DESCRIPTION
BACKGROUND OF THE INVENTION
Technical Field
The present invention relates to an image generation apparatus,
method, and program that generates a pseudo three-dimensional image
by performing volume rendering on a three-dimensional image.
Background Art
Heretofore, processing for generating and displaying a pseudo
three-dimensional image has been performed in which
three-dimensional image data of a subject obtained by CT equipment,
MRI equipment, ultrasonic diagnostic equipment, or the like are
stereoscopically visualized on a two-dimensional plane using
computer graphics technologies in order to facilitate understanding
of a three-dimensional structure of the subject and the like. As
a method for generating such a pseudo three-dimensional image, a
volume rendering method is known in which opacity and color
information of R, G, B are set to each pixel value (voxel value)
constituting a three-dimensional image and the visualization is
achieved by performing ray-casting on each pixel on a projection
plane from the observation side.
For the opacity or color information set (referred to as
opacity curve and color map respectively) which is set to each pixel
value, it is common to give one set to one three-dimensional image.
But this causes a problem that different tissues having the same
signal value cannot be displayed distinguishably. In response to
this, when generating one image representing a plurality of objects
(e.g., bone, blood vessel, heart, liver, and the like) by volume
render, Non-Patent Document 1 proposes that different tissues having
the same signal value are represented by different color or opacity
by applying a different color map and/or opacity curve to each object
region.
[Prior Art Documents]
[Patent Documents]
2
Patent Document 1: Japanese Unexamined Patent Publication No.
2011-212219
[Non-Patent Documents]
Non-Patent Document 1: H. Imai "I see! ! The Bible of the Way
of Thinking and Processing - Three-Dimensional Medical Image",
Shujunsha, P105, 2003
DISCLOSURE OF THE INVENTION
In the meantime, in CT imaging of a blood vessel, for example,
a method that further visualizes and makes that portion extractable
by performing imaging after injecting a special liquid called a
contrast agent is sometimes used so that pixel values of the blood
vessel differ from those of other organs in a three-dimensional image.
In this case, however, a variation in pixel value occurs from place
to place even in the same blood vessel region due to the imaging
timing or the amount of the contrast agent. Therefore, when
generating an image that represents a wide range of blood vessel
region, even for one target object tissue of blood vessel, the simple
application of one color map and/or an opacity curve causes a problem
that a three-dimensional morphology of the entire blood vessel cannot
be represented accurately.
The problem due to the variation in pixel value may possibly
occur by a partial volume effect. In particular, for thin blood
vessels of a diameter represented by several pixels, there may be
a case where a blood vessel which should have actually a high pixel
value is imaged with a pixel value lower than the real value influenced
by the surrounding pixel values at the time of the imaging. This
may result in that, in an image generated by applying one color map
and opacity curve which is based on the pixel value distribution
of a thick blood vessel, a thick blood vessel portion appears clearly
but a thin blood vessel portion is not visualized, as shown, for
example, in Figure 6, while, in an image generated by one color map
and opacity curve, which are the aforementioned color map and opacity
curve translated according to the pixel value distribution of a thin
blood vessel, a thin blood vessel portion is visible (refer to the
portion enclosed by the white circle) but a thick blood vessel portion
3
appears over inflated, as shown in Figure 7.
Patent Document 1 proposes a method for dynamically correcting
color map and/or opacity curve with respect to each image, but the
method is based on the assumption that the depiction range of each
image is so small that the variation in pixel value is negligible
and related to how to decide one opacity curve used for generating
on image. Therefore, the method does not solve the aforementioned
problem when generating an image representing a wider object range.
In view of the circumstances described above, it is an object
of the present invention to provide an image generation apparatus,
method, and program capable of generating an image which depicts
three-dimensional morphology of a predetermined target object more
accurately when generating a pseudo three-dimensional image by
performing volume rendering.
An image generation apparatus of the present invention
includes an image generation section that generates a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using an opacity curve that defines the
relationship between pixel value and opacity, a region
identification section that identifies a whole region representing
a predetermined target object from the three-dimensional image, and
an opacity curve setting section that sets a base opacity curve to
the identified whole region, obtains, with respect to each of at
least some pixels in the identified whole region, a representative
value of pixel values in an adjacent region of a pixel concerned,
and sets an opacity curve obtained by modifying the base opacity
curve using the obtained representative value as the opacity curve
to be applied to the pixel concerned in the volume rendering (first
image generation apparatus).
The opacity curve setting section may be a section that obtains,
with respect to each of the at least some pixels, a shift amount
of opacity curve based on a subtraction value obtained by subtracting
a representative value of pixel values in the whole region from the
representative value of pixels in the adjacent region of the pixel
concerned and sets an opacity curve obtained by shifting the base
4
opacity curve by the obtained shift amount in a pixel value direction
as the opacity curve to be set to the pixel concerned in the volume
rendering.
The term "pixel value direction" as used herein refers to,
when the opacity curve is set in a coordinate system with one axis
(e.g., horizontal axis) representing the pixel value and the other
axis (e.g., vertical axis) representing the opacity, a direction
of the axis representing the pixel value.
When obtaining the shift amount based on the subtraction value,
a value having the same positive or negative sign as that of the
subtraction value is obtained as the shift amount. Here, the absolute
value of the shift amount may be the same as the absolute value of
the subtraction value or a value obtained by multiplying the value
with a predetermined coefficient other than 1. Thus, "shifting by
the shift amount" refers to shifting to plus direction of the axis
representing the pixel value if the sign of the shift amount is plus
while shifting to minus direction of the axis representing the pixel
value if the sign of the shift amount is minus.
In the image generation apparatus of the present invention,
the opacity curve setting section may be a section that obtains,
with respect to each of some pixels obtained by sampling the pixels
in the whole region at a predetermined interval, a representative
value of pixels in an adjacent region of a pixel concerned, obtains
a shift amount of opacity curve based on a difference between the
obtained representative value and a representative value of pixel
values in the whole region, and sets an opacity curve obtained by
shifting the base opacity curve by the obtained shift amount to a
pixel value direction as the opacity curve to be applied to the pixel
concerned, and obtains, with respect to each of pixels in the whole
region other than the sampled some pixels, an estimated value of
shift amount at a pixel concerned by interpolation using the shift
amount of each of the two or more sampled pixels located adjacent
to the pixel concerned and sets an opacity curve obtained by shifting
the base opacity curve by the determined estimated value in a pixel
value direction as the opacity curve to be applied to the pixel
5
concerned in the volume rendering.
The representative value of the pixel values in the adjacent
region may be a mode value, a median value, or an average value of
the pixel values in the adjacent region, an average value of pixel
values of all pixel values in the adjacent region which fall in a
range of predetermined width of pixel values from the mode value
or the median value of the pixels in the adjacent region, or a mode
value or a median value of pixel values of all pixel values in the
adjacent region which fall in a range of predetermined width of pixel
values from the mode value or the median value of the pixels in the
adjacent region.
At this time, the representative value of the pixels in the
whole region may be a value of the same type as that of the
representative value of the pixels in the adjacent region or a value
of different type from that of the representative value of the pixels
in the adjacent region. For example, if the representative value
of the pixels in the adjacent region is the average value, the
representative value of the pixels in the whole region may be the
average value or the mode value of the pixels in the whole region.
In the image generation apparatus of the present invention
described above, the image generation section may be a section that
generates a pseudo three-dimensional image by performing volume
rendering on the three-dimensional image using a color map that
defines the relationship between pixel value and display color, and
the apparatus may further includes a color map setting section that
sets a base color map to the identified whole region and calculates,
with respect to each of at least some pixels in the identified whole
region, a representative value of pixel values in an adjacent region
of a pixel concerned, and sets a color map obtained by modifying
the set base color map using the calculated representative value
as the color map to be applied to the pixel concerned in the volume
rendering.
An image generation apparatus of the present invention
includes an image generation section that generates a pseudo
three-dimensional image by performing volume rendering on a
6
three-dimensional image using a color map that defines the
relationship between pixel value and display color, a region
identification section that identifies a whole region representing
a predetermined target object from the three-dimensional image, and
a color map setting section that sets a base color map to the
identified whole region, calculates, with respect to each of at least
some pixels in the identified whole region, a representative value
of pixel values in an adjacent region of a pixel concerned, and sets
a color map obtained by modifying the set base color map using the
calculated representative value as the color map to be applied to
the pixel concerned in the volume rendering (second image generation
apparatus).
In the first and the second image generation apparatuses, the
predetermined target object may be a blood vessel.
First and second image processing methods of the present
invention are methods, each performs the processing performed by
each section of the first or the second image generation apparatus
with at least one computer.
First and second image processing programs of the present
invention are programs, each causes at least one computer to perform
the processing performed by each section of the first or the second
image generation apparatus. The programs are recorded on recording
media, such as CD-ROM, DVD, and the like, or recorded on an auxiliary
storage of a server computer or a network storage in a downloadable
manner and supplied to the user.
According to the first image generation apparatus, method,
and program of the present invention, when generating a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using an opacity curve that defines the
relationship between pixel value and opacity, a whole region
representing a predetermined target object is identified from the
three-dimensional image, a base opacity curve is set to the
identified whole region, with respect to each of at least some pixels
in the identified whole region, a representative value of pixel
values in an adjacent region of a pixel concerned is obtained, and
7
an opacity curve obtained by modifying the base opacity curve using
the obtained representative value is set as the opacity curve to
be applied to the pixel concerned in the volume rendering. This allows
an opacity curve appropriate for depicting the pixel to be applied
to each of all pixels in the region representing the predetermined
target object, whereby an image more accurately depicting
three-dimensional morphology of the predetermined target object may
be generated. This effect is more significant if the predetermined
target object is a blood vessel and further significant if the
three-dimensional image is an image obtained by injecting a contrast
agent into the blood vessel.
In the image generation apparatus, method, and program of the
present invention described above, if an arrangement is adopted in
which, with respect to each of some pixels obtained by sampling the
pixels in the whole region at a predetermined interval, a
representative value of pixels in an adjacent region of a pixel
concerned is obtained, a shift amount of opacity curve based on a
difference between the obtained representative value and a
representative value of pixel values in the whole region is obtained,
and an opacity curve obtained by shifting the base opacity curve
by the obtained shift amount to a pixel value direction is set as
the opacity curve to be applied to the pixel concerned, and, with
respect to each of pixels in the whole region other than the sampled
some pixels, an estimated value of shift amount at a pixel concerned
is obtained by interpolation using the shift amount of each of the
two or more sampled pixels located adjacent to the pixel concerned,
and an opacity curve obtained by shifting the base opacity curve
by the determined estimated value in a pixel value direction is set
as the opacity curve to be applied to the pixel concerned in the
volume rendering, the arithmetic operation may be speeded up.
Further, if the representative value of the pixel values in
the adjacent region is the mode value or the median value of the
pixel values in the adjacent region, an extreme value due to noise,
if present in the adjacent region, is less likely to be reflected
to the representative value.
8
Further, if the representative value of the pixel values in
the adjacent region is an average value, a mode value, or a median
value of pixel values of all pixel values in the adjacent region
which fall in a range of predetermined width of pixel values from
the mode value or the median value of the pixels in the adjacent
region, an extreme value that presents on the maximum or the minimum
side of the pixel values in the adjacent region is less likely to
be reflected to the representative value.
Still further, in the image generation apparatus, method, and
program of the present invention described above, if an arrangement
is adopted in which a base color map is set to the identified whole
region and, with respect to each of at least some pixels in the
identified whole region, a representative value of pixel values in
an adjacent region of a pixel concerned is calculated, and a color
map obtained by modifying the set base color map using the calculated
representative value is set as the color map to be applied to the
pixel concerned in the volume rendering, a color map appropriate
for depicting the pixel may be applied to each of all pixels in the
region representing the predetermined target object, whereby an
image more accurately depicting three-dimensional morphology of the
predetermined target object may be generated.
According to the second image generation apparatus, method,
and program of the present invention, when generating a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using a color map that defines the
relationship between pixel value and opacity, a whole region
representing a predetermined target object is identified from the
three-dimensional image, a base color map is set to the identified
whole region, with respect to each of at least some pixels in the
identified whole region, a representative value of pixel values in
an adjacent region of a pixel concerned is calculated, and a color
map obtained by modifying the set base color map using the calculated
representative value is set as the color map to be applied to the
pixel concerned in the volume rendering. This allows a color map
appropriate for depicting the pixel to be applied to each of all
9
pixels in the region representing the predetermined target object,
whereby an image more accurately depicting three-dimensional
morphology of the predetermined target object may be generated.
BRIEF DESCRIPTION OF DRAWINGS
Figure 1 is a schematic block diagram of an image generation
apparatus of a first embodiment.
Figure 2 is a flowchart illustrating an operation of the image
generation apparatus of the first embodiment.
Figure 3 is a schematic block diagram of an image generation
apparatus of a second embodiment.
Figure 4 is a flowchart illustrating an operation of the image
generation apparatus of the second embodiment.
Figure 5 is a drawing illustrating an example image generated
by the image generation apparatus of the second embodiment.
Figure 6 is a drawing illustrating an example image generated
by conventional technology.
Figure 7 is a drawing illustrating an example image generated
by conventional technology.
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, a first embodiment of the present invention will
be described with reference to the accompanying drawings. Figure
1 is a block diagram of an image generation apparatus 1, illustrating
a schematic configuration thereof. The configuration of the image
generation apparatus 1 illustrated in Figure 1 is realized by
executing an image generation program read into an auxiliary storage
unit on a computer. Here, the image generation program is recorded
on a recording medium, such as CD-ROM or the like, or distributed
via a network, such as the Internet or the like, and installed on
the computer. The image processing program defines image obtaining
processing, image generation processing, region specification
processing, opacity curve setting processing, and display control
processing, as the processing to be performed by CPU of the computer.
When each processing described above is executed by the CPU according
to the definition described above, the computer functions as an image
obtaining section 11, an image generation section 12, a region
10
identification section 13, an opacity curve setting section 14, and
a display control section 16, which will be described later.
The image generation apparatus 1 is connected to a storage
unit 2, such as a hard disk drive or the like, and a display unit
3, such as a display or the like. The storage unit 2 stores
three-dimensional image data (three-dimensional images) obtained
by imaging predetermined target objects with imaging equipment, such
as CT, MRI, PET, SPECT, ultrasonic images, and the like. A
three-dimensional image is a collection of pixel data in a
three-dimensional space and may be obtained by stacking a plurality
of tomographic images obtained by imaging equipment. The image
obtaining section 11 obtains the three-dimensional image stored in
the storage unit 2 and stores the image in a storage device built
in the image generation apparatus 1 or a storage unit connected to
the image generation apparatus, such as a hard disk drive.
The region identification section 13 identifies a whole region
representing a predetermined target object from the
three-dimensional image obtained by the image obtaining section 11.
Here, a whole region representing a predetermined target object
refers to a region that includes all regions representing the
predetermined target object and does not include other regions.
Hereinafter, a description will be made of a case in which the
predetermined target object is a blood vessel. The region
identification section 13 extracts a region representing the blood
vessel (blood vessel region) from the three-dimensional image and
identifies the extracted region as a whole region. Here, the
extraction of the blood vessel region may be obtained by threshold
method, Region Growing method, Level Set method, and other various
types of image processing.
For example, the region identification section 13 searches
a linear structure with respect of each local region in the three
dimensional image by calculating eigenvalues of a 3x3 Hessian matrix.
In a region where a linear structure is included, one of the three
eigenvalues is close to 0 and the other two are relatively large.
The eigenvector corresponding to the eigenvalue close to 0 indicates
11
the principal axis direction of the linear structure. The region
identification section 13 makes use of this relationship and
determines the likelihood of linear structure with respect to each
local region based on eigenvalues of the Hessian matrix. For a local
region where a linear structure is identified, the center point
thereof is detected as a candidate point. Then, the candidate points
detected by the search are connected based on a given algorithm.
This builds a tree structure formed of the candidate points and blood
vessel branches (edges) connecting the candidate points. The
coordinate information of the plurality of detected candidate points
and vector information indicating the directions of the blood vessel
branches are stored in a memory with the identifiers of the candidate
points and blood vessel branches. Then, a blood vessel contour (blood
vessel external wall) is identified on a cross-section perpendicular
to the blood vessel route with respect to each detected candidate
point based on the surrounding pixel values. The shape may be
identified by known segmentation methods as represented by the
Graph-Cuts. The blood vessel region which is a tubular structure
is extracted, and information necessary to identify the extracted
blood vessel region is generated and stored in a memory through the
aforementioned processing.
The opacity curve setting section 14 sets an opacity curve
to be applied, in volume rendering, to each pixel in the blood vessel
region identified by the region identification section 13. The
opacity curve defines the relationship between pixel value and
opacity and may be represented by a function of pixel value. More
specifically, a base opacity curve is set to the blood vessel region
first. For example, a pixel value distribution of the blood vessel
region and a pixel value distribution of the neighboring region are
checked with the three-dimensional image, and pixel values serving
as the boundary separating the blood vessel region from other regions
are obtained based on these distributions. Then, an opacity curve
in which opacity "0" changes to a value other than "0" or opacity
"1" changes to a value other than "1" is set adjacent to the pixel
values. The opacity curve may be a curve in which the opacity changes
12
in stepwise from "0" to "1" or a curve in which the opacity changes
at a predetermined slope according to increase or decrease in the
pixel value.
Then, the opacity curve setting section 14 obtains, with
respect to each of some pixels obtained by three-dimensionally
sampling the pixels in the blood vessel region at a predetermined
interval, a subtraction value obtained by subtracting the average
value of the pixel values in the blood vessel region from the average
value of the pixel values of an adjacent region (e.g., region in
the range of lcm in up, down, left, right, front, and back directions)
of a pixel concerned as a shift amount of the opacity curve, and
sets an opacity curve obtained by shifting the base opacity curve
by the determined shift amount in a pixel value direction as the
opacity curve to be applied to the pixel concerned. That is, the
opacity curve setting section 14 sets an opacity curve represented,
when the pixel value is taken as a variable v, the base opacity curve
is taken as OD (v), and the determined shift amount is taken as m,
by 0(v) = OD(v-m) to each of the sampled pixels.
Further, the opacity curve setting section 14 obtains, with
respect to each of pixels in the blood vessel region other than the
sampled some pixels, an estimated value of shift amount at a pixel
concerned by interpolation using the shift amount of each of the
two or more sampled pixels located adjacent to the pixel concerned,
and sets an opacity curve obtained by shifting the base opacity curve
by the determined estimated value in a pixel value direction as the
opacity curve to be applied to the pixel concerned.
The image generation section 12 generates a pseudo
three-dimensional image by performing volume rendering on the
three-dimensional image using the opacity curve set by the opacity
curve setting section 14. The display control section 16 displays
the pseudo three-dimensional image generated by the image generation
section 12 on the display unit 3.
Figure 2 is a flowchart illustrating an operation of the image
generation apparatus 1. As illustrated, first, the image obtaining
section 11 obtains a three-dimensional image stored in the storage
13
unit 2 (SI). Then, the region identification section 13 extracts
a whole region representing a blood vessel from the three-dimensional
image obtained by the image obtaining section 11 and identifies the
extracted region as a whole region (S2) . Then, the opacity curve
setting section 14 sets a base opacity curve to the blood vessel
region (S3) . Further, the opacity curve setting section 14 obtains,
with respect to each of some pixels obtained by three-dimensionally
sampling the pixels in the blood vessel region at a predetermined
interval, a subtraction value obtained by subtracting the average
value of the pixel values in the blood vessel region from the average
value of the pixel values of the adjacent region of a pixel concerned
as a shift amount of the opacity curve and obtains, with respect
to each of pixels in the blood vessel region other than the sampled
some pixels, an estimated value of the shift amount at a pixel
concerned by interpolation using the shift amount of each of the
two or more sampled pixels located adjacent to the pixel concerned
(S4) .
Further, the opacity curve setting section 14 sets, with
respect to each of pixels other than the sampled some pixels, an
opacity curve obtained by shifting the base opacity curve by the
shift amount determined in step S3 in a pixel value direction as
the opacity curve to be applied to the pixel concerned and sets,
with respect to each of pixels other than the sampled some pixels,
an opacity curve obtained by shifting the base opacity curve by the
estimated value determined in step S3 in a pixel value direction
as the opacity curve to be applied to the pixel concerned (S5).
Thereafter, the image generation section 12 generate a pseudo
three-dimensional image by performing volume rendering on the
three-dimensional image using the opacity curve set by the opacity
curve setting section 14 (S6). Then, the display control section
16 displays the pseudo three-dimensional image generated by the image
generation section 12 on the display unit 3 (S7), and the processing
is completed.
In the present embodiment, the region identification section
13 identifies the whole region representing a blood vessel from a
14
three-dimensional image, the opacity curve setting section 14 sets
a base opacity curve to the identified whole region, and obtains,
with respect to each of at least some of the pixels in the whole
region, a representative value of pixel values in an adjacent region
of a pixel concerned and sets an opacity curve obtained by changing
the base opacity curve using the obtained representative value as
the opacity curve to be applied to the pixel concerned in volume
rendering, the image generation section 12 generates a pseudo
three-dimensional image by performing volume rendering using the
opacity curves set by the opacity curve setting section 14. This
allows an opacity curve appropriate for depicting the pixel to be
applied to each of all pixels in the region representing the blood
vessel, whereby an image more accurately depicting three-dimensional
morphology of the blood vessel may be generated.
Next, a second embodiment of the present invention will be
described. Figure 3 illustrates an image generation apparatus
according to the second embodiment of the present invention. The
image generation apparatus 100 of the present embodiment includes
a color map setting section 15 in addition to the configuration of
the projection image generation apparatus 10 according to the first
embodiment illustrated in Figure 1. Other aspects are identical to
those of the first embodiment.
The color map setting section 15 sets a color map to be applied
to each pixel in the blood vessel region identified by the region
identification section 13 in volume rendering. The color map defines
the relationship between the pixel value and display color and may
be represented by a function of pixel value. More specifically, a
base color map is first set to the blood vessel region. Then, the
color map setting section 15 obtains, with respect to each of some
pixels obtained by three-dimensionally sampling the pixels in the
blood vessel region at a predetermined interval, a subtraction value
obtained by subtracting the average value of the pixel values in
the blood vessel region from the average value of the pixel values
of an adjacent region (e.g., region in the range of 1cm in up, down,
left, right, front, and back directions) of a pixel concerned as
15
a shift amount of the color map, and sets a color map obtained by
shifting the base color map by the determined shift amount in a pixel
value direction to the pixel concerned as the color map to be applied
to the pixel concerned. That is, the color map setting section 15
sets a color map represented, when the pixel value is taken as a
variable v, the base color map is taken as ClrMapD (v), and determined
shift amount is taken as m, by ClrMap (v) = ClrMapD (v-m) to each
of the sampled pixels.
Further, the color map setting section 15 obtains, with
respect to each of pixels other than the sampled some pixels in the
blood vessel region, an estimated value of the shift amount at a
pixel concerned by interpolation using the shift amount of each of
the two or more sampled pixels located adjacent to the pixel concerned,
and sets color map obtained by shifting the base color map by the
determined estimated value in a pixel value direction as the color
map to be applied to the pixel concerned.
At this time, if the shift amount and the estimated value have
already obtained by the opacity curve setting section 14, the color
map setting section 15 set a color map to be applied to each pixel
in the blood vessel region using the shift amount and the estimated
value obtained by the opacity curve setting section 14 without
calculating these values. It should, of course, be appreciated that
an arrangement may be made in which the opacity curve setting section
14 sets the opacity curve to be applied to each pixel in the blood
vessel region using the shift amount and the estimated value obtained
by the color map setting section 15.
The image generation section 12 generates a pseudo
three-dimensional image by performing volume rendering on the
three-dimensional image using the opacity curve set by the opacity
curve setting section 14 and the color map set by the color map setting
section 15. Figure 5 shows an example pseudo three-dimensional image
generated by the image generation apparatus 100 of the present
embodiment. In Figure 5, the three-dimensional morphology of the
entire blood vessel is depicted more accurately in comparison with
the images generated by performing volume rendering on the same
16
three-dimensional image according to conventional technology shown
in Figures 6 and 7.
Figure 4 is a flowchart illustrating an operation of the image
generation apparatus 100. As illustrated, first, the image obtaining
section 11 obtains a three-dimensional image stored in the storage
unit 2 (Sll) . Then, the region identification section 13 extracts
a whole region representing a blood vessel from the three-dimensional
image obtained by the image obtaining section 11 and identifies the
extracted region as a whole region (S12) . Then, the opacity curve
setting section 14 sets a base opacity curve to the blood vessel
region and the color map setting section 15 sets a base color map
to the blood vessel region (S13).
Further, the opacity curve setting section 14 or the color
map setting section 15 obtains, with respect to each of some pixels
obtained by three-dimensionally sampling the pixels in the blood
vessel region at a predetermined interval, a subtraction value
obtained by subtracting the average value of the pixel values in
the blood vessel region from the average value of the pixel values
of the adjacent region of a pixel concerned as a shift amount and
obtains, with respect to each of pixels other than the sampled some
pixels, an estimated value of the shift amount at a pixel concerned
by interpolation using the shift amount of each of the two or more
sampled pixels located adjacent to the pixel concerned (S14).
Then, the opacity curve setting section 14 sets, with respect
to each of the sampled some pixels, an opacity curve obtained by
shifting the base opacity curve by the shift amount determined in
step S14 in a pixel value direction as the opacity curve to be applied
to the pixel concerned and sets, with respect to each of pixels other
than the sampled some pixels, an opacity curve obtained by shifting
the base opacity curve by the estimated value determined in step
S14 in a pixel value direction as the opacity curve to be applied
to the pixel concerned. Further, the color map setting section 15
sets, with respect to each of the sampled some pixels, a color map
obtained by shifting the base color map by the shift amount determined
in step S14 in a pixel value direction as the color map to be applied
17
to the pixel concerned and sets, with respect to each of pixels other
than the sampled some pixels, a color map obtained by shifting the
base color map by the estimated value determined in step S14 in a
pixel value direction as the color map to be applied to the pixel
concerned (S15).
Thereafter, the image generation section 12 generate a pseudo
three-dimensional image by performing volume rendering on the
three-dimensional image using the opacity curves set by the opacity
curve setting section 14 and the color maps set by the color map
setting section 15 (S16). Then, the display control section 16
displays the pseudo three-dimensional image generated by the image
generation section 12 on the display unit 3 (S17), and the processing
is completed.
In the present embodiment, the region identification section
13 identifies the whole region representing a blood vessel from a
three-dimensional image, the opacity curve setting section 14 sets
a base opacity curve to the identified whole region, and obtains,
with respect to each of at least some of the pixels in the whole
region, a representative value of pixel values in an adjacent region
of a pixel concerned and sets an opacity curve obtained by changing
the base opacity curve using the obtained representative value as
the opacity curve to be applied to the pixel concerned in volume
rendering, the color map setting section 15 sets a base color map
to the identified whole region, obtains, with respect to each of
at least some of the pixels in the whole region, a representative
value of pixel values in an adjacent region of a pixel concerned,
and sets an color map obtained by changing the base color map using
the obtained representative value as the color map to be applied
to the pixel concerned in volume rendering, the image generation
section 12 generates a pseudo three-dimensional image by performing
volume rendering using these opacity curve and color map. This allows
an opacity curve and a color map appropriate for depicting the pixel
to be applied to each of all pixels in the region representing the
blood vessel, whereby an image depicting the three-dimensional
morphology of the blood vessel more accurately may be generated.
18
In the present embodiment, the description has been made of
a case in which both the opacity curve and the color map to be applied
to each pixel in the blood vessel region are obtained by modifying
the base opacity curve and the base color map respectively but an
arrangement may be adopted in which, for example, the base opacity
curve is used directly as the opacity curve to be applied to each
pixel and only the color map to be applied to each pixel is obtained
by modifying the base color map.
Further, in the embodiment described above, the description
has been made of a case in which the shift amount is calculated for
only some of the pixels in the blood vessel region
three-dimensionally sampled at a predetermined interval and the
estimated value of the shift amount is obtained by interpolation
for each of the other pixels, but an arrangement may be adopted in
which, with respect to each of all pixels in the blood vessel region,
a shift amount is obtained based on a subtraction value obtained
by subtracting a representative value of the pixel values in the
blood vessel region from a representative value of the pixel values
in an adjacent region of a pixel concerned and the opacity curve
and/or the color map to be applied to the pixel concerned is set
using the determined shift amount.
Still further, in the embodiment described above, the
description has been made of a case in which the shift amount or
the estimated value is obtained with respect to each pixel, but an
arrangement may be adopted in which a shift amount or an estimated
value is obtained with respect to each partial region constituted
by two or more pixels and the opacity curve and/or the color map
obtained based on the obtained shift amount or the estimated value
is applied to all pixels in the partial region. That is, the opacity
curve setting section 14 may be a section that obtains, with respect
to each of partial regions having a predetermined size and
constituting the whole region identified by the region
identification section 13, a representative value of pixel values
in partial region concerned and sets an opacity curve obtained by
modifying the base opacity curve using the obtained representative
19
value as the opacity curve to be applied to the partial region
concerned in volume rendering, or a section that obtains, with
respect to each of partial regions having a predetermined size and
constituting the whole region identified by the region
identification section 13, a representative value of pixel values
in an adjacent region that includes a partial region concerned and
sets an opacity curve obtained by modifying the set base opacity
curve using the obtained representative value as the opacity curve
to be applied to the partial region concerned in the volume rendering.
Likewise, the color map setting section 15 may be a section
that obtains, with respect to each of partial regions having a
predetermined size and constituting the whole region identified by
the region identification section 13, a representative value of pixel
values in partial region concerned and sets a color map obtained
by modifying the base color map using the obtained representative
value as the color map to be applied to the partial region concerned
in volume rendering, or a section that obtains, with respect to each
of partial regions having a predetermined size and constituting the
whole region identified by the region identification section 13,
a representative value of pixel values in an adjacent region that
includes a partial region concerned and sets color map obtained by
modifying the set base color map using the obtained representative
value as the color map to be applied to the partial region concerned
in the volume rendering.
Further, in each of the aforementioned embodiments, the
description has been made of a case in which the predetermined target
object is a blood vessel, but the predetermined target object may
be a structure having tubular structure, such as an intestine, a
bronchus, and the like, or various organs, such as a heart, a liver,
and the like.
Still further, in each of the aforementioned embodiments, the
description has been made of a case in which each of the representative
value of pixels in the adjacent region and the representative value
of pixels in the whole region is an average value, but each of the
representative value of pixels in the adjacent region and the
20
representative value of pixels in the whole region may be a mode
or median value of the pixels in each region, an average value of
pixels of all the pixels in each region which fall in a range of
predetermined width of pixel values from the mode or median value
of each region, or a mode or median value of pixels of all the pixels
in each region which fall in a range of predetermined width of pixel
values from the mode or median value of each region. At this time,
the representative value of the pixels in the whole region may be
a value of the same type as that of the representative value of the
pixels in the adjacent region or a value of different type from that
of the representative value of the pixels in the adjacent region.
Further, in a case where only a predetermined target object
of the entire three-dimensional image is to be the visualization
target region when generating an image by the volume rendering, if
the region representing the predetermined target object identified
by the region identification section 13 is dilated by one voxel or
so and the volume rendering is performed on the dilated region, the
surface of the predetermined target object may also be depicted by
the volume rendering.
21
We Claim:
1. An image generation apparatus, comprising:
an image generation section that generates a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using an opacity curve that defines the
relationship between pixel value and opacity;
a region identification section that identifies a whole region
representing a predetermined target object from the
three-dimensional image; and
an opacity curve setting section that sets a base opacity curve
to the identified whole region, obtains, with respect to each of
at least some pixels in the identified whole region, a representative
value of pixel values in an adjacent region of a pixel concerned
and sets an opacity curve obtained by modifying the base opacity
curve using the obtained representative value as the opacity curve
to be applied to the pixel concerned in the volume rendering.
2. The image generation apparatus as claimed in claim 1,
wherein the opacity curve setting section obtains, with respect to
each of the at least some pixels, a shift amount of opacity curve
based on a subtraction value obtained by subtracting a representative
value of pixel values in the whole region from the representative
value of pixels in the adjacent region of the pixel concerned, and
sets an opacity curve obtained by shifting the base opacity curve
by the obtained shift amount in a pixel value direction as the opacity
curve to be set to the pixel concerned in the volume rendering.
3. The image generation apparatus as claimed in claim 1 or
2, wherein the opacity curve setting section
obtains, with respect to each of some pixels obtained by
sampling the pixels in the whole region at a predetermined interval,
a representative value of pixels in an adjacent region of a pixel
concerned, obtains a shift amount of opacity curve based on a
difference between the obtained representative value and a
representative value of pixel values in the whole region, and sets
an opacity curve obtained by shifting the base opacity curve by the
obtained shift amount to a pixel value direction as the opacity curve
to be applied to the pixel concerned; and
obtains, with respect to each of pixels in the whole region
other than the sampled some pixels, an estimated value of shift amount
at a pixel concerned by interpolation using the shift amount of each
of the two or more sampled pixels located adjacent to the pixel
concerned and sets an opacity curve obtained by shifting the base
opacity curve by the determined estimated value in a pixel value
direction as the opacity curve to be applied to the pixel concerned
in the volume rendering.
4. The image generation apparatus as claimed in any of claims
1 to 3, wherein the representative value of the pixel values in the
adjacent region is a mode value, a median value, or an average value
of the pixel values in the adjacent region.
5. The image generation apparatus as claimed in any of claims
1 to 3, wherein the representative value of the pixel values in the
adjacent region is an average value of pixel values of all pixel
values in the adjacent region which fall in a range of predetermined
width of pixel values from the mode value or the median value of
the pixels in the adjacent region.
6. The image generation apparatus as claimed in any of claims
1 to 3, wherein the representative value of the pixel values in the
adjacent region is a mode value or a median value of pixel values
of all pixel values in the adjacent region which fall in a range
of predetermined width of pixel values from the mode value or the
median value of the pixels in the adjacent region.
7. The image generation apparatus as claimed in any of claims
1 to 6, wherein:
the image generation section generates a pseudo
three-dimensional image by performing volume rendering on the
three-dimensional image using a color map that defines the
relationship between pixel value and display color; and
the apparatus further comprises a color map setting section
that sets a base color map to the identified whole region and
calculates, with respect to each of at least some pixels in the
identified whole region, a representative value of pixel values in
an adjacent region of a pixel concerned, and sets a color map obtained
by modifying the set base color map using the calculated
representative value as the color map to be applied to the pixel
concerned in the volume rendering.
8. An image generation apparatus, comprising:
an image generation section that generates a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using a color map that defines the
relationship between pixel value and display color;
a region identification section that identifies a whole region
representing a predetermined target object from the
three-dimensional image; and
a color map setting section that sets a base color map to the
identified whole region, calculates, with respect to each of at least
some pixels in the identified whole region, a representative value
of pixel values in an adjacent region of a pixel concerned, and sets
a color map obtained by modifying the set base color map using the
calculated representative value as the color map to be applied to
the pixel concerned in the volume rendering.
9. The image generation apparatus as claimed in any of claims
1 to 8, wherein the predetermined target object is a blood vessel.
10. An image generation method that generates a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using an opacity curve that defines the
relationship between pixel value and opacity, wherein the method
comprises the steps of:
identifying a whole region representing a predetermined
target object from the three-dimensional image; and
setting a base opacity curve to the identified whole region,
obtaining, with respect to each of at least some pixels in the
identified whole region, a representative value of pixel values in
an adjacent region of a pixel concerned, and setting an opacity curve
obtained by modifying the base opacity curve using the obtained
representative value as the opacity curve to be applied to the pixel
concerned in the volume rendering.
11. An image generation method that generates a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using a color map that defines the
relationship between pixel value and display color, wherein the
method comprises the steps of:
identifying a whole region representing a predetermined
target object from the three-dimensional image; and
setting a base color map to the identified whole region,
calculating, with respect to each of at least some pixels in the
identified whole region, a representative value of pixel values in
an adjacent region of a pixel concerned, and setting a color map
obtained by modifying the base color map using the calculated
representative value as the color map to be applied to the pixel
concerned in the volume rendering.
12. An image generation program that causes a computer to
function as:
an image generation section that generates a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using an opacity curve that defines the
relationship between pixel value and opacity;
a region identification section that identifies a whole region
representing a predetermined target object from the
three-dimensional image; and
an opacity curve setting section that sets a base opacity curve
to the identified whole region, obtains, with respect to each of
at least some pixels in the identified whole region, a representative
value of pixel values in an adjacent region of a pixel concerned
and sets an opacity curve obtained by modifying the base opacity
curve using the obtained representative value as the opacity curve
to be applied to the pixel concerned in the volume rendering.
13. An image generation program that causes a computer to
function as:
an image generation section that generates a pseudo
three-dimensional image by performing volume rendering on a
three-dimensional image using a color map that defines the
relationship between pixel value and display color;
a region identification section that identifies a whole region
representing a predetermined target object from the
three-dimensional image; and
a color map setting section that sets a base color map to the
identified whole region, calculates, with respect to each of at least
some pixels in the identified whole region, a representative value
of pixel values in an adjacent region of a pixel concerned, and sets
a color map obtained by modifying the set base color map using the
calculated representative value as the color map to be applied to
the pixel concerned in the volume rendering.