Limits...
Whole body MRI: improved lesion detection and characterization with diffusion weighted techniques.

Attariwala R, Picker W - J Magn Reson Imaging (2013)

Bottom Line: Theory, b-value selection, common artifacts and target to background for optimized viewing will be reviewed for applications in the neck, chest, abdomen, and pelvis.DWI, when used in conjunction with routine imaging, can assist in detecting hemorrhagic degradation products, infection/abscess, and inflammation in colitis, while aiding with discrimination of free fluid and empyema, while limiting the need for intravenous contrast.DWI in conjunction with routine anatomic images provides a platform to improve lesion detection and characterization with findings rivaling other combined anatomic and functional imaging techniques, with the added benefit of no ionizing radiation.

View Article: PubMed Central - PubMed

Affiliation: AIM Medical Imaging, Vancouver, BC, Canada. attariwala@gmail.com

Show MeSH

Related in: MedlinePlus

Diffusion imaging of a cross section through the abdomen at b-values of 0 (a), 200 (b), 800 (c), and 1000(d) s/mm2. At a b-value of 0, note the dark areas representing fluid, flowing or static. At a b-value of 200, flowing fluids are no longer visible. As b-values increase, there is a resultant loss of background tissue, with minimal normal background liver seen at b-value of 1000 (d).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3795449&req=5

fig03: Diffusion imaging of a cross section through the abdomen at b-values of 0 (a), 200 (b), 800 (c), and 1000(d) s/mm2. At a b-value of 0, note the dark areas representing fluid, flowing or static. At a b-value of 200, flowing fluids are no longer visible. As b-values increase, there is a resultant loss of background tissue, with minimal normal background liver seen at b-value of 1000 (d).

Mentions: The log signal intensity versus b-values for a specific tissue can be obtained, and is frequently termed the tissue diffusion curve, which essentially serves as a water-based fingerprint of the tissue being examined 14. The simplest diffusion model assumes tissue is homogeneous, and results in a straight monoexponential line of signal intensity versus b-value. However, the body tissue microenvironment at which water diffusion is measured is inhomogeneous (non-Gaussian), due to cellular compartments, membranes, and vascularity, which result in the tissue diffusion curve being nonlinear. The most pronounced nonlinear effects are due to vascular perfusion, predominantly capillary flow 15,16, and occurs in addition to the bulk diffusion (Figs. 3, and 4]. Thus a bi-exponential tissue diffusion model obtained by a minimum of three b-values resulting in two logarithmic linear components: one for perfusion/flow effects at b-values from 0 to 100 s/mm2 and another for tissue b-values above 100 s/mm2 for bulk diffusion, can suffice for most clinical applications of diffusion body imaging. Because of the additive effects of perfusion and bulk diffusion, there is a steepening of the slope of the diffusion curve that occurs at b-values less than 100 s/mm2 (Fig. 5). Extrapolating and then subtracting the b < 100 s/mm2 bulk diffusion component from the diffusion component, a technique known as curve stripping, at b = 0 yields the flow or perfusion fraction which compares well with in vivo experimental data 17. Using multiple b-values can yield further refinements to the tissue diffusion curve, and this is typically referred to as intravoxel incoherent motion (IVIM). When the IVIM curve for a tissue is obtained, the bi-exponential data can also readily be extracted. More complex methods such as kurtosis, which assesses the probability extent of non-Gaussian behavior, and diffusion tensor imaging (DTI) have also been applied to body tissue diffusion investigation.


Whole body MRI: improved lesion detection and characterization with diffusion weighted techniques.

Attariwala R, Picker W - J Magn Reson Imaging (2013)

Diffusion imaging of a cross section through the abdomen at b-values of 0 (a), 200 (b), 800 (c), and 1000(d) s/mm2. At a b-value of 0, note the dark areas representing fluid, flowing or static. At a b-value of 200, flowing fluids are no longer visible. As b-values increase, there is a resultant loss of background tissue, with minimal normal background liver seen at b-value of 1000 (d).
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3795449&req=5

fig03: Diffusion imaging of a cross section through the abdomen at b-values of 0 (a), 200 (b), 800 (c), and 1000(d) s/mm2. At a b-value of 0, note the dark areas representing fluid, flowing or static. At a b-value of 200, flowing fluids are no longer visible. As b-values increase, there is a resultant loss of background tissue, with minimal normal background liver seen at b-value of 1000 (d).
Mentions: The log signal intensity versus b-values for a specific tissue can be obtained, and is frequently termed the tissue diffusion curve, which essentially serves as a water-based fingerprint of the tissue being examined 14. The simplest diffusion model assumes tissue is homogeneous, and results in a straight monoexponential line of signal intensity versus b-value. However, the body tissue microenvironment at which water diffusion is measured is inhomogeneous (non-Gaussian), due to cellular compartments, membranes, and vascularity, which result in the tissue diffusion curve being nonlinear. The most pronounced nonlinear effects are due to vascular perfusion, predominantly capillary flow 15,16, and occurs in addition to the bulk diffusion (Figs. 3, and 4]. Thus a bi-exponential tissue diffusion model obtained by a minimum of three b-values resulting in two logarithmic linear components: one for perfusion/flow effects at b-values from 0 to 100 s/mm2 and another for tissue b-values above 100 s/mm2 for bulk diffusion, can suffice for most clinical applications of diffusion body imaging. Because of the additive effects of perfusion and bulk diffusion, there is a steepening of the slope of the diffusion curve that occurs at b-values less than 100 s/mm2 (Fig. 5). Extrapolating and then subtracting the b < 100 s/mm2 bulk diffusion component from the diffusion component, a technique known as curve stripping, at b = 0 yields the flow or perfusion fraction which compares well with in vivo experimental data 17. Using multiple b-values can yield further refinements to the tissue diffusion curve, and this is typically referred to as intravoxel incoherent motion (IVIM). When the IVIM curve for a tissue is obtained, the bi-exponential data can also readily be extracted. More complex methods such as kurtosis, which assesses the probability extent of non-Gaussian behavior, and diffusion tensor imaging (DTI) have also been applied to body tissue diffusion investigation.

Bottom Line: Theory, b-value selection, common artifacts and target to background for optimized viewing will be reviewed for applications in the neck, chest, abdomen, and pelvis.DWI, when used in conjunction with routine imaging, can assist in detecting hemorrhagic degradation products, infection/abscess, and inflammation in colitis, while aiding with discrimination of free fluid and empyema, while limiting the need for intravenous contrast.DWI in conjunction with routine anatomic images provides a platform to improve lesion detection and characterization with findings rivaling other combined anatomic and functional imaging techniques, with the added benefit of no ionizing radiation.

View Article: PubMed Central - PubMed

Affiliation: AIM Medical Imaging, Vancouver, BC, Canada. attariwala@gmail.com

Show MeSH
Related in: MedlinePlus