diffusion, but also microcirculation of blood in the capillary network. The intraxovel incoherent motion model has been introduced to describe these combined diffusion and microcirculation effects in diffusion weighted imaging. Analysis of the multicompartmental water diffusion is mostly performed by applying a biexponential fit function to the diffusion curve and evaluating the diffusion and perfusion components separately. However, this technique often suffers from high standard fit errors, especially for the perfusion fraction f. In 2003, Bennett et al. proposed a stretched exponential model to account for the multiexponential behavior of diffusion curves in the brain. In this work, we extended the stretched exponential model to the abdomen and present fit results from the kidneys of healthy subjects.

2650. Interference of Inversion Recovery with Diffusion Weighted Imaging: Negative Apparent Diffusion Coefficients!

Thomas Gaass¹, Bram Stieltjes, Frederik Laun<sup>1

1</sup>Medical Physics in Radiology, German Cancer Research Center, Heidelberg, Germany

The aim of this work was to evaluate whether diffusion weighted imaging (DWI) and inversion recovery (IR) may be applied without interference. DWI of the liver shows that the apparent diffusion coefficient (ADC) measured in a single voxel is clearly dependent on the inversion time and ADCs vary between -0.007 s/mm² and 0.009 s/mm² in a 100 ms TI interval. The counterintuitive negative diffusion is observed in the liver and in regions with incomplete fat saturation. This can be explained by the here proposed two compartment model. Thus, DWI and IR can generally not be applied without interference.

2651. Diffusion-Weighted Imaging of the Kidney: Beyond Mono- And Bi-Exponential Models

Anna Caroli¹, Luca Antiga¹, Giuseppe Petralia², Massimo Bellomi², Andrea Remuzzi^1,3, Paul Summers<sup>2

1</sup>Bioengineering Department, Mario Negri Institute for Pharmacological Research, Bergamo, Italy; ²Division of Radiology, European Institute of Oncology, Milan, Italy; ³University of Bergamo, Bergamo, Italy

Mono-exponential models do not accurately predict diffusion-weighted signal decay in the kidney, while bi-exponential models are unable to differentiate contributions. We propose a “piece-wise” exponential model, separately fitting low and high b-values with two exponentials, expressive of fast and slow transport components. Ten healthy volunteers underwent DWI both pre- and post-lunch, and acquisitions were repeated within one of the two sessions. The model was stable, and accurately fit signal attenuation. Diffusion parameters showed high repeatability, but significant differences between pre- and post-meal acquisitions. These results point out the need for more complete interpretations of DWI signal in describing the complex transport in the kidney.

2652. Impact of Low and High B-Value MR Diffusion in HIV/HCV-Coinfected, HIV-Monoinfected and Uninfected Subjects

Susan Moyher Noworolski^1,2, Phyllis Tien^3,4, Michelle Nystrom¹, Suchandrima Banerjee⁵, Aliya Qayyum<sup>1

1</sup>Radiology and Biomedical Imaging, University of California, San Francisco, CA, United States; ²The Graduate Group in Bioengineering, University of California, San Francisco and Berkeley, CA, United States; ³Medicine, University of California, San Francisco, CA, United States; ⁴Medicine, Veteran Affairs Medical Center, San Francisco, CA, United States; ⁵MR Applied Science Lab, GE Healthcare, Menlo Park, CA, United States

The impact of a perfusion regime, low b-value ADC, and a tissue regime, high b-value ADC were evaluated in comparison to a conventional ADC in three groups of subjects: HIV/HCV (hepatitis C) coinfection, HIV-monoinfection, and without infection. Liver ADC was measured using b values of 0 and 150 (ADC^low), 150 and 600 (ADC^high) and 0 and 600 (ADC^conv) in one breathhold sequence. ADC^low and ADC^high provided unique information. HIV tended to have the highest ADC levels and was significantly higher than HIV/HCV for ADC^low and ADC^conv. HIV status may thus be an important consideration in interpretation of liver ADC.

2653. SSFP Diffusion Prepared SSFSE