Towards task-based assessment of CT performance: System and object

MTF across different reconstruction algorithms

Samuel Richarda)
Carl E. Ravin Advanced Imaging Laboratories, Department of Radiology, Duke University,
Durham, North Carolina 27705
Daniela B. Husarik
Duke University Medical Center, Department of Radiology, Duke University, Durham, North Carolina 27705

Girijesh Yadava
GE Healthcare, Waukesha, Wisconsin 53188
Simon N. Murphy
Carl E. Ravin Advanced Imaging Laboratories, Department of Radiology, Duke University,
Durham, North Carolina 27705

Ehsan Samei
Carl E. Ravin Advanced Imaging Laboratories, Department of Radiology, Duke University,
Durham, North Carolina 27705
(Received 21 July 2011; revised 15 May 2012; accepted for publication 16 May 2012; published 11
June 2012)
Purpose: To investigate a measurement method for evaluating the resolution properties of CT imag-
ing systems across reconstruction algorithms, dose, and contrast.
Methods: An algorithm was developed to extract the task-based modulation transfer function (MTF)
from disk images generated from the rod inserts in the ACR phantom (model 464 Gammex, WI).
These inserts are conventionally employed for HU accuracy assessment. The edge of the disk objects
was analyzed to determine the edge-spread function, which was differentiated to yield the line-spread
function and Fourier-transformed to generate the object-specific MTF for task-based assessment, de-
noted MTFTask . The proposed MTF measurement method was validated against the conventional wire
technique and further applied to measure the MTF of CT images reconstructed with an adaptive statis-
tical iterative algorithm (ASIR) and a model-based iterative (MBIR) algorithm. Results were further
compared to the standard filtered back projection (FBP) algorithm. Measurements were performed
and compared across different doses and contrast levels to ascertain the MTFTask dependencies on
those factors.
Results: For the FBP reconstructed images, the MTFTask measured with the inserts were the same as
the MTF measured from the wire-based method. For the ASIR and MBIR data, the MTFTask using the
high contrast insert was similar to the wire-based MTF and equal or superior to that of FBP. However,
results for the MTFTask measured using the low-contrast inserts, the MTFTask for ASIR and MBIR
data was lower than for the FBP, which was constant throughout all measurements. Similarly, as a
function of mA, the MTFTask for ASIR and MBIR varied as a function of noise—with MTFTask being
proportional to mA. Overall greater variability of MTFTask across dose and contrast was observed for
MBIR than for ASIR.
Conclusions: This approach provides a method for assessing the task-based MTF of a CT system
using conventional and iterative reconstructions. Results demonstrated that the object-specific MTF
can vary as a function of dose and contrast. The analysis highlighted the paradigm shift for iterative
reconstructions when compared to FBP, where iterative reconstructions generally offer superior noise
performance but with varying resolution as a function of dose and contrast. The MTFTask generated
by this method is expected to provide a more comprehensive assessment of image resolution across
different reconstruction algorithms and imaging tasks. © 2012 American Association of Physicists in
Medicine. [http://dx.doi.org/10.1118/1.4725171]

Key words: MTF, NPS, NEQ, detectability index, CT, iterative reconstruction, statistical reconstruc-
tion, quality assurance, image quality assessment

I. INTRODUCTION
The evaluation of spatial resolution of imaging systems has
played a central role in imaging performance evaluation.1
The advent of more complex imaging systems has posed
an increasing challenge in our ability to assess their perfor-
mance, including spatial resolution. For example, iterative
and statistical reconstructions can exhibit nonlinear signal
characteristics, which can affect system resolution properties

4115 Med. Phys. 39 (7), July 2012 0094-2405/2012/39(7)/4115/8/$30.00 © 2012 Am. Assoc. Phys. Med. 4115
4116 Richard et al.: Towards task-based assessment of CT performance 4116


differently than with standard reconstruction algorithms,
I (x, y) = f (x − x  , y − y  )PSF(x  , y  )dx  dy  , (1)
i.e., filtered back projection (FBP). One of the most
comprehensive metric used to measure and report spatial where PSF(x, y) is defined as the PSF such that
resolution of imaging systems is the modulation transfer func- 
tion (MTF).2 The MTF provides a measure of how well the PSF(x, y)dxdy = 1. (2)
system transfers contrast across spatial-frequencies. The MTF
is also essential to evaluating other key imaging performance The Fourier representation of the PSF is the optical transfer
metrics such as the detective quantum efficiency (DQE), function (OTF),
noise equivalent quanta (NEQ), and detectability index.3 OTF(u, v) = F {PSF(x, y)}, (3)
Previous methods for MTF measurements have used a va-
riety of “known” objects to generate an input to measure the where F{} denotes the Fourier transform. The MTF is the
point spread function (PSF). Ideally, a delta-function-like ob- modulus of the OTF as
ject would be employed; however, this approach is gener- MTF(u, v) = |OTF(u, v)| , (4)
ally not practical due to its low-signal power. In CT, a wire
describing the magnitude of transferred signal at each spatial
method provides a pseudo-point object across multiple slices
frequency. Note that (1) Eq. (2) implies that MTF(0) = 1 (dc
and can provide a 2D in-slice MTF by averaging realization
signal magnitude is unchanged by the transfer function); and
across multiple slices. This approach has been investigated
(2) an ideal system has MTF = 1 at all spatial frequencies.
since the early years of CT.4, 5 Even so, most current qual-
The MTF can be employed to characterize the resolution of an
ity assurance (QA) methods have adopted line-pair pattern
imaging system in terms of MTF( f) or to reflect the resolution
methods due to their simpler and more intuitive approach.6, 7
in a scalar form, for example, the frequency ( f50 ) at which
In radiography, the most common method for measuring the
MTF( f) reduces to 0.50.
MTF is the angled edge technique, which provides an edge
Experimentally, it is impractical to measure the PSF di-
across multiple rows or columns that can generate a “super-
rectly. That would require imaging of an infinitesimally small
sampled” edge-spread function (ESF).8, 9 The ESF can be dif-
object, a 2D delta function, so that I(x,y) = PSF(x,y). If one
ferentiated to yield the line-spread function (LSF) from which
is to do so, the mean signal magnitude would be correspond-
the Fourier transform yields the MTF. This method has been
ingly small and the resulting image would be dominated by
adapted to CT using sharp objects or edges to measure in-slice
noise. More convenient impulse function techniques are typi-
resolution.10
cally used for such measurements, such as the LSF, which is
Both the wire and edge techniques have limited utility in
the Radon transform of the PSF,
advanced CT systems. These systems offer nonlinear image 
processing and reconstruction algorithms which can lead to a LSF(x) = PSF(x, y)dy. (5)
contrast and dose dependency for the MTF. The conventional
wire and edge methods typically use highly dense materials The LSF can be measured from the image of a narrow
and very low noise to simulate an infinite impulse (i.e., sig- slit.11 Alternatively, the ESF, given by the antiderivative of
nal) to maximize CNR and improve measurement precision. the LSF, may be employed as
 x
However, this corresponds to objects that are atypical in clin-
ESF(x) = LSF(x  )dx  , (6)
ical images. Therefore, for these CT systems, it is important −∞
to use test objects with (lower) contrast and (higher) noise
which can be measured from the image of a sharp edge. This
to yield MTFs relevant to realistic objects. This paper intro-
can be rewritten as
duces a novel method for evaluating the spatial resolution of
∂
CT systems using a conventional QA phantom. This work was LSF(x) = ESF(x). (7)
motivated by a desire to: (1) assess MTF at various contrast ∂x
and noise levels; (2) measure the MTF using a standard QA Hence, the PSF, LSF, ESF, and MTF are all related. For a ra-
phantom without any additional equipment; and (3) report on dially symmetric MTF, Eqs (2)–(7) together with the Fourier
the MTF performance of two new reconstruction algorithms slice theorem suggest that the one-dimensional (1D) MTF
in terms of their performance compared to standard recon- (i.e., a slice of the full 2D MTF) is given by
 
struction.  ∂ ESF(x) exp(−2π if x)dx 
 ∂x 
MTF(f ) =  ∂ , (8)
 ∂x
ESF(x)dx 
II. METHODS AND MATERIAL where the denominator ensures that MTF(0) = 1 [equivalent
II.A. Theory to Eq. (2)]. Therefore, the MTF can be experimentally as-
sessed by imaging a sharp edge object to yield an estimate
Fourier based metrology is a prevalent approach to char- of the ESF, which in turn is related to the LSF (by derivative)
acterizing medical imaging performance. This includes the and to the MTF as in Eq. (8).
MTF, which characterizes the resolution of the imaging sys- The requirements for measuring the MTF of an imaging
tem. The output signal I(x,y) (i.e., image) is related to the in- system include: (1) a linear signal response and (2) a shift-
put signal f(x,y) (i.e., object) by the impulse response function invariant signal (i.e., signal independent of location in the im-
as age). However, in the strictest sense of the definition most

Medical Physics, Vol. 39, No. 7, July 2012

4117 Richard et al.: Towards task-based assessment of CT performance
F IG . 1. (a) Axial slice of the ACR (model 464) phantom and (b) one of the insert zoomed in. (c) Image of the GE performance phantom (GEPP), which includes
a 50 μm tungsten wire, zoomed in (d).
imaging systems are nonlinear and shift-variant. Neverthe-
less, under certain conditions, linearity and shift-invariance
can be a reasonable assumption. These conditions often in-
clude low-contrast and localized measurements. As a result,
it is often necessary to qualify under which conditions the
MTF results are valid. Measurement of MTF for nonlinear
system has been the focus of past research. For example, the
measurements of MTF on cathode ray tube (CRT) display
as discussed in the AAPM task group 18,12 which employed
low-contrast edge object for measuring the MTF. The MTF
for object-specific assessment was also essential in predict-
ing image performance in dual-energy imaging using noise
reduction algorithms. In that work, the MTF was measured
using materials that mimicked the contrast of various types of
tissues (i.e., soft tissue or bone; Ref. 13) and was employed to
fairly accurately predict the detectability index for detection
and discrimination tasks.14 Similarly, in this work, the MTF
is measured under predefined conditions such as contrast and
noise levels. This yields a specific task-based transfer function
that can be used to compute performance assessment metrics
such as the detectability index. This work is further justified
by the need to define the MTF on a locally linearizable region
of the object to assess the performance of task-specific perfor-
mance in the presence of nonlinear reconstruction algorithms.
Medical Physics, Vol. 39, No. 7, July 2012
4117
II.B. MTF measurement method
An experimental methodology was developed to assess
the MTF from the image of a 2D circular object as illustrated
in Fig. 1. The method was largely adapted from the edge tech-
nique developed for radiography but with incorporation of a
circular object.9 The ACR phantom was used as the basis of
the methodology. As described above, a motivation of this
work was the fact that nonlinear reconstruction algorithms in
CT may display spatial resolution dependence on contrast and
noise level in the image. The circular objects of varying con-
trast present in the ACR phantom (Fig. 1) enables the eval-
uation of that dependence. The dependence of the MTF on
signal and noise represents a different type of measurement
than those made on linear systems. Therefore, to highlight this
important difference between the conventional MTF and the
MTF measured for the purpose a task-based assessment on
nonlinear systems, we introduce a notation change to the MTF
as MTFTask to denote this distinction. In this sense, the object
used to measure MTFTask essentially mimics the contrast and
edge of the detected signal as perceived by the observer. This
MTFTask is quite different from what is conventionally known
as the system MTF, rather MTFTask refers to an object depen-
dent measure of sharpness in the Fourier domain analogous to
4118 Richard et al.: Towards task-based assessment of CT performance
F IG . 2. (a) Diagram depicting one quadrant of the image data showing
the reordering of the pixel intensities as a function of distance from center
(b) and uniformly resampled in (c).
the MTF but represents a different quantity with the intent to
model perceived image sharpness for the evaluation of task-
based imaging performance.
The ACR phantom contains three inserts and one air
cavity for HU accuracy measurement. Only the three solid
inserts were used for MTFTask measurement. The inserts pro-
vided nominal HU values of 955, 120, and −95 at 120 kVp for
the bone, acrylic, and polyethylene inserts, respectively. The
background material was water equivalent “solid water” with
a nominal HU value of 0 at 120 kVp. The inserts provided
a range of hyper- and hypodense materials for evaluating the
change in MTFTask across different materials.
A computer program was developed based on MATLAB
(MathWorks, MA) to generate a one-dimensional ESF from
the 2D image of a circular disk. The region of interest around
the disk was fitted to a 2D planer fit, which was subtracted
from the original image to remove low frequency trends. A
threshold method combined with center of mass algorithm
was first employed to determine the center of the disk. This
provided a radial coordinate system relative to the edge of the
disk. As shown in Fig. 2, each pixel was arranged based on
their distance from the center or equivalently, from the edge.
The transformation of the 2D image to a parametric line func-
tion yielded a nonuniformly sampled ESF. The ESF was re-
sampled using 1st order polynomial interpolation, with bins
of one-tenth that of the reconstructed pixel size.8 This bin-
ning factor was found empirically to offer an ideal tradeoff
between sample size and noise. To reduce noise in the ESF,
the resampled ESF was averaged across multiple ESF real-
izations (measured across up to ten consecutive CT slices) to
generate the ensemble ESF.
The derivative of the ensemble ESF yielded the LSF.
The resultant LSF was multiplied by a Hann window to re-
move the noise in the tails. The width of the Hann window
matched the length of the ESF, which was approximately
30-mm wide for the ACR phantom employed. This step was
devised to remedy the noise asymmetry in the LSF, as the tails
corresponding to the interior of the disk were noisier and less
densely sampled than those corresponding to the disk exte-
rior. The resultant LSF had a width approximately twice the
radius of the disk. The fast Fourier transform (FFT) of the LSF
yielded the MTFTask . The MTFTask across multiple ensembles
Medical Physics, Vol. 39, No. 7, July 2012
4118
(5) was averaged and the standard deviation was calculated
and reported via error bars. The MTFTask was generated from
50 individual slices.
II.C. Experimental measurements
The ACR QA phantom was scanned on the GE
Discovery CT750 HD scanner (Waukesha, WI) across a wide
range of mA (25, 50, 75, 100, and 200) stations while keeping
other acquisition parameters constant: 120 kVp, 0.5 s per rota-
tion, 1.375 pitch, and 0.625 mm reconstructed slice thickness.
Images were reconstructed with three different algorithms:
FBP, adaptive statistical iterative reconstruction (ASIR), and
model-based iterative reconstruction (MBIR). All FBP recon-
struction were reconstructed with the GE “standard” filter.
The ASIR is a commercially available reconstruction which
blends an iterative reconstruction with FBP; in this work, all
images used “ASIR50 ,” which corresponds to a 50/50 blend of
FBP and the iterative reconstruction. The MBIR is an inves-
tigational reconstruction pending FDA approval at the time
of this writing. The reconstruction takes into consideration
the finite size of the detector, the x-ray focal spot, and re-
constructed voxels, elements that are conventionally treated
as infinite points in space in FBP reconstructions. Further de-
tails on the MBIR algorithms can be found in Thibault et al.15
The MTF measurements were made according to the disk
methodology described above at multiple noise levels. To pro-
vide a comparison with more conventional wire technique, a
wire phantom was also employed to provide a reference point
for comparing MTF measurement methods for the various
reconstruction algorithms. The phantom (GE performance
phantom (GEPP), Waukesha, WI), shown in Figs. 1(c) and
1(d), incorporates a 80 μm tungsten wire. The phantom was
imaged on a GE 750 HD scanner with a conventional tech-
nique at 120 kVp and 200 mA with 0.5 s per rotation, a
1.375 pitch, and 0.625 mm reconstructed thickness using FBP,
ASIR, and MBIR. The image of the wire was used as the
PSF from which the 2D FFT yielded the 2D MTF. This was
performed with custom software written in MATLAB. The 2D
MTF was radially averaged to generate the final 1D MTF pro-
file for comparisons.
III. RESULTS
Figure 3 plots the MTFTask measured with the disk-based
and the wire-based method. The disk measurements were per-
formed on the high contrast insert (i.e., bone insert). The
MTFTask measured from the FBP and ASIR reconstructed
images shows little difference between the two measuring
techniques. For the FBP data, the largest discrepancy is at
high spatial frequencies, where the circular edge method dis-
plays slightly higher MTFTask values (less than 0.15) but well
within the measurement error. Excellent agreement was found
between the disk and wire methods for ASIR images. The
MBIR results show higher MTFTask values for both methods
but a significant difference between the two methods. This is
mainly due to the object-specific characteristics of MBIR, as
further highlighted in the results below.
4119 Richard et al.: Towards task-based assessment of CT performance 4119

1 FBP 1 ASIR 1 MIRB

50
08
0.8 08
0.8 Disc 08
0.8

Wire

MTFTask

MTFTask
MTFTask
0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

(a) (b) (c)
0 0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm)

F IG . 3. Comparison of the disk edge and wire based method for measuring the MTFTask for images reconstructed with (a) FBP, (b) ASIR, and (c) MBIR. Data
were acquired at 120 kVp and 200 mA.

Figure 4 demonstrates the MTFTask for FBP, ASIR, and
MBIR reconstructions acquired at the highest mA level
(200 mA) for various contrast inserts. The error bars display
the variability in the measured due to both stochastic sources
such as image noise but also due to more deterministic
sources such as the variability and nonuniformity in the
reconstruction. For the FBP measurements, the lower contrast
insert (i.e., acrylic and PE) provide low signal-to-noise ratio
resulting in noisier MTFTask measurements. Nonetheless,
the FBP-based MTFTask results agree across the different
inserts; they are well within the measurement error. This is an
expected result as the FBP is a linear reconstruction technique
and as such the MTFTask is expected to be independent of
contrast or noise. The ASIR results exhibit similar loss in
MTFTask precision as a function of object contrast. But most
importantly, they further exhibit a dependency on contrast
with higher resolution noted for the highest contrast object.
The MBIR exhibits the largest dependence on object contrast,
where the MTF is dramatically enhanced for the bone object
while the MTFTask for the PE object is comparable to the
FBP-based MTFTask . These results illustrate that for iterative
reconstructions, the MTF needs to be determined as an
object-based MTFTask rather that a system-based MTF.
Figure 5 plots the MTFTask measured using ASIR50 re-
constructed data across different materials (PE, acrylic, and
bone) and across five different tube currents. The data dis-
play a modest dependence on tube current for all contrast lev-
els. The MTFTask is highest for the highest contrast object and
lowest noise setting. Similarly, Fig. 6 shows the MTFTask mea-
sured using MBIR reconstructed data across different materi-
als and tube currents. The data display a strong dependence on
contrast, as noted above, but even a more pronounced depen-
dence on tube current, where the MTFTask seems to increase
with tube current across all materials. The highest MTFTask
are noted for the highest tube current, but they appear to
plateau towards highest mA settings. The data show clearly
that a measurement of MTFTask at high dose and high contrast
would provide a “best case” representation of the system’s
resolution.
Figure 7 summarizes the results by plotting the spa-
tial frequency at which the MTFTask reduces to 50% ( f50 ).
Figures 7(a)–7(c) plot f50 as a function of tube current (mA)
for the MTFTask measurements across the different inserts and
reconstruction algorithms. Results for FBP show no depen-
dence with f50 fixed at 0.35 mm−1 . ASIR50 , in contrast, dis-
plays a small dependence with f50 ranging from 0.22 at very
low tube currents and low contrast (i.e., PE at 10 mA) to
0.35 mm−1 for high contrast and high tube current (i.e., bone
at 200 mA). Results for MBIR display the largest dependence,
with f50 ranging from 0.21 to 0.6 mm−1 . For comparative

1 1 1

0.8 0.8 0.8

MTFTask
MTFTask
MTFTask

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm)

F IG . 4. Comparison of MTFTask measured across the different inserts (bone, acrylic, and PE) for images reconstructed with (a) FBP, (b) ASIR50 , and (c) MBIR.
Data were acquired at 120 kVp and 200 mA.

Medical Physics, Vol. 39, No. 7, July 2012

4120 Richard et al.: Towards task-based assessment of CT performance 4120

1 1 1

0.8 0.8 0.8

MTFTask

MTFTask
MTFTask

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm)

F IG . 5. Plots of MTFTask for ASIR50 at five tube current levels (25, 50, 75, 100, and 200 mA) and contrast inserts (a) PE, (b) acrylic, and (c) bone. Data were
acquired at 120 kVp.

purposes, Fig. 7(d) plots the dependency of noise on mA. The
significant correlation between mA and resolution for the it-
erative reconstructions, there is a clear paradigm shift with
advanced reconstructions, where compared to standard FBP,
both noise and resolution vary as a function of tube current
and dose.
IV. DISCUSSION
Advances in CT, including iterative algorithms, are
expected to require more comprehensive performance as-
sessment methods. Spatial resolution will be central to this
evaluation and the MTF is essential for assessing Fourier-
based system performance metrics such as the NEQ and
the detectability index.16, 17 The presented results highlight
an important paradigm shift in the clinic with iterative
reconstructions—as the pixel noise can be constrained by
the reconstruction algorithm, this causes distortions in the
image at low x-ray signal and lowers the spatial resolution
as characterized by the MTFTask . Current MTF assessment
methods only focus on a single contrast and noise level.4, 5
As a result, this work presented a method for measuring
the MTFTask for CT systems for different reconstruction
algorithms and across a range of noise and contrast levels.
In the case of FBP, no dependence was observed due to the
system’s linearity (as expected). For iterative reconstructions,
the MTFTask was found to be dependent on both object
contrast and image noise. This makes the comparison of
MTF between CT systems and protocols more challenging.
The contrast dependence of the task-based MTF further
implies that the results presented above depend not only on
the material used for the MTF measurement but also the
surrounding material. Therefore, the MTFTask results shown
in this work are representative only of the conditions used
in this study. Furthermore, the shape of the object may also
impact the MTFTask but was beyond the scope of this work.
This points toward the need to define specific guideline for
measuring the MTFTask for iterative algorithms, for example,
by specifying the noise or contrast at which the MTFTask is
being evaluated. It is essential to establish that dependence
as otherwise traditional wire-based MTF techniques would
only provide the best case scenario in most cases. The abil-
ity to provide an object-based MTF also provides a more ac-
curate description of the imaging task when computing the
detectability index.18 For example, a large low-contrast ob-
ject and a small high contrast object would have different
representative MTFs. This work provides a methodology for

1 1 1

0.8 0.8 0.8

MTFTask

MTFTask
MTFTask

0.6 0.6 0.6

0.4 0.4 0.4

0.2 0.2 0.2

0 0 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm) Spatial Frequency (cycles/mm)

F IG . 6. Plots of MTFTask for MBIR at five tube current levels (25, 50, 75, 100, and 200 mA) and contrast inserts (a) PE, (b) acrylic, and (c) bone. Data were
acquired at 120 kVp.

Medical Physics, Vol. 39, No. 7, July 2012

4121 Richard et al.: Towards task-based assessment of CT performance
1 1
0.8
(cycles/mm)
0.6
0.4
0.2
0 0
0 50 100 150 200
Tube current (mA)
1
0.8
(cycles/mm)
0.6
0.4
0.2
0
0 50 100 150 200
Tube current (mA)
F IG . 7. Plots of f50 for MTFTask for FBP, ASIR50 , and MBIR at across tube current for the (a) PE, (c) acrylic, and (c) bone insert. (d) Plots of the corresponding
pixel noise measured in the background. Data were acquired at 120 kVp.
computing task-based MTF for evaluating observer model fig-
ure of merits for CT performance assessment using iterative
reconstruction as presented in works by Zeng and Myers19
and Yendiki and Fessler,20 which employed mathematical ob-
servers to assess statistically reconstructed images.
Applications for MTFTask naturally include QA. The ACR
phantom used in this study is currently available in many clin-
ics as a QA phantom. Extending QA evaluations with a task-
based MTF measurements could be automated, as is currently
underway at our institution. In addition to evaluating the rel-
ative reliability of the MTF over time for QA purposes, the
interpretation of the MTFTask in more absolute terms may also
be desirable. This can be achieved by combining the MTFTask
with a task function to generate a surrogate of image quality,
such as the detectability index,3 which could provide a useful
approach for comparing and optimizing image reconstruction
performance, albeit under very specific imaging conditions.
Notwithstanding the findings of the study, certain limita-
tions need to be acknowledged. First, in this work, the MTF
was applied not as a generic reflection of system resolution
but rather as an object-specific quantity. This was done so to
accommodate the nonlinearity of the iterative reconstruction
methods and is reflected by the notation change—MTFTask .
However, provided that the MTFs are annotated with quali-
Medical Physics, Vol. 39, No. 7, July 2012
4121
0.8
(cycles/mm)
0.6
0.4
0.2
0 50 100 150 200
Tube current (mA)
100
80
Noise (HU)
60
40
20
0
0 50 100 150 200
Tube current (mA)
fying information such as, but not limited to, the noise and
contrast level at which the MTFTask is measured, they can pro-
vide some reflection of local resolution for clinically relevant
tasks. Note that this is analogous to the MTF of dual-energy
images using noise reduction where the MTF is dependent on
the contrast of the object.13 In that case, the proper descrip-
tion of the MTF was essential to accurately model and predict
system and human performance.14 Another limitation of this
work is that the MTF was radially averaged. This was per-
formed to minimize the noise in the measurements, especially
those made at a low mA. Therefore, the result ignores any
anisotropy and effectively averages out the MTF in the axial
plane. One way to overcome this limitation is to only average
a certain angular extent, which is certainly possible within the
framework presented but was outside the scope of this work. It
should also be noted that the error bars describe the stochas-
tic variability as well as the deterministic variability associ-
ated with the nonuniformity of the MTF across realizations.
While noise and contrast strongly affect the MTFTask , other
factors could be analyzed. For example, the MTFTask could
be measured as a function of other parameters such as the
contrast-to-noise ratio. Furthermore, the study is related to its
application to only one CT system; therefore, the MTF was
measured for only two types of iterative reconstructions—
4122 Richard et al.: Towards task-based assessment of CT performance
there are many parameters in iterative reconstructions, such
as the regularizer and edge preservers, that could affect the
MTFTask —future studies are warranted to include other CT
system and reconstruction algorithms.
V. CONCLUSION
In summary, the method described in this work provides
a methodology for evaluation of the task-based MTF across
various CT reconstruction algorithms. The method employs a
conventional image quality phantom for CT with an adapted
circular edge technique to generate the ESF across vari-
ous contrast and noise levels. The method provides a robust
method for characterizing resolution of different types of re-
constructed objects. The findings show how iterative recon-
structions apply different tradeoffs across noise and resolution
as a function of tube current. In all cases, the noise was lower
with ASIR and MBIR than with FBP. Nevertheless, the task-
based MTF for MBIR was greater than for FBP at high-dose
and high-contrast objects, however, the performance dimin-
ished for low-dose, low-contrast objects. In contrast, MTFTask
for ASIR remained equal or lower than FBP in the set of imag-
ing conditions investigated in this study. As a result, the MTF
assessment presented in this work begins to provide a method
for evaluating task-based resolution metrics for system opti-
mization and comparison across standard and iterative recon-
struction algorithms.
a) Author
to whom correspondence should be addressed. Electronic mail:
samu.richard@gmail.com; Telephone: 919-684-1461.
1 B. Jacob, L. K. Harold, and L. V. M. Richard, Handbook of Medical Imag-
ing, Volume 1. Physics and Psychophysics (SPIE, 2000).
2 C. E. Metz and K. Doi, “Transfer function analysis of radiographic imaging
systems,” Phys. Med. Biol. 24(6), 1079–1106 (1979).
3 International Commision on Radiation Units and Measurements, “Prescrib-
ing, recording, and reporting photon beam therapy,” Supplement to ICRU
Report No. 50 (ICRU, Washington, DC, 1999).
4 C. J. Bischof and J. C. Ehrhardt, “Modulation transfer function of the EMI
CT head scanner,” Med. Phys. 4(2), 163–167 (1977).
5 R. T. Droege and R. L. Morin, “A practical method to measure the MTF of
CT scanners,” Med. Phys. 9(5), 758–760 (1982).
Medical Physics, Vol. 39, No. 7, July 2012
4122
6 E. L. Nickoloff and R. Riley, “A simplified approach for modulation trans-
fer function determinations in computed tomography,” Med. Phys. 12(4),
437–442 (1985).
7 S. Rathee, B. G. Fallone, and D. Robinson, “Modulation transfer func-
tion of digitally reconstructed radiographs using helical computed tomog-
raphy,” Med. Phys. 29(1), 86–89 (2002).
8 E. Samei, M. J. Flynn, and D. A. Reimann, “A method for measuring the
presampled MTF of digital radiographic systems using an edge test de-
vice,” Med. Phys. 25(1), 102–113 (1998).
9 H. Fujita et al., “A simple method for determining the modulation trans-
fer function in digital radiography,” IEEE Trans. Med. Imaging 11(1),
34–39.(1992).
10 I. Mori and Y. Machida, “Deriving the modulation transfer function of CT
from extremely noisy edge profiles,” Radiol. Phys. Technol. 2(1), 22–32
(2009).
11 I. A. Cunningham and B. K. Reid, “Signal and noise in modulation transfer
function determinations using the slit, wire, and edge techniques,” Med.
Phys. 19(4), 1037–1044 (1992).
12 E. Samei, A. Badano, D. Chakraborty, K. Compton, C. Cornelius,
K. Corrigan, M. J. Flynn, B. Hemminger, N. Hangiandreou, J. Johnson,
M. Moxley, W. Pavlicek, H. Roehrig, L. Rutz, J. Shepard, R. Uzenoff,
J. Wang, and C. Willis, “Assessment of display performance for medical
imaging systems,” Report of the American Association of Physicists in
Medicine (AAPM) Task Group 18, 2005.
13 S. Richard and J. H. Siewerdsen, “Cascaded systems analysis of noise re-
duction algorithms in dual-energy imaging,” Med. Phys. 35(2), 586–601
(2008).
14 S. Richard and J. H. Siewerdsen, “Comparison of model and human ob-
server performance for detection and discrimination tasks using dual-
energy x-ray images,” Med. Phys. 35(11), 5043–5053 (2008).
15 J. B. Thibault et al., “A three-dimensional statistical approach to improved
image quality for multislice helical CT,” Med. Phys. 34(11), 4526–4544
(2007).
16 J. H. Siewerdsen, I. A. Cunningham, and D. A. Jaffray, “A framework for
noise-power spectrum analysis of multidimensional images,” Med. Phys.
29(11), 2655–2671 (2002).
17 D. J. Tward and J. H. Siewerdsen, “Cascaded systems analysis of the
3D noise transfer characteristics of flat-panel cone-beam CT,” Med. Phys.
35(12), 5510–5529 (2008).
18 International Commission on Radiation Units and Measurements,
“Medical imaging—The assessment of image quality,” Report I (Interna-
tional Commission on Radiation Units and Measurements, Bethesda, MD,
1996).
19 R. Zeng and K. J. Myers, “Task-based comparative study of iterative image
reconstruction methods for limited-angle x-ray tomography,” Proc. SPIE
7961, 796137 (2011).
20 A. Yendiki and J. A. Fessler, “Analysis of observer performance in
unknown-location tasks for tomographic image reconstruction,” J. Opt.
Soc. Am. A Opt. Image Sci. Vis. 24(12), B99–B109 (2007).