Information Technology Reference
In-Depth Information
10 Image Quality
Image quality is a function of the interplay between the image noise, contrast, and
spatial resolution. For purposes of brevity and illustration, we will limit our dis-
cussion here to CT and MR imaging modalities.
In an ideal medical image, tissue which is homogeneous in density or signal
intensity would have a uniform gray-scale appearance. However,
images
contain noise, arising from a number sources including thermally based radiofre-
quency energy emission in MRI, scattering of X-ray photons in CT, and due to
detector noise. This noise is represented as
real
fluctuations of pixel intensity units about
some intensity average for a homogeneous tissue. The ability to assess the object or
structure of interest (displayed in the image by detection of the signal from the
object of interest) may be limited by the noise, which degrades the image quality
and also the object representation. Thus, the magnitude of the noise relative to the
signal of the image is an important concept in medical image interpretation and
analysis. The power of the image signal is derived from the square of the average
pixel intensity. The image noise is determined from the standard deviation of the
pixel intensities in a region of the image with homogeneous signal intensity. The
ratio of the power of the image signal to the power of the noise is the signal-to-noise
ratio (SNR) [ 25 ].
For a particular imaging modality, the signal-to-noise ratio may be improved by
adjustment of scanning parameters for a particular machine, or by the use of a
machine with different inherent technology. For instance, the SNR for magnetic
resonance imaging may be increased by using a machine with a higher intrinsic
magnetic
fl
eld
strength versus 1.5 T. SNR may also be increased in a particular machine by
adjustment of scanning parameters, such as decreasing the matrix size, increasing
the section thickness, decreasing the RF bandwidth, and increasing the number of
image acquisitions [ 26 , 27 ]. In CT imaging, the SNR may be increased by
increasing the mA, kVp, voxel size, and scan time [ 26 ].
The image contrast in medical images may be thought of as the difference in
pixel intensity between different anatomic structures. In CT imaging, the contrast is
in general increased by administering iodine-based contrast media (for soft tissues),
and decreasing the kVp. MR imaging contrast is connected to the magnetic prop-
erties of the body tissues being imaged with the particular MR sequence being used,
such as T1, versus PD, versus T2. As with CT imaging, the contrast of the MR
images may be increased by the administration of contrast material, in this case,
gadolinium based materials.
The spatial resolution of an image is related to voxel size, and is determined by
the ability to discriminate adjacent anatomic structures. When performing a CT
scan of a patient, the in-plane spatial resolution (axial plane, perpendicular to the
long axis of the scanner) may be increased by decreasing the
field, which may be accomplished by using a machine with a 3.0 T
field of view and
Search WWH ::




Custom Search