R_Bibb_Medical_Modelling_The_Application_of_Adv.pdf

Medical imaging for rapid prototyping 23
three-dimensional series scans, the data should be continuous. Noncontinuous
sets of data may be satisfactorily combined in software later,
however as the patient has usually shifted position slightly and the separate
series may not align perfectly.
2.3.3 Missing data
Even with a very small scan distance, some detail may be lost where thin
sections of tissue exist between the scan planes. In addition, parts that are
very small or connected only by thin sections may not survive the build
process.
Due to the time taken to acquire each image, fl owing fl uids will have
moved between the excitation and emission stages of the scan. With multiple
images being taken, this may result in the signal being reduced or reinforced.
Therefore, blood vessels, for example, may appear too dark or too
bright.
2.3.4 Scan distance
This is the distance between the scans taken to form the three-dimensional
scan series (unlike CT data capture is not limited to the axial plane), and
it may also be referred to as ‘pitch’ or ‘distance between cuts’. To maximise
the data available to produce a smooth model this distance should be minimised.
Typically, distances of 1–1.5 mm produce good results. A scan distance
greater than 2 mm will give increasingly poor results as the scan
distance increases. However, taking thinner slices results in less signal
strength per pixel being detected by the scanner. Therefore, more echoes
are required to boost the signal strength, resulting in signifi cantly longer
scan times.
Unlike most CT scanners, the number of pixels used in a cross section
is a variable parameter. Typically, the cross section will be broken down
in to a relatively small number of larger pixels compared to CT. For
example, a typical CT image may be 512 × 512 pixels at a pixel size of
0.5 mm, whereas an MR image may be 256 × 256 pixels at a pixel size of
1 mm. The main reason for this is to maintain signal strength. A larger
number of smaller pixels results in less signal strength per pixel. Once again,
therefore, more echoes are required to boost the signal strength, increasing
scan times.
For three-dimensional modelling, it may be necessary to alter the compromise
between scan time and signal strength compared with the protocols
normally used for imaging only. MR is often a preferred imaging methodology
due to its inherent safety compared to CT. However, the application
of MR for three-dimensional modelling should be carefully considered due
to the increased scan times. Although MR is safe, the procedure may