FIELD
The present invention relates to the technical field of computed tomography (CT), particularly to a method and apparatus for computing noise on a CT image.
BACKGROUND
When post-processing is performed on an image generated by computed tomography (CT), noise therein usually needs to be estimated. One existing method for estimating noise is to use a high-frequency information value within one fixed region of interest on the image as noise, and then to perform a de-noising processing on the entire image by said high-frequency information value.
When there is a larger change in a proportion of soft tissues in an object on which CT scan is performed in a scanning direction, the use of the above method for estimating noise will result in a too smooth image on which de-noising processing has been performed based thereupon, i.e., some details will be lost. For example, when CT scan is performed on the head, since the posterior cranial fossa or cranial base portions have denser bones while the calvarium has more soft tissues, there will be an obvious change in a proportion of soft tissues thereof along a Z-axis direction of CT scan. If these CT images of denser bones are de-noised by the above existing method, a problem that the image is excessively smoothed will occur.
SUMMARY
An objective of the present invention is to provide a novel method and apparatus for computing noise on a CT image, which is capable of solving the technical problem that the CT image is excessively smoothed.

One embodiment of the present invention provides a method for computing noise on a CT image, comprising: selecting a region of interest on a current CT image; computing a proportion of soft tissues within the region of interest; and when the proportion is less than a preset proportion threshold of soft tissues, computing said noise according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image.
Another embodiment of the present invention provides an apparatus for computing noise on a CT image, comprising: a region-of-interest selecting module for selecting a region of interest on a current CT image; a proportion of soft tissues within a region of interest computing module for computing a proportion of soft tissues within the region of interest; and a module that computes noise by a noise model, for when the proportion is less than a preset proportion threshold of soft tissues, computing said noise according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image.
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention can be better understood in light of the following description of embodiments of the present invention with reference to the accompanying drawings, in which:
Fig. 1 is a schematic flow chart illustrating one embodiment of a method for computing noise on a CT image according to the present invention;
Fig.2 is a schematic flow chart illustrating one embodiment of a step of selecting a region of interest on a current CT image during the process of computing noise on a CT image according to the present invention;
Fig. 3 is a schematic flow chart illustrating one embodiment of determining a range of the current CT image during the process of selecting a region of interest on the current CT image according to the present invention;

Fig. 4 is a schematic flow chart illustrating one embodiment of a step of determining a proportion threshold of soft tissues according to a relationship between a change in a proportion of soft tissues of a scanned object along a certain scanning direction and a change in a corresponding noise value during the process of computing noise on a CT image according to the present invention;
Fig. 5 is a schematic flow chart illustrating one embodiment of a step of computing noise according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image when the proportion is less than a preset proportion threshold of soft tissues during the process of computing noise on a CT image according to the present invention;
Fig. 6 is a schematic block diagram illustrating one embodiment of an apparatus for computing noise on a CT image according to the present invention;
Fig. 7A is a CT image of a region with more head bones whose noise has not been removed;
Fig. 7B is a CT image of a region with more head bones whose noise has been removed by the prior art;
Fig. 7C is a CT image of a region with more head bones whose noise has been removed by a technical solution of the present invention;
Fig. 8 is a schematic diagram illustrating a range of the CT image and a region of interest selected based upon this range;
Fig. 9A is a diagram illustrating distribution of accumulated values that is obtained by accumulating pixels along an X-axis direction of the CT image;
Fig. 9B is a diagram illustrating distribution of accumulated values that is obtained by accumulating pixels along a Y-axis direction of the CT image;
Fig. 10 is a diagram illustrating a plurality of CT images that are obtained by scanning a head along a Z-axis direction of CT scan;

Fig. 11A is a diagram illustrating a change in a proportion of soft tissues of the plurality of images as shown in Fig. 10;
Fig. 11B is a diagram illustrating a change in noise values of the plurality of images as shown in Fig. 10.
DETAILED DESCRIPTION
Hereafter, a detailed description will be given for preferred embodiments of the present disclosure. It should be pointed out that in the detailed description of the embodiments, for simplicity and conciseness, it is impossible for the Description to describe all the features of the practical embodiments in details. It should be understood that in the process of a practical implementation of any embodiment, just as in the process of an engineering project or a designing project, in order to achieve a specific goal of the developer and in order to satisfy some system-related or business-related constraints, a variety of decisions will usually be made, which will also be varied from one embodiment to another. In addition, it can also be understood that although the effort made in such developing process may be complex and time-consuming, some variations such as design, manufacture and production on the basis of the technical contents disclosed in the disclosure are just customary technical means in the art for those of ordinary skilled in the art relating to the contents disclosed in the present invention, which should not be regarded as insufficient disclosure of the present invention.
Unless defined otherwise, all the technical or scientific terms used in the Claims and the Description should have the same meanings as commonly understood by one of ordinary skilled in the art to which the present invention belongs. The terms “first”, “second” and the like in the Description and the Claims of the present utility model do not mean any sequential order, number or importance, but are only used for distinguishing different components. The terms “a”, “an” and the like do not denote a limitation of quantity, but denote the

existence of at least one. The terms “comprises”, “comprising”, “includes”, “including” and the like mean that the element or object in front of the “comprises”, “comprising”, “includes” and “including” covers the elements or objects and their equivalents illustrated following the “comprises”, “comprising”, “includes” and “including”, but do not exclude other elements or objects. The term “coupled” or “connected” or the like is not limited to being connected physically or mechanically, nor limited to being connected directly or indirectly.
In order to make the purpose, the technical solutions and the advantages of the invention more apparent, the technical solutions of the present invention will be set forth clearly and fully in the following by combining with specific embodiments of the invention and the corresponding accompanying drawings. Obviously, the described embodiments are merely part— not all— of the embodiments in the present invention. In view of the embodiments in the present invention, other embodiments made by one of ordinary skilled in the art without inventive work all fall within the scope of protection of the invention.
According to an embodiment of the present invention, a method for computing noise on a CT image is provided.
With reference to Fig. 1, Fig. 1 is a schematic flow chart illustrating one embodiment of a method 100 for computing noise on a CT image according to the present invention. The method 100 may comprise the following Steps 101-103.
As shown in Fig. 1, in Step 101, a region of interest is selected on a current CT image.
The so-called current CT image means a CT image whose noise needs to be computed.
One objective of selecting the region of interest is to count up a proportion of soft tissues of a scanned object (e.g., a certain organ of a body) within the region of interest.

Since the scanned object might deviate from a scanning center during the process of CT scan, in one embodiment of the present invention, Step 101 may further include the following Sub-steps 201-202 with reference to Fig. 2.
In Sub-step 201, a range of the current CT image is determined.
The range of the CT image as described herein means a distribution range of the scanned object on said CT image. With reference to Fig. 8, a region enclosed by a larger rectangular block in Fig. 8 is just one embodiment of the range of the CT image as described herein.
In one embodiment of the present invention, with reference to Fig. 3, Sub-step 201 may further comprise Sub-steps 301-302.
In Sub-step 301, pixel values of the current CT image are accumulated along a coordinate axis direction.
In one embodiment of the present invention, the pixel values of the current CT image may be accumulated along X-axis and Y-axis directions of the current CT image respectively. Accumulation results may be represented by distribution diagrams of accumulated values as shown in Fig. 9A and Fig. 9B respectively. A horizontal axis of the distribution diagram may represent coordinate values along the X-axis and Y-axis directions and a vertical axis thereof may represent the accumulated values.
In Sub-step 302, a range is determined according to the accumulation result.
In the distribution diagram obtained by Sub-step 301, a corresponding horizontal axis region in which the accumulated values are not zero or greater than a certain preset threshold may be regarded as the range of the current CT image.
In Sub-step 202, a central point of the range is used as a central point of the region of interest.

For one geometric figure whose range has been determined, the position of its central point is determined, and thus said central point may be used as the central point of the region of interest.
Certainly, the region of interest may also be of other shapes in addition to rectangle. No matter what shape is used, a size of the region of interest (e.g., a length and a width of a rectangle, a radius of a circle, etc.) will usually be set in advance. Hence, as long as the position of the central point of the region of interest has been determined, the region of interest may also be determined . As shown in Fig. 8, the smaller rectangular block therein is just the region of interest obtained by performing Step 101.
In Step 102, a proportion of soft tissues within the region of interest is computed.
In one embodiment of the present invention, a proportion of pixel points of a low-frequency portion to the total pixel points within the region of interest may be estimated, which is used as the proportion of soft tissues with the region of interest. Or, the number of pixel points of a high-frequency portion within the region of interest may also be estimated at first, and then the number of the remaining pixel points may be obtained, such that a ratio of the number of the remaining pixel points to the total pixel points may be used as the proportion of soft tissues within the region of interest.
In one embodiment of the present invention, when the proportion of soft tissues within the region of interest is greater than a preset proportion threshold of soft tissues, noise on the CT image may be estimated by the existing method.
In one embodiment of the present invention, the proportion threshold of soft tissues may be determined according to a relationship between a change in the proportion of soft tissues of the scanned object along a certain scanning direction and a change in a corresponding noise value. The setting of the proportion threshold of soft tissues may be performed at any time before Step 103. In one

embodiment of the present invention, with reference to Fig. 4, the proportion threshold of soft tissues may be set by performing the following Sub-steps 401-404.
In Sub-step 401, for the same scanned object, a plurality of CT images are acquired along a direction in which the proportion of soft tissues of the scanned object changes.
For example, as shown in Fig. 10A, if the scanned object on the current CT image whose noise needs to be computed is a head, a plurality of CT images of the head may be acquired along a Z-axis direction of CT scan when the proportion threshold of soft tissues is preset. This is because that the proportion of soft tissues of the head along the Z-axis direction has a more obvious change.
In Sub-step 402, a proportion of soft tissues of each of the plurality of CT images is computed.
For each of the plurality of CT images acquired in Sub-step 401, its proportion of soft tissues may be computed by the afore-mentioned method for computing the proportion of soft tissues within the region of interest, and then a diagram illustrating a change in the proportion of soft tissues in the Z-axis direction as shown in Fig. 11A may be obtained.
In Sub-step 403, a noise value of each of said plurality of CT images is computed.
For each of the plurality of CT images acquired in Sub-step 401, its noise value may be computed by the method for estimating the high-frequency information thereon, and then a diagram illustrating a change in the noise value in the Z-axis direction as shown in Fig. 11B may be obtained.
In Sub-step 404, a proportion of soft tissues of a CT image whose noise value is greater than the preset noise value is used as the proportion threshold of soft tissues.

Usually, the noise produced when CT scan is performed on a certain scanned object will be in a relatively stable, predictable range. For example, a normal noise value for scanning the head is usually within 3 dB. Hence, a proportion of soft tissues of a CT image whose noise value is just beyond 3 dB may be used as the proportion threshold of soft tissues.
Specifically for Fig. 11A and Fig. 11B, the proportion of soft tissues in Fig. 11A, of the CT image on the left dashed line in Fig. 11B, or the proportion of soft tissues in Fig. 11A, of the CT image on the right dashed line in Fig. 11B may be used as the proportion threshold of soft tissues.
In Step 103, when the proportion is less than the preset proportion threshold of soft tissues, noise is computed according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image.
In one embodiment of the present invention, the plurality of scanning parameters may include a voltage (kV) used for producing an X-ray by a CT machine, a product value (mAs) of a current intensity used for producing an X-ray by the CT machine and a duration of a current, a scan mode, a helical pitch. Among the above parameters, the scan mode may include a full scan mode and a plus scan mode.
In one embodiment of the present invention, the noise model may contain a plurality of lists of noise ratios, each of which may contain a ratio of a noise value of a historical scanning image to a noise value of a reference image for the same scanned object.
The reference image may be determined by people. For example, an image obtained by performing axial scan with an X-ray whose intensity is 120 kV on a water mold whose radius is 20 centimeters may be used as the reference image. There may be a plurality of such reference images because other scanning parameters may also be different and be combined correspondingly, e.g., between different scan modes, different mAs.

The historical scanning image may be a plurality of images obtained by previously employing some common combinations of the above scanning parameters for the same scanned object (e.g., the head).
The plurality of lists of noise ratios in the noise model may be set with respect to the scanning parameters. For example, a first list of noise ratios may be set with respective to mAs, which may contain noise ratios of images obtained by performing historical scanning on the head and scanning the water mold of 20 centimeters respectively when kV is fixed as an intensity (e.g., 120 kV) used by the reference image, under the conditions of various common combinations of the mAs, the helical pitch and the scan modes. For another example, a second list of noise ratios may be set with respective to the scan modes, which may contain noise ratios between images obtained by performing historical scanning on the head in the case that the other scanning parameters are the same while only the scan modes are different, and may also contain noise ratios between images obtained by scanning the water mold of 20 centimeters in the case that the other scanning parameters are the same while only the scan modes are different. For yet another example, a third list of noise ratios may be set with respect to kV, which may contain noise ratios of images obtained by performing historical scanning on the head and scanning the water mold of 20 centimeters in the case that the other scanning parameters are the same while only the kV values are different. Since the noise values and their ratios in the above lists are all obtained in the case of axial scan, if helical scan is employed for the current CT image, mAs used by the helical scan may also be converted equivalent to mAs of the axial scan according to an mAs conversion relationship between the helical scan and the axial scan (mAs of the axial scan is equal to mAs of the helical scan divided by the helical pitch).
In one embodiment of the present invention, with reference to Fig. 5, Step 103 may further include the following Sub-steps 501-502.

In Sub-step 501, a plurality of corresponding noise ratios are acquired in the list of ratios according to the plurality of scanning parameters employed when the current CT image is generated.
For example, if the current CT image is generated by an axial scan in which the scanning parameters are 180 mAs, 100 kV, a plus scan mode, a helical pitch of 0.625, then a first noise ratio of axial scan on the head and axial scan on the water mold of 20 centimeters in which the scanning parameters are 180 mAs, a plus scan mode, a helical pitch of 0.625 may be firstly found out in the above first list of noise ratios. Then a third noise ratio of axial scan on the head and axial scan on the water mold of 20 centimeters in which the scanning parameter is 100 kV may be found out in the above third list of noise ratios.
In Sub-step 502, noise of the current CT image is obtained by multiplying a product of a plurality of noise ratios with a noise value of the reference image.
Since the noise values of the axial scan on the water mold of 20 centimeters in which the scanning parameters are 180 mAs, 120 kV, a plus scan mode, a helical pitch of 0.625 may be data obtained in advance by experiments, in Sub-step 502, a noise value of the current CT image may just be estimated by multiplying the first noise ratio obtained in Sub-step 501 with the third noise ratio obtained in Sub-step 501 and further with the noise value of the reference image obtained in Sub-step 501.
So far, a method for computing noise on a CT image according to an embodiment of the present invention has been described. As seen from comparison between Fig. 7A and Fig. 7B, after the noise is computed and removed by the existing method, the image within the gray circle is excessively smoothed and detail information is lost, whereas as seen from comparison between Fig. 7B and Fig. 7C, after the noise is computed and removed by the method of the present invention, the image within the gray circle is merely appropriately smoothed so that considerable detail information is retained while the noise is removed. Therefore, the method of the present invention is capable

of more accurately computing the noise on the CT image in the case that the change in the proportion of soft tissues of the scanned object along the direction of the CT scan is larger, avoiding the problem that the image is excessively smoothed in the subsequent process of de-noising. Moreover, the method of the present invention is also capable of adaptively selecting an region of interest so as to avoid improper selection of the region of interest due to the deviation of the position of the scanned object.
Similar to said method, the present invention also provides a corresponding apparatus.
Fig. 6 is a schematic block diagram illustrating one embodiment of an apparatus for computing noise on a CT image according to the present invention.
As shown in Fig. 6, the apparatus 600 may includes: a region-of-interest selecting module 601 for selecting a region of interest on a current CT image; a proportion of soft tissues within a region of interest computing module 602 for computing a proportion of soft tissues within the region of interest; and a module 603 that computes noise by a noise model, for when the proportion is less than a preset proportion threshold of soft tissues, computing the noise according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image.
In one embodiment of the present invention, the region-of-interest selecting module 602 may further include: a range determining module for determining a range of the current CT image; a central point selecting module for using a central point of the range as a central point of the region of interest.
In one embodiment of the present invention, the range determining module may further include: a pixel value accumulating module for accumulating pixel values of the current CT image along a coordinate axis direction; and a module for determining the range according to an accumulation result.

In one embodiment of the present invention, the apparatus 600 may further include: a proportion threshold of soft tissues setting module for determining a proportion threshold of soft tissues according to a relationship between a change in a proportion of soft tissues of a scanned object along a certain scanning direction and a change in a corresponding noise value.
In one embodiment of the present invention, the proportion threshold of soft tissues setting module may further include: a multiple-CT-image acquiring module for, for the same scanned object, acquiring a plurality of CT images along a direction in which a proportion of soft tissues of the scanned object changes; a proportion of soft tissues of each of CT images computing module for computing a proportion of soft tissues of each of the plurality of CT images; a noise value computing module for computing a noise value of each of the plurality of CT images; and a module for using the proportion of soft tissues of the CT image whose noise value is greater than a preset noise value as the proportion threshold of soft tissues.
In one embodiment of the present invention, the plurality of scanning parameters may include a voltage used for producing an X-ray by a CT machine, a product value of a current intensity used for producing an X-ray by the CT machine and a duration of a current, a scan mode, a helical pitch.
In one embodiment of the present invention, the noise model may contain a list of noise ratios that contains a ratio of a noise value of a historical scanning image to a noise value of a reference image for the same scanned object.
In one embodiment of the present invention, the module 603 that computes noise by a noise model may further include: a noise ratio acquiring module for acquiring a plurality of corresponding noise ratios in a list of ratios according to the plurality of scanning parameters employed when the current CT image is generated; and a module for obtaining noise of the current CT image by multiplying a product of the plurality of noise ratios with the noise value of the reference image.

So far, an apparatus for computing noise on a CT image according to an embodiment of the present invention has been described. As seen from comparison between Fig. 7A and Fig. 7B, after the noise is computed and removed by the prior art, the image within the gray circle is excessively smoothed and detail information is lost, whereas as seen from comparison between Fig. 7B and Fig. 7C, after the noise is computed and removed by the apparatus of the present invention, the image within the gray circle is merely appropriately smoothed so that considerable detail information is retained while the noise is removed. Therefore, the apparatus of the present invention is capable of more accurately computing the noise on the CT image in the case that the change in the proportion of soft tissues of the scanned object along the direction of the CT scan is larger, avoiding the problem that the image is excessively smoothed in the subsequent process of de-noising. Moreover, the apparatus of the present invention is also capable of adaptively selecting a region of interest so as to avoid improper selection of the region of interest due to the deviation of the position of the scanned object.
The above descriptions are merely embodiments of the invention and are not intended to restrict the scope of the invention. All kinds of variations and modifications could be made to the present invention to those skilled in the art. Any modifications, alternatives and improvements made within the spirit and principles of the present invention shall fall within the scope of the appended claims.

We claim:
1. A method for computing noise on a CT image, comprising:
selecting a region of interest on a current CT image;
computing a proportion of soft tissues within said region of interest; and
when said proportion is less than a preset proportion threshold of soft tissues, computing said noise according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image.
2. The method as claimed in Claim 1, wherein said step of selecting a region
of interest on a current CT image further comprises:
determining a range of the current CT image; and
using a central point of said range as a central point of said region of interest.
3. The method as claimed in Claim 2, wherein said step of determining a
range of the current CT image further comprises:
accumulating pixel values of the current CT image along a coordinate axis direction; and
determining said range according to an accumulation result.
4. The method as claimed in Claim 1, further comprising:
determining said proportion threshold of soft tissues according to a relationship between a change in a proportion of soft tissues of a scanned object along a certain scanning direction and a change in a corresponding noise value.

5. The method as claimed in Claim 4, wherein said step of determining said
proportion threshold of soft tissues according to a relationship between a change
in a proportion of soft tissues of a scanned object along a certain scanning
direction and a change in a corresponding noise value further comprises:
for the same scanned object, acquiring a plurality of CT images along a direction in which a proportion of soft tissues of the scanned object changes;
computing a proportion of soft tissues of each of said plurality of CT images;
computing a noise value of each of said plurality of CT images; and
using a proportion of soft tissues of a CT image whose noise value is greater than a preset noise value as said proportion threshold of soft tissues.
6. The method as claimed in Claim 1, wherein said plurality of scanning parameters comprise a voltage used for producing an X-ray by a CT machine, a product value of a current intensity used for producing an X-ray by said CT machine and a duration of a current, a scan mode, a helical pitch.
7. The method as claimed in Claim 6, wherein said noise model contains a list of noise ratios that contains a ratio of a noise value of a historical scanning image to a noise value of a reference image for the same scanned object.
8. The method as claimed in Claim 7, wherein said step of computing noise according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image when said proportion is less than a preset proportion threshold of soft tissues further comprises:
acquiring a plurality of corresponding noise ratios in said list of ratios according to the plurality of scanning parameters employed when the current CT image is generated; and

obtaining noise of the current CT image by multiplying a product of said plurality of noise ratios with the noise value of the reference image.
9. An apparatus for computing noise on a CT image, comprising:
a region-of-interest selecting module for selecting a region of interest on a current CT image;
a proportion of soft tissues within a region of interest computing module for computing a proportion of soft tissues within said region of interest; and
a module that computes noise by a noise model, for when said proportion is less than a preset proportion threshold of soft tissues, computing said noise according to a preset noise model and a plurality of scanning parameters employed for generating the current CT image.
10. The apparatus as claimed in Claim 9, wherein said region-of-interest
selecting module further comprises:
a range determining module for determining a range of the current CT image;
a central point selecting module for using a central point of said range as a central point of said region of interest.
11. The apparatus as claimed in Claim 10, wherein said range determining
module further comprises:
a pixel value accumulating module for accumulating pixel values of the current CT image along a coordinate axis direction; and
a module for determining said range according to an accumulation result.
12. The apparatus as claimed in Claim 9, further comprising:

a proportion threshold of soft tissues setting module for determining the proportion threshold of soft tissues according to a relationship between a change in a proportion of soft tissues of a scanned object along a certain scanning direction and a change in a corresponding noise value.
13. The apparatus as claimed in Claim 12, wherein said proportion threshold
of soft tissues setting module further comprises:
a multiple-CT-image acquiring module for, for the same scanned object, acquiring a plurality of CT images along a direction in which a proportion of soft tissues of the scanned object changes;
a proportion of soft tissues of each of CT images computing module for computing a proportion of soft tissues of each of said plurality of CT images;
a noise value computing module for computing a noise value of each of said plurality of CT images; and
a module for using a proportion of soft tissues of a CT image whose noise value is greater than a preset noise value as said proportion threshold of soft tissues.
14. The apparatus as claimed in Claim 9, wherein said plurality of scanning parameters comprise a voltage used for producing an X-ray by a CT machine, a product value of a current intensity used for producing an X-ray by said CT machine and a duration of a current, a scan mode, a helical pitch.
15. The apparatus as claimed in Claim 14, wherein said noise model contains a list of noise ratios that contains a ratio of a noise value of a historical scanning image to a noise value of a reference image for the same scanned object.
16. The apparatus as claimed in Claim 15, wherein said module that computes noise by a noise model further comprises:

a noise ratio acquiring module for acquiring a plurality of corresponding noise ratios in said list of ratios according to the plurality of scanning parameters employed when the current CT image is generated; and
a module for obtaining noise of the current CT image by multiplying a product of said plurality of noise ratios with the noise value of the reference image.