DESCRIPTION
BACKGROUND OF THE INVENTION
Field of the Invention
The present invention relates to a three-dimensional display apparatus and a
three dimensional display method for displaying a volume rendering image of a
three-dimensional image composed of a plurality of tomographic images which have
been obtained by tomographic imaging of an object, as well as a program for causing a
computer to execute the three-dimensional image display method.
Description of the Related Art
In recent years, three-dimensional images of high quality are increasingly used
for diagnostic imaging, accompanying advances in medical instruments (for example,
multi-detector CT’s, and the like). Further, analysis of such three-dimensional image
facilitates grasping the three-dimensional shapes of various organs present in the
interior of the body, and further enables the relative positional relationships among the
respective tissues of arteries, veins, and tumors present in the organs and
three-dimensional structures thereof to be understood. In such a case, a specific organ
and a specific structure in the organ are extracted by using various image processing
algorithms and a three-dimensional shape is projected onto a two-dimensional plane by
display method such as volume rendering (VR) and the like so that the
three-dimensional structure thereof can be understood.
Here, when the three-dimensional display is displayed by VR, an organ, a
tissue, a structure, and the like of interest are extracted, and a color (R, G, B) and an
opacity level (opacity) are set for the signal value of each pixel, based on the signal
value (a CT value if the image is a CT image) at each voxel position in the
three-dimensional image of the extracted structure. In such a case, color templates, in
each of which a color and an opacity level are set according to a region of interest, are
preliminarily prepared, and a desired color template is selected depending on regions.
This enables a region of interest to be visualized in a volume rendering image (VR
image).
3
Meanwhile, there are also cases in which arrows that indicate the presence of
tumors, text indicating the name of each structure included in three-dimensional images,
and the like are added to the position of a corresponding structure as a label. Further, a
method in which a spine is extracted and a label is automatically added to the extracted
spine, and a method in which a bronchus is extracted and an anatomical medical term is
added to the extracted bronchus as a label have been proposed.
A method in which when displaying each of the three-dimensional images, to
which labels are added in such a manner as described above, text described in each label
is displayed by pointing to a position where the label is added has been proposed (refer
to Patent Document 1). In this method, the text includes the name of a segmented
anatomical structure, descriptions thereof, or abnormality thereof. Further, a method in
which while a doctor makes an observation by using three-dimensional images and
examines a subject by utilizing an endoscope, a label on which the doctor's observation
is described is displayed on an endoscopic image being displayed in the case that the
endoscope approaches the position to which the observation is added has been proposed
(refer to Patent Document 2).
[Prior Art Documents]
[Patent Documents]
[Patent Document 1]
PCT Japanese Publication No. 2010-500089
[Patent Document 2]
Japanese Unexamined Patent Publication No. 2011-206168
SUMMARY OF THE INVENTION
In the case that a small number of structures are included in a
three-dimensional image, it is possible to display all of the labels without any problems
because there are a small number of labels added to the structures. However, in the case
that labels are added to all of the various structures which are objects contained over a
wide range in a three-dimensional image, such as a chest or a chest and abdominal part
of the human body, there is a possibility that all labels cannot be displayed on a display
screen when the three-dimensional image is displayed. Further, in the case that all of the
labels are displayed, the labels added to the structures in the interior of the organs are
displayed in a state that only the appearance of the organs are visualized and the
4
interiors of the organs cannot be viewed. Therefore, it is impossible to understand to
which structure the labels are added. That is, there is no point in displaying the labels in
such a case. In this case, it can be considered to automatically switch between a display
mode or a non-display mode of each label according to the structures visualized in a
three-dimensional image. However, in the case that there are many structures contained
in a three-dimensional image, it is extremely troublesome to perform such an operation
because there are a large number of labels.
The present invention has been developed in view of the foregoing
circumstance. It is an object of the present invention to enable the display of labels
added to a three-dimensional image to be controlled without imposing a burden on a
user.
A three-dimensional image display apparatus according to the present
invention that displays a three-dimensional image of an object composed of a plurality
of structures, to each of which at least one label is added, comprising:
an image display control means that displays the three-dimensional image by
volume rendering;
a label display determination means that determines at least one label to be
displayed from a plurality of labels based on the opacity of the three-dimensional image
to be displayed by volume rendering;
a label display control means that adds at least one label determined to be
displayed to a corresponding structure and displays the label with the three-dimensional
image to be displayed by volume rendering.
The structures refer to various structures contained in the object represented by
the three-dimensional image. For example, in the case of a three-dimensional image of a
human body, the structures are not limited to structures, such as tumors and various
organs (a lung, a liver, a heart, a spleen, a pancreas, and the like) in the interior of a
human body, which constitute a specific region. The structures also include specific
positions, such as the center positions of tumors, vascular bifurcations, and the center
points of various organs.
Note that in the three-dimensional image display apparatus according to the
present invention, the label display determination means may determine that labels
added to structures are to be displayed in the case that the distance between a position at
5
which the three-dimensional image becomes opaque according to the opacity thereof
and a structure to which a label has been added is less than or equal to a specified value.
Further, in the three-dimensional image display apparatus according to the
present invention, the label display control means may be means that controls the
position of the at least one label to be displayed for each structure when the at least one
label determined to be displayed is added to a plurality of structures.
Further, in the three-dimensional image display apparatus according to the
present invention, the label display control means may cause the at least one label to be
added only to a portion having a specified area or greater and displayed when the
structure is divided into a plurality of portions having the identical label to be displayed
and is present in the three-dimensional image to be displayed by volume rendering.
Further, in the three-dimensional image display apparatus according to the
present invention, the label display control means may be means that causes the at least
one label to be added only to a portion having the largest area and to be displayed when
the special creature is divided into a plurality of portions having the identical label to be
displayed and is present in the three-dimensional image to be displayed by volume
rendering.
Further, the three-dimensional image display apparatus according to the present
invention may further include label adding means that adds at least one label to the
three-dimensional image.
A three-dimensional image display method according to the present invention,
of displaying a three-dimensional image of an object composed of a plurality of
structures, to each of which at least one label is added, comprising:
displaying the three-dimensional image by volume rendering;
determining at least one label to be displayed from a plurality of labels based
on the opacity of the three-dimensional image to be displayed by volume rendering; and
adding the at least one label determined to be displayed to a corresponding
structure and displaying the label with the three-dimensional image to be displayed by
volume rendering.
Note that the three-dimensional image display method may be provided as a
program for causing a computer to execute the three-dimensional image display method.
According to the present invention, when a three-dimensional image is
6
displayed by volume rendering, the labels to be displayed are determined from among a
plurality of labels based on the opacity of the three-dimensional image to be displayed
by volume rendering, and then the labels determined to be displayed are added to the
respective corresponding structures and displayed with the three-dimensional image.
This enables the display of the labels added to the three-dimensional images to be
controlled without the necessity of a user’s operation, resulting in the burden on the user
when displaying labels being reduced.
BRIEF DESCRIPTION OF THE DRAWINGS
Figure 1 is a block diagram that schematically illustrates the configuration of a
three-dimensional image display apparatus of an embodiment of the present invention.
Figure 2 is a diagram that describes the process for determining a label to be
displayed.
Figure 3 is a flow chart that illustrates the process carried out in the present
embodiment.
Figure 4 is a diagram that illustrates a state in which a label is displayed in a
VR image.
Figure 5 is a diagram that illustrates a state in which the opacity of the body
surface is changed in the VR image illustrated in Figure 4.
Figure 6 is a diagram that illustrates a state in which the opacity of the right
upper lobe is changed in the VR image illustrated in Figure 5.
Figure 7 is a diagram that illustrates a state in which a line of sight is changed
in the VR image illustrated in Figure 6.
Figure 8 is a diagram that illustrates a state in which a line of sight is changed
in the VR image illustrated in Figure 7.
DESCRIPTION OF THE PREFERRED EMBODIMENTS
Hereinafter, an embodiment of the present invention will be described with
reference to the drawings. Figure 1 is a block diagram that schematically illustrates the
configuration of a three-dimensional image display apparatus of an embodiment of the
present invention. Note that the configuration of a three-dimensional image display
apparatus 1 illustrated in Figure 1 is realized by causing a three-dimensional image
display program read into an auxiliary storage device to execute on a computer. This
program is recorded in recording media such as CD-ROM’s and the like or is
7
distributed via a network such as the Internet to be installed in computers.
The three-dimensional image display apparatus 1 according to the present
embodiment includes an image obtainment unit 10, a storage unit 12, a structure
extraction unit 14, a label adding unit 16, an image display control unit 18, a label
display determination unit 20, a label display control unit 22, an input unit 24 and a
display unit 26.
The image obtainment unit 10 functions as a communication interface which
obtains a three-dimensional image V0 acquired by imaging a chest of a subject in a
modality 2 such as multi-slice CT apparatuses, MRI apparatuses, or the like. Note that
the modality 2 is a multi-slice apparatus in this embodiment. Further, the
three-dimensional image group V0 is delivered via a LAN from the modality 2.
Here, the three-dimensional image V0 is obtained by laminating
two-dimensional tomographic images which are sequentially obtained along the
direction perpendicular to the tomographic sections of the chest which is a target for
diagnosis. In the present embodiment, the three-dimensional image V0 is generated by
overlapping a plurality of tomographic images acquired by the modality 2. Note that a
three-dimensional image which has been obtained by using the CT apparatus is data in
which the amount of X-ray absorption is stored for each voxel (i.e., a pixel position)
that constitutes lattice points in a three-dimensional space. In the data, one signal value
(when the CT apparatus is applied for imaging, the value represents the amount of the
X-ray absorption) is provided for each pixel position.
Note that the three-dimensional image V0 is added to with supplemental
information specified by the DICOM (Digital Imaging and Communications in
Medicine) specification. For examples, the supplemental information may include an
image ID for identifying a three-dimensional image, a patient ID for identifying a
subject, an examination ID for identifying an examination, a unique ID (UID) assigned
to each piece of image information, the examination date on which the image
information has been generated, the examination time, the kind of a modality which has
been used in the examination to obtain the image information, information regarding the
patient such as the name of the patient, age, gender, and the like, a site to be examined
(a site to be imaged, a chest in the present embodiment), imaging conditions (whether
contrast agent is used or not, the amount of radiation, and the like), a series number or
8
an obtainment number when a plurality of images has been obtained for one
examination.
The storage unit 12 is a large capacity storage device such as a hard disk and
stores three-dimensional images V0 therein.
The structure extraction unit 14 extracts a body surface region, a lung region, a
bronchus, and a pulmonary nodule from a three-dimensional image V0 of a chest, as
structures. The body surface region is extracted by estimating the range of a signal value,
in which a body surface is considered to be present, with respect to a signal value (i.e.,
CT value) at each pixel position of the three-dimensional image V0 and by performing
threshold value processing using the estimated value of the range. A method for
extracting the lung region, in which air is present in the lung field, may apply an
arbitrary method such as a method in which a histogram of a signal value of each pixel
position in the three-dimensional image V0 is generated and the lung region is subjected
to the threshold processing, a region expanding method which is based on seed points
that represent the lung region, and the like. The extracted lung region is then separated
into five lobes: a right upper lobe, a right middle lobe, a right lower lobe, a left upper
lobe and a left lower lobe. This separation may be performed by causing the display unit
26 to display the extracted lung field and by the user manually tracing interlobar
membrane from the input unit 24. Alternatively, a method for automatically extracting
interlobar membrane described in Y. Sato et al., “Extraction of Lung Lobes in X-ray CT
images and its Application to Evaluation of Heavy Ion Radiation Therapy”, “MEDICAL
IMAGING TECHNOLOGY, Vol. 22, No.5, 2004, and the like may be applied. Note
that the methods for extracting the lung region is not limited to these methods described
above, but an arbitrary method may be applied.
A method for extracting the bronchus may be a method in which an assembly
of pixels within the bronchus region is extracted by the region expanding method, and
thinning processing is conducted on the extracted bronchus region. Further, in such a
method, the respective pixels on thin lines are classified into end points, edges (sides),
and branch points based on the connecting relationship of the thin lines that represent
the obtained bronchus so that tree-structure data which represents the bronchus is
obtained. Further, as described in N. Kawamura et al., "Examination of Bronchus
Extraction Algorithm using Multi Slice CT Images", THE INSTITUTE OF
9
ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS,
Technical Report, MBE, ME and bio-cybernetics, Vol. 105, No. 221, pp. 11-14, 2005,
and the like, a method for automatically extracting the bronchus may be applied.
Further, in the present embodiment, anatomical name of eyes is performed for
each bifurcation in the extracted bronchus. In this case, a user may perform such
nomenclature manually. Alternatively, a method for automated nomenclature as
described in K. Mori et al., “A method for automated nomenclature of bronchial
branches extracted from CT images”, International Congress Seried, Vol. 1281, pp.
86-91, 2005, and the like may be applied. Through such methods, the right and left sides
of the bronchus are named as apical branch (B1), posterior bronchus (B2), anterior
bronchus (B3) … posterior basal (B10).
Methods for extracting pulmonary nodules as described in Y. Li et al.,
“Interactive Segmentation of Lung Nodules using AdaBoost and Graph Cuts”,
FOURTH INTERNATIONAL WORKSHOP ON PULMONARY IMAGE ANALYSIS,
pp. 125-133, C. Schneider et al., “Automated lung nodule detection and segmentation”,
Medical Imaging, Proc. of SPIE, Vol. 7260, 2009, and the like may be applied.
Alternatively, a user may manually extract pulmonary nodules by using the input unit
24.
The label adding unit 16 adds labels to the extracted structures in response to
the user’s input from the input unit 24. The contents of each label may include
observations in addition to an anatomical name when the structure is a tumor, or the like.
Alternatively, an arrow that represents the position of a tumor, or the like may be
included. In particular, a label describing the text “skin” is added to a body surface
region, labels respectively describing the texts “right upper lobe”, “right middle lobe”,
“right lower lobe”, “left upper lobe”, and “left lower lobe” are added to five lobes of a
lung region, and a label describing the text “bronchus” is added to a bronchus region.
Further, the anatomical nomenclature is performed on the bronchus in the present
embodiment, and labels of the anatomical names, “B1” through “B10” are added to the
bronchus accordingly. In addition, a label of an observation, indicating the text “a solid
shade of 10 mm”, is added to a pulmonary nodule region.
Note that, the addition of labels refers to mean that a plurality of pixels, which
belong to an extracted structure in the three-dimensional image V0, are correlated to the
10
text of a label. Such an operation enables the added label to be viewed when any one of
the pixels included in the extracted structure is designated. Conversely, when the label
is designated, the structure to which the label has been added will be viewed. The
three-dimensional image V0 to which labels are added will be stored in the storage unit
12.
The image display control unit 18 displays a volume rendering (VR) image of a
three-dimensional image V0 by using the volume rendering method. In other words, the
image display control unit 18 emits a virtual light beam from a projection plane toward
the three-dimensional image V0 and generates a three-dimensional image by virtual
reflected light from the interior of the object, based on the colors (R, G, B) and opacity
corresponding to the respective signal values in the three-dimensional image V0. Then,
the image display control unit 18 further generates a projection image, which enables
seeing through a three-dimensional structure in the interior of the object, on the
projection plane from the three-dimensional image and displays this projection image as
a volume rendering image. Note that the colors and opacity are defined in a
predetermined color template, and the signal values at the respective pixel positions in
the three-dimensional image V0 are converted into pixel values of the projection image,
based on the colors and opacity set according to the predetermined color template by the
alpha blending method. Note that during the display of the volume rendering (VR)
image, when a user issues an instruction to change the color template or an instruction
to change the opacity from the body surface toward the interior such that the structure of
the interior of a chest is gradually displayed, the image display control unit 18 changes
an aspect of the VR image of the three-dimensional image V0, based on the colors and
opacity set according to a color template or based on a designated opacity.
The label display determination unit 20 determines a label/labels to be
displayed together with a three-dimensional image to be displayed by VR (hereinafter
referred to as a VR image) from a plurality of labels added to the three-dimensional
image based on the opacity of the three-dimensional image V0 to be displayed by VR.
Hereinafter, the process for determining label(s) to be displayed will be described.
In the present embodiment, colors for the pixels of a VR image are determined
by the alpha blending method when displaying the VR image. In other words, a ray
travels to the position where the ray attenuates to become 0, i.e., to the surface of an
11
object while signal values within the three-dimensional image V0, which are present on
the ray vector represented by the ray, and the values of the opacity respectively
corresponding to the signal values are subjected to alpha blending with respect to pixels
on the projection plane. The label display determination unit 20 calculates the distances
between a pixel Pij of the surface of an object in the three-dimensional image V0 and
pixels Li along the ray vector of the three-dimensional image V0. The pixels L1 are
pixels in the interior of all of the structures to which labels are added. Then, the label
display determination unit 20 calculates a pixel Li_min at which the distance to the
pixel Pij is the shortest within each structure, and compares the calculated distance
Dmin between the shortest point Li_min and the pixel Pij to a threshold value Th1 (for
example, 1cm). If the distance Dmin is less than or equal to the threshold value, label(s)
added to the structure Li will be displayed.
Figure 2 is a diagram that describes the process for determining a label to be
displayed. Note that Figure 2 illustrates a two-dimensional image to describe the
process. As illustrated in Figure 2, structures L1 and L2, to which labels are added, are
disposed in this order along the ray vector Ve passing through a point Pij on the surface
of an object. Further, if the distance between the pixel Pij on the surface and the
structure L1 is less than or equal to the threshold value Th1 and the distance between
the pixel Pij and the structure L2 exceeds the threshold Th1, the label display
determination unit 20 will determine that label(s) added to the structure L1 are to be
displayed.
All the pixels (x, y) of the surface over the projection plane are subjected to the
process described above, and thereby a label map is generated, in which whether
label(s) are displayed on the projection plane is defined. It is preferable for this label
map to indicate 1 for each pixel (x, y) on the projection plane when label(s) are to be
displayed and to indicate 0 for each pixel (x, y) on the projection plane when a label is
not to be displayed. Further, pixels indicated as 1 are subjected to connected component
processing and connected components composed of the pixels indicated as 1 are
generated. Then, the number of pixels for each connected component is counted. If the
number of the pixels is greater than or equal to a threshold value Th2 (for example, 100),
label(s) added to a structure corresponding to the connected component will be
determined to be displayed on a region composed of the connected component.
12
The label display control unit 22 overlays and displays label(s) added to a
structure/structures corresponding to the connected component, label(s) added to which
have been determined to be displayed, on a VR image being displayed. The position at
which a label is displayed may be anywhere within the region of the connected
component, the label added to which is to be displayed. However, when the position of
the center of gravity of the connected component is within the connected component,
the label should be displayed at the position. When the position of the center of gravity
is not within the connected component, the label should be displayed at a position
within the connected component, which is closest to the position of the center of gravity.
The label may be displayed as it is at the display position. However, it is preferable for
the label to be displayed with a reference line drawn from the display position. Further,
when a plurality of labels are displayed, it is preferable for the labels to be displayed in
the range that radially expands from the center of the VR image (i.e., the center of the
projection plane).
Note that when the plurality of the labels are displayed, there are cases that
display positions of the labels overlap with each other. In such a case, the label display
control unit 22 moves the display position of the connected component having a larger
area to a position where the labels do not overlap with each other, and thereby controls
display of the labels so as to enable all the labels to be seen at the same time.
The input unit 24 includes a known input device such as a keyboard, a mouse,
and the like.
The display unit 26 includes a known display device such as a liquid crystal,
CRT, and the like.
Next, the process carried out in the present invention will be described. Figure
3 is a flow chart that illustrates the process carried out in the present embodiment. Note
that the image obtainment unit 10 obtains the three-dimensional image V0, and the label
adding unit 16 adds label(s) thereto before the storage unit 12 stores the
three-dimensional image V0 therein. Further, the following will describe a case, in
which a body surface is displayed, and then the opacity of the body surface is changed
such that structures in the interior of the body surface will be sequentially displayed by
VR. When a user operates the input unit 24, the image display control unit 18 causes the
display unit 26 to display a VR image, onto which a structure/structures designated by
13
the user have been projected (step ST1). Then, the label display determination unit 20
generates a label map as described above, and performs the connected composition
processing to determine label(s) to be displayed (step ST2). Further, the label display
control unit 22 displays label(s) on the VR image (step ST3).
Figure 4 is a diagram that illustrates a state in which a label is displayed in a
VR image. The VR image of the body surface, which is opaque, is displayed in the first
place. As illustrated in Figure 4, the label of the text "skin" is displayed with a leader
line drawn from the position of the center of gravity of the body surface. In this case, as
the body surface is projected onto the substantially entire surface of the VR image, the
leader line is drawn from the substantially center position of the VR image to display
the label. Note that when the body surface is opaque, the organs in the interior thereof
cannot be viewed.
In such a state, a determination is made whether an instruction to change the
opacity or a line of sight is issued (step ST4). If an affirmative determination is made at
ST 4, the operation returns to step ST2 so that step ST2 and step ST3 will be repeated.
If a user issues an instruction to gradually decrease the opacity of the body surface from
the state illustrated in Figure 2, the body surface in the VR image will gradually change
from opaque to transparent. Then, the rays on the projection plane pass through the
body surface and travel to the surface of the lung region and bronchus so that a VR
image, onto which the lung region and bronchus are projected, will be displayed. In
addition, labels are added to the respective five lobes of the lung region and the
bronchus in the present embodiment. When the VR image of the surfaces of the lung
region and the bronchus are displayed, the labels are added to the respective five lobes,
which constitute the lung region, and the bronchus as illustrated in Figure 5. Note that
the labels are displayed with leader lines drawn from the position of the center of
gravity of the respective five lobes and the bronchus, the leader lines being radially
drawn with the center of the VR image as reference.
Further, if a user issues an instruction to change the opacity of the right upper
lobe to be transparent, the portion of the right upper lobe will not be displayed and a VR
image including the bifurcations of the bronchus, which were hidden under the right
upper lobe, will be displayed instead. In this case, a VR image illustrated in Figure 6 do
not display the label of the right upper lobe, but displays the labels of the anatomical
14
nomenclature B1 through B3 added to the bronchus and the label of the observation
added to a pulmonary nodule which has been found in the bronchus.
The line of sight is changed and the projection plane is rotated from the state
illustrated in Figure 6 in such a manner to display a VR image with the lung region
being observed from slightly upward. Then, it can be seen that left lower lobe is divided
into two regions by the bronchus. In this case, the size of connected components (which
have a value of 1) corresponding to the left lower lobe are compared to each other in the
label map, and a larger region is added to with a label at the position of the center of
gravity thereof and a smaller region is added to with a semitransparent label at the
position of the center of gravity thereof. Figure 7 illustrates this state. Figure 7
represents the semitransparent label surrounded by the broken lines. Note that the label
may be added only to the larger region, not to the smaller region.
Further, Figure 8 illustrates a state in which a line of sight has been changed
from the VR image illustrated in Figure 7. Figure 8 does not include the right middle
lobe region. Therefore, the lable of the right middle lobe is not to be displayed.
Returning to the flow chart, if a negative determination is made in step ST4, a
determination is made whether an instruction to complete the process was issued (step
ST5). If a negative determination is made in step ST5, the operation will return to step
ST4. If an affirmative determination is made in step ST5, the process will be completed.
In such a manner as described above, in the present embodiment, label(s) to be
displayed are determined from a plurality of labels based on the opacity when
displaying a three-dimensional image V0 by VR. Then, label(s) determined to be
displayed are added to a corresponding structure and displayed by VR together with the
three-dimensional image V0. This enables the display of the labels added to the
three-dimensional image V0 to be controlled without the necessity of a user's work,
resulted in the burden on the user being reduced when displaying the labels.
Note that in the embodiment above, pulmonary arteries, pulmonary veins, ribs,
and a spine may be extracted further and labels may be added thereto. In such a case as
well, label(s) to be displayed are determined, and displayed with overlapped on the VR
image in the same manner as described above.
Further, the process to be carried out for when a VR image of a lung region is
displayed was described in the embodiment above. It is a matter of course that the
15
present invention can be applied to the case in which a heart is extracted and a VR
image of the heart region is displayed. Further, the present invention can be also applied
to the case in which a three-dimensional image of an abdominal part is displayed. For
example, a target in the abdominal part is designated as a liver region. The liver region,
a hepatic an artery, a hepatic vein, a portal vein, and a tumor are extracted from the
three-dimensional image V0, and a label is added to each of the liver region, hepatic an
artery, hepatic vein, portal vein, and tumor. Labels to be displayed may be determined
based on the opacity of structures to be displayed when the VR image is displayed.
Here, the heart region is extracted by estimating the range of a signal value, in
which the heart is present in the three-dimensional image V0, and then carrying out
threshold processing using the value of the range.
Further, a method for extracting the liver region may apply a method in which
the range of CT values where the liver is present in the three-dimensional image V0 is
estimated, the threshold processing is conducted by using the value of the range, and a
morphology filter is applied to the extracted region. As described in J. Masumoto et al.,
"Automated Liver Segmentation Method for Dynamic CT Data Using Non-Rigid
Registration", Journal of Computer Aided Diagnosis of Medical Images", Vol.7, No.4-1,
2003, a method in which contrast patterns of the liver region are detected by utilizing a
plurality of phase images of the liver taken in chronological order and the liver region is
detected by using the detected contrast patterns is also applied. Further, level set
methods as described in P.S. Sulaiman et al., "A Liver Level Set (LLS) Algorithm for
Extracting Liver's Volume Containing Disconnected Regions Automatically”, IJCSNS
International Journal of Computer Science and Network Security", Vol. 8, No. 12, 2008
and T. Hitosugi et al., "Development of a liver extraction method using a level set
method and its performance evaluation", Computer Aided Diagnosis of Medical Images,
Vol. 7, No. 4-2, 2003 can be applied. Note that the method for extracting the liver
region of the present invention is not limited to these methods, but an arbitrary method
may be applied.
Further, as a method for extracting a hepatic artery, a hepatic vein, and a portal
vein (hereinafter, there are cases that these are referred to simply as blood vessels), a
method in which a main axis direction and the positional information regarding a
plurality of candidate points that represent a target tissue composed by a linear structure
16
are calculated and the plurality of the candidate points are reconstructed to be connected
with each other by using a cost function which is a variable based on the calculated
positional information and main axis direction, as disclosed in Japanese Unexamined
Patent Publication No. 2010-220742, for example may be applied. Further, a method for
automatically discriminating blood vessels from each other and extracting them as
disclosed in Japanese Unexamined Patent Publication No. 2011-212314 may also be
applied.
Further, a method for extracting blood vessels may apply a method in which a
tree-structure is generated by connecting the respective nodes with each other from a
first root node corresponding to a first tree-structure root node and a second root node
corresponding to a second tree-structure root node based on the characteristics of the
blood vessels which repeatedly branch from an origin of each of a first and a second
linear structures and extends in directions away from the origin in such a manner to
become wider. In this method, a cost function is used such that a cost which represents
ease of connection with respect to a plurality of edges, each of which is capable of
connecting with each node, and which bind a plurality of nodes together is weighted for
each node. Further, in this method, the first and second linear structures are designated
as the hepatic artery and the hepatic vein, respectively so that the hepatic artery and the
hepatic vein can be discriminated and extracted. Further, in this method, the first and
second linear structures are designated as the portal vein and hepatic artery, and the
hepatic vein, respectively so that the portal vein and hepatic artery, and the hepatic vein
can be discriminated and extracted. Note that in this method, the origin may be
identified by an arbitrary method, and the root node corresponding to the origin may be
identified by a well-known method based on the origin. For example, the origin may be
designated on a displayed image by an input device such as a mouse, or the like.
Alternatively, an origin detection unit may be applied for detecting an origin. The origin
detection unit detects the origin by mechanically learning a plurality of teacher data,
which represents that the origin is a known predetermined structure. Note that various
known methods for extracting a root node by mechanically learning teacher data may be
applied. For example, the Adaboost method can detect an origin based on the amount of
characteristics of a known origin in teacher data.
Further, a tumor can be extracted by methods that utilize the Voxel
17
Classification described in M. Freiman et al., "Liver tumors segmentation from CTA
images using voxels classification and affinity constraint propagation", Int J CARS,
2010. Note that methods for extracting hepatic arteries, hepatic veins, portal veins, and
tumors are not limited to these methods, but an arbitrary method may be applied.
Further, the labels including the texts are displayed in the embodiments above.
Only the arrows that represent the positions of tumors may be displayed as the labels.
The reproduction of the additional information when the additional information
is added to the three-dimensional image of the human body was described in the
embodiment mentioned above. It is a matter of course that the present invention can be
applied to the case in which additional information is added to a three-dimensional
image of the topography data, the case in which additional information is added to a
three-dimensional image of cloud in the weather data, or the case in which additional
information is added to three-dimensional images of various components in
nondestructive inspection.

We Claim
1. A three-dimensional image display apparatus that displays a
three-dimensional image of an object composed of a plurality of structures, to each of
which at least one label is added, comprising;
image display control means that displays the three-dimensional image by
volume rendering;
label display determination means that determines at least one label to be
displayed from a plurality of labels based on the opacity of the three-dimensional image
to be displayed by volume rendering;
label display control means that adds the at least one label determined to be
displayed to a corresponding structure and displays the label with the three-dimensional
image to be displayed by volume rendering.
2. The three-dimensional image display apparatus of Claim 1, wherein the label
display determination means determines that labels added to structures are to be
displayed in the case that the distance between a position at which the three dimensional
image becomes opaque according to the opacity thereof and a structure to which a label
has been added is less than or equal to a specified value.
3. The three-dimensional image display apparatus of Claim 1 or 2, wherein the
label display control means controls the position of the at least one label to be displayed
for each structure when the at least one label determined to be displayed is added to a
plurality of structures.
4. The three-dimensional image display apparatus of any one of Claims 1
through 3, wherein the label display control means causes the at least one label to be
added only to a portion having a specified area or greater and displayed when the
structure is divided into a plurality of portions having the identical label to be displayed
and is present in the three-dimensional image to be displayed by volume rendering.
5. The three-dimensional image display apparatus of any one of Claims 1
through 3, wherein the label display control means causes the at least one label to be
added only to a portion having the largest area and to be displayed when the structure is
divided into a plurality of portions having the identical label to be displayed and is
present in the three-dimensional image to be displayed by volume rendering.
6. The three-dimensional image display apparatus of any one of Claims 1
19
through 4 that further comprises label adding means that adds at least one label to the
three-dimensional image.
7. A three-dimensional image display method of displaying a
three-dimensional image of an object composed of a plurality of structures, to each of
which at least one label is added, comprising:
displaying the three-dimensional image by volume rendering;
determining at least one label to be displayed from a plurality of labels based
on the opacity of the three-dimensional image to be displayed by volume rendering; and
adding the at least one label determined to be displayed to a corresponding
structure and displaying the label with the three-dimensional image to be displayed by
volume rendering.
8. A program for causing a computer to execute a three-dimensional image
display method for a three-dimensional image of an object composed of a plurality of
structures, to each of which at least one label is added, comprising the steps of:
displaying the three-dimensional image by volume rendering;
determining at least one label to be displayed from a plurality of labels based
on the opacity of the three-dimensional image to be displayed by volume rendering; and
adding the at least one label determined to be displayed to a corresponding
structure and displaying the label with the three-dimensional image to be displayed by
volume rendering.