Blurring of the CT image can be described by the PSF of a system [3,
4].
The PSF is assumed separable into a two-dimensional (2D) PSF in the x-y scan plane and the SSP in the z-direction perpendicular to the scan plane [5-8].
Then,
the three dimensional (3D) CT image I (x,
y,
z) can be expressed as [4,
6,
7].
I(x,y,z) = [O(x,y,z)**PSF(x,y)]*SSP(z) (1)
where O(x,
y,
z) is the object function,
PSF(x,
y) and SSP(z) are the 2D PSF and slice sensitivity profile,
respectively; and ** and * are 2D and 1D convolutions,
respectively.
Noise and artifact components are neglected in this representation.
Screening images of two different screening centers were used for the study.
Plaka Healthcare Center at Niigata has done the screening with thin slices.
In this screening clinic,
applicants have been subjected to lung cancer CT screening examinations with a multidetector-row CT scanner without contrast enhancement (Aquilion,
Toshiba Medical Systems,
Tokyo,
Japan).
The imaging parameters were as follows: reconstruction kernel FC51,
120 kV,
60 mA tube current,
1.0 mm slice thickness,
16 detector rows,
0.5 s rotation time and a pitch factor of 0.9375.
Azumi General Hospital at Nagano has done screening with thick slices.
Screenings have done with a multidetector-row CT scanner without contrast enhancement (Asteion,
Toshiba Medical Systems,
Tokyo,
Japan).
The imaging parameters were as follows: reconstruction kernel FC50,
120 kV,
30 mA tube current,
8.0 mm slice thickness,
4 detector rows,
0.75 s rotation time and a pitch factor of 1.375.
Sets of five healthy subjects’ screening images were selected from both clinics for the study.
Approval of the institutional review board was taken from both clinics for accessing the database.
PSF and SSP of both screening clinics were measured with routine scan/reconstruction settings and used in simulation process.
Reference virtual nodule set consists of 150 nodules with object function diameter 6 mm and density (ΔCT) 300 HU.
∆CT was defined as the density difference between background and the object function.
Separate reference nodule sets were generated for each clinic based on measured PSF and SSP.
For the accuracy simulated images were computed with a fine digital sampling pitch.
However,
clinical images do not have such fine intervals as in simulated images.
Superimposing process needs the similar intervals on both simulated and screening images.
Therefore,
resampling should be performed before superimposing the simulated images on clinical images.
Then,
the simulated blurred images were transformed into practical images that can be used in clinical evaluations [9,
10].Then set of 30 nodules were superimposed on images of each subject.
The whole lung field was covered selecting three levels from upper,
middle and lower lung Figure 1.
Finally,
CAD detections were performed (Figures 2and 3) and FROC curves were obtained (Figure 4).