Cj. Sutherland et De. Gayou, ARTIFACTS AND THRESHOLDING IN X-RAY CT OF A CORTICAL BONE AND TITANIUM COMPOSITE, Journal of computer assisted tomography, 20(3), 1996, pp. 496-503
Purpose: X-Ray CT has the potential to provide precise and accurate da
ta from which the mechanical properties of bone can be calculated. Suc
h data would be useful in understanding the response of bone tissue to
implants. Various artifacts can, however, degrade accuracy of the dat
a. The aim of this study was to measure the artifactual errors produce
d in CT of a simulated femoral bone-titanium composite and to propose
a method to correct for them. Method: A composite phantom that simulat
es cortical bone and a titanium implant was designed and constructed.
The phantom was scanned and the image data were analyzed over a range
of thresholds with image analysis software developed for this study. T
he outer (OD) and inner (ID) diameter and the CT number of the cortica
l bone, with and without titanium, were measured over a range of corti
cal thicknesses. Results: While ID can be accurately measured by choos
ing the proper threshold (800 HU), OD, even at optimal threshold, will
be underestimated by similar to 2%, If a proper threshold is selected
, CT number can also be accurately determined. Errors of up to 5%, how
ever, are produced by titanium unless corrected by proper threshold se
lection. Conclusion: Intramedullary titanium is not a deterrent to obt
aining accurate measurements of cortical bone dimensions and propertie
s. Proper choice of thresholds for image analysis of CT scan data can
yield accuracy and precision of less than or equal to 2%.