KR20130045544A - Method and apparatus for analyzing magnetic resonance imaging, and recording medium for executing the method - Google Patents

Method and apparatus for analyzing magnetic resonance imaging, and recording medium for executing the method Download PDF

Info

Publication number
KR20130045544A
KR20130045544A KR1020110109808A KR20110109808A KR20130045544A KR 20130045544 A KR20130045544 A KR 20130045544A KR 1020110109808 A KR1020110109808 A KR 1020110109808A KR 20110109808 A KR20110109808 A KR 20110109808A KR 20130045544 A KR20130045544 A KR 20130045544A
Authority
KR
South Korea
Prior art keywords
tumor
candidate
magnetic resonance
image
tumor candidate
Prior art date
Application number
KR1020110109808A
Other languages
Korean (ko)
Other versions
KR101284388B1 (en
Inventor
김동현
박재석
김응엽
양승욱
Original Assignee
연세대학교 산학협력단
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 연세대학교 산학협력단 filed Critical 연세대학교 산학협력단
Priority to KR1020110109808A priority Critical patent/KR101284388B1/en
Publication of KR20130045544A publication Critical patent/KR20130045544A/en
Application granted granted Critical
Publication of KR101284388B1 publication Critical patent/KR101284388B1/en

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06KRECOGNITION OF DATA; PRESENTATION OF DATA; RECORD CARRIERS; HANDLING RECORD CARRIERS
    • G06K9/00Methods or arrangements for reading or recognising printed or written characters or for recognising patterns, e.g. fingerprints
    • G06K9/62Methods or arrangements for recognition using electronic means
    • G06K9/6201Matching; Proximity measures
    • G06K9/6202Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/30Determination of transform parameters for the alignment of images, i.e. image registration
    • G06T7/32Determination of transform parameters for the alignment of images, i.e. image registration using correlation-based methods
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/10Image acquisition modality
    • G06T2207/10072Tomographic images
    • G06T2207/10088Magnetic resonance imaging [MRI]
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30004Biomedical image processing
    • G06T2207/30016Brain

Abstract

The present invention relates to a method for analyzing a magnetic resonance image using a magnetic resonance black blood flow imaging technique.
Magnetic resonance image analysis method according to an embodiment of the present invention comprises the steps of obtaining an image by applying a magnetic resonance black blood flow imaging technique; Extracting an object to be inspected from the obtained image; Extracting a tumor candidate from the extracted object by applying a template collating technique; Extracting features of the extracted tumor candidate; And determining whether the tumor candidate is a tumor tissue by using the extracted feature.

Description

METHOD AND APPARATUS FOR ANALYZING MAGNETIC RESONANCE IMAGING, AND RECORDING MEDIUM FOR EXECUTING THE METHOD}

The present invention relates to a method for analyzing a magnetic resonance image, and more particularly, to a method for analyzing a magnetic resonance image using a magnetic resonance black blood flow imaging technique.

The number of people dying from cancer worldwide will increase from 6 million per year to 10 million by 2020, and the world health organization forecasts that 10 million people with cancer will increase to 15 million by 2020. Doing. In addition, more than 12% of the world's deaths are cancer patients, and in developed countries, cancer is the second leading cause of death after heart disease.

Recently, the use of magnetic resonance imaging for the diagnosis of cancer is increasing.

In general, in order to photograph metastatic tumors by magnetic resonance imaging, it is necessary to inject a contrast agent through a blood vessel to reduce T1 time of tumor tissue. In this case, the tumor tissue has a higher signal intensity than the other human tissues in the image and appears bright. However, the blood vessels included in the imaging volume also have high signal intensity due to the contrast medium administered, and it is difficult to read the image because it is similar to the tumor tissue in the final image depending on the position and slice orientation of the vessel. .

In addition, there is a lack of trained specialists who can read and diagnose the MR image compared to the trend of increasing the number of MR images taken every year as part of the health examination.

To this end, research on the technology that can assist the reading task is required.

SUMMARY OF THE INVENTION An object of the present invention is to provide a method, an apparatus for analyzing a magnetic resonance image using a magnetic resonance black blood flow imaging technique, and a recording medium on which a program for executing the method is recorded.

It is an object of the present invention to detect a tumor candidate group using magnetic resonance black blood flow imaging technique and template control technique, and to perform a magnetic resonance imaging analysis method, apparatus, and method for determining whether the tumor candidate group is tumor tissue. There is a recording medium on which the program for recording is recorded.

According to an embodiment of the present invention to achieve the above object, obtaining an image by applying a magnetic resonance black blood flow imaging technique; Extracting an object to be inspected from the obtained image; Extracting a tumor candidate from the extracted object by applying a template collating technique; Extracting features of the extracted tumor candidate; And determining whether the tumor candidate is tumor tissue by using the extracted features.

According to an embodiment of the present invention to achieve the above object, an image acquisition unit for obtaining an image by applying a magnetic resonance black blood flow imaging technique; A preprocessor extracting an object to be inspected from the acquired image; A tumor candidate extracting unit extracting a tumor candidate from the extracted object by applying a template collating technique; A feature extraction unit for extracting features of the extracted tumor candidates; A tumor determination unit that determines whether the tumor candidate is tumor tissue using the extracted features; And a controller configured to control the image acquirer, the preprocessor, the tumor candidate extractor, the feature extractor, and the tumor determiner.

According to an embodiment of the present invention to achieve the above object, obtaining an image by applying a magnetic resonance black blood flow imaging technique; Extracting an object to be inspected from the obtained image; Extracting a tumor candidate from the extracted object by applying a template collating technique; Extracting features of the extracted tumor candidate; And a recording medium having recorded thereon a program for executing the step of determining whether the tumor candidate is a tumor tissue using the extracted feature.

According to one embodiment of the present invention, by applying the magnetic resonance black blood flow imaging technique and template contrast technique together, it is possible to reduce the number of false candidates, to detect more accurate tumor candidates when detecting a tumor target candidate.

1 is a block diagram of a magnetic resonance image analysis apparatus according to an embodiment of the present invention.
2 is a flowchart illustrating a method for analyzing magnetic resonance images according to an embodiment of the present invention.
FIG. 3 is a diagram illustrating a process of extracting an object to be inspected from among magnetic resonance image analysis methods according to an exemplary embodiment of the present invention.
4 is a view for comparing the conventional magnetic resonance image and the image to which the magnetic resonance black blood flow imaging technique is applied.
5 is a diagram illustrating a process of extracting a tumor candidate group through an image.

Hereinafter, a magnetic resonance image analysis method and analysis apparatus according to an embodiment of the present invention will be described with reference to the accompanying drawings. According to an embodiment of the present invention, magnetic resonance image analysis may determine whether cancer tissue (or tumor tissue) is present in a specific object.

1 is a block diagram of an apparatus for analyzing magnetic resonance imaging (MRI) according to an embodiment of the present invention.

The illustrated MRI apparatus 100 includes an image acquirer 110, a preprocessor 120, a tumor candidate extractor 130, a feature extractor 140, a tumor determiner 150, and a controller 160. ) May be included.

The image acquirer 110 may acquire an image by applying a magnetic resonance black blood flow imaging technique.

In the present specification, magnetic resonance black blood flow imaging technique refers to a magnetic resonance imaging technique that prevents blood signals from appearing in a final image by using mobility of blood flowing out or flowing from blood vessels located in an image or an image volume (3D image). Say. The magnetic resonance black blood flow imaging technique may be performed using a pulse train. Pulse trains are pulses that have similar properties and follow regularly. For example, a group of pulses that occur sequentially in time at a point in a circuit.

The preprocessing unit 120 performs a preprocessing process on an image obtained by applying the magnetic resonance black blood flow imaging technique in the image acquisition unit 100. The preprocessing process is performed before the tumor candidate is extracted, and may include segmentation of an image, noise removal, and normalization.

According to an embodiment of the present invention, the preprocessor 120 may correct the signal of the image obtained by applying the magnetic resonance black blood flow imaging technique and extract the object to be examined. The object to be tested refers to an object to which cancer tissue (or tumor tissue) is to be determined. For example, the object under test may include a brain. Therefore, according to one embodiment of the present invention, it is possible to determine whether tumor tissue is present in the brain.

The tumor candidate extracting unit 130 may extract a tumor candidate by applying a template collation technique to the test target object extracted by the preprocessing unit 120.

In the present specification, a template collation technique is a method of extracting a correlation using a collation with an image with a specific template. The specific template may be prepared in advance by predicting the growth form of the tumor tissue. For example, because metastatic brain tumors grow spherically, the particular template can be made spherical. In addition, since the tumor size may vary, a plurality of specific templates may be manufactured and used.

The feature extractor 140 may extract various features of the tumor candidate. The feature extractor 140 may parameterize various information about each tumor candidate to generate a table or store a parameterized table. In this case, the feature extractor 140 may extract the feature of the tumor candidate by considering only the image signal intensity of each tumor candidate.

The tumor determination unit 150 may determine whether the corresponding tumor candidate is tumor tissue or normal tissue by using the features of the tumor candidate extracted by the feature extraction unit 140. For example, machine learning methods can determine whether a tumor candidate is tumor tissue. Machine learning is a field of artificial intelligence that develops algorithms and technologies that enable computers to learn. An example is an artificial neural network. An artificial neural network (ANN) is a mathematical model that aims to represent some of the features of brain function in computer simulations.

The controller 160 may organically control the overall functions performed by the image acquirer 110, the preprocessor 120, the tumor candidate extractor 130, the feature extractor 140, and the tumor determiner 150. Can be.

In addition, when determining that the tumor candidate extracted by the tumor determination unit 150 is a tumor tissue, the controller 160 may display the tumor candidate on the original image to identify the tumor candidate. The original image refers to an MRI photographed image to which the magnetic resonance black blood flow imaging technique is not applied.

2 is a flowchart illustrating a method for analyzing magnetic resonance images according to an embodiment of the present invention.

The image acquirer 110 may acquire an image by applying a magnetic resonance black blood flow imaging technique (S210). The acquired image is no signal of blood. By applying the magnetic resonance black blood flow imaging technique as described above, false positives can be reduced. In other words, in the brain image obtained by the imaging technique without the magnetic resonance black blood flow imaging technique, more false positives are generated due to signals generated in blood vessels, cerebrospinal fluid, etc., thereby lowering the accuracy of diagnosis. This problem can be solved by obtaining an image obtained by removing a signal from a blood vessel using a black blood imaging technique related to an embodiment of the present invention.

The preprocessing unit 120 may preprocess the image acquired through applying the magnetic resonance black blood flow imaging technique (S220). The preprocessing process is performed before the tumor candidate is extracted, and may include segmentation of an image, noise removal, and normalization.

The preprocessor 120 may correct the signal of the image obtained by applying the magnetic resonance black blood flow imaging technique.

The magnetic resonance black blood flow imaging technique may use a non-selective excitation pulse. This should ideally excite all spindles in the image volume to have a constant flip angle.

However, the pulses actually used are incomplete due to physical limitations in the design, resulting in non-uniform flip angles of the spindles in the image volume, resulting in non-uniform signal strength in the image.

On the other hand, since the signal strength is used as an important variable for the object to be inspected or the region of interest (ROI), it is necessary to correct the nonuniform signal strength to perform the correct diagnosis.

For example, the signal is transformed into a frequency domain (or k-space) through a three-dimensional Fourier transform for signal correction, and then, from energy information gathered in a low frequency region to the entire image region. The signal intensity distribution can be obtained and divided by the original image. Through the above method it is possible to correct the non-uniformity of the signal.

In addition, the preprocessor 120 may extract only the object to be inspected. This makes it possible to exclude signals due to non-retrievable objects that are not completely removed despite the magnetic resonance black blood flow imaging technique. By eliminating the signals due to the non-test object, a more accurate diagnosis can be made.

Hereinafter, the case where the searched object is the brain will be described with an example.

In this case, the preprocessor 120 may perform a brain extraction process. Through this, even though magnetic resonance black blood flow imaging technique is applied, signals generated from lateral ventricle, 3rd ventricle, or cerebrospinal fluid (CSF) that are not completely removed can be excluded.

3 is a diagram illustrating a process of extracting a brain.

As shown in the drawing, signals generated in the lateral ventricle, the third ventricle, or the cerebrospinal fluid (CSF), which have not been completely removed despite the application of magnetic resonance black blood flow imaging techniques, can be removed. . In this manner, the scope of the search object can be reduced.

Next, the tumor candidate extracting unit 130 may extract the tumor candidate using a template collation technique (S230). An algorithm for generating a tumor candidate may be applied to the 3D brain image extracted through the brain extraction process.

Since metastatic brain tumors generally grow spherically, a template having a three-dimensional spherical shape can be used as a template for tumor candidate extraction.

For example, the tumor candidate extracting unit 130 may obtain a normalized cross-correlation coefficient (NCCC) by comparing the template having the shape of a three-dimensional sphere with the brain image. The tumor candidate extractor 130 may create an NCCC map in which pixels having a high cross correlation coefficient are bright and low dots are dark.

However, since the correlation coefficient may vary depending on the size of the tumor, a result of producing a plurality of templates may be obtained. Candidates that repeatedly exhibit high NCCC values for a plurality of templates may be regarded as tumors mapped to the size of the template having the highest NCCC value among them.

The template control technique could be used in one embodiment of the present invention because the tumor candidate group (or candidate group to be examined) was significantly reduced by the magnetic resonance black blood flow imaging technique. This can also be explained through FIG. 4.

4 is a view for comparing the conventional magnetic resonance image and the image to which the magnetic resonance black blood flow imaging technique is applied. Images a and c are images taken by a magnetic resonance imaging technique to which the magnetic resonance black blood flow imaging technique is not applied, and images b and d are images to which the magnetic resonance black blood flow imaging technique is applied.

As shown, the magnetic resonance imaging technique without the magnetic resonance black blood imaging technique generates high signal tissues (vessels, cerebrospinal fluid, etc.). . In addition, since the false positive candidate groups generated in the tumor tissue determination step to be described later to be excluded all depend on the tumor tissue determination system.

In addition, the boundary between the brain metastasis tumor and the surrounding brain tissue is not clear in the image obtained through the magnetic resonance imaging technique without the magnetic resonance black blood flow imaging technique. Because of this, the growth of tumors in spherical shape is difficult to confirm without going through biopsy (biopsy).

The tumor candidate extractor 130 leaves only representative voxels having an NCCC value greater than or equal to a predetermined threshold value among the candidate voxels shown in the NCCC map and excludes the rest from the candidates. Only the remaining representative voxels can be extracted as tumor candidates.

The voxel refers to an extension of the concept of a pixel to a three-dimensional space with a volume pixel. This refers to a sample of data of the actual volume obtained precisely.

On the other hand, according to one embodiment of the present invention, the tumor candidate extracting unit 130 may group the tumor candidates according to a specific criterion when the extracted tumor candidates are plural.

For example, when the first tumor candidate and the second tumor candidate are located within a predetermined distance, the first tumor candidate and the second tumor candidate may be grouped into one group. The predetermined distance may be determined in consideration of the size of a specific template used for tumor candidate extraction. For example, a template size of 10 may group tumor candidates that are relatively further apart than a template size of 8 into a group.

The tumor candidate grouping process is a process of grouping voxels suspected of a tumor group into one candidate.

If a plurality of templates are used to apply the template collation technique, the voxels may be grouped based on the size of the template having the highest NCCC value.

When the tumor candidate extractor 130 extracts the tumor candidate through the method, the feature extractor 140 may extract various information about the extracted tumor candidate (S240). In this case, the feature extractor 140 may extract the feature of the tumor candidate in consideration of only the intensity of the signal of the tumor candidate.

Since the number of false candidates is reduced through the application of magnetic resonance black blood flow imaging techniques, it may be possible to determine whether the tumor tissue is extracted by using the extracted features even if the features are extracted only by considering the signal intensity of the tumor candidates.

The feature extractor 140 may extract various features of the tumor candidate. The feature extractor 140 may parameterize various information about each tumor candidate to generate a table or store a parameterized table.

As an example, as shown in Table 1, various information about tumor candidates may be parameterized.

Number Description One Information used in Neural Network training, the 'correct' answer for that candidate (0 or 1). 2 Data ID 3 X-axis position of representative voxels 4 Z-axis position of representative voxels 5 Y-axis position of representative voxel 6 NCCC value 7 Signal intensity of representative voxels 8 Normalized signal intensity of representative voxels 9 Average signal intensity of the candidate volume 10 Standard deviation of the signal intensity of the candidate volume 11 Number of voxels with higher than average signal intensity among candidate volumes 12 The number of voxels that have a value higher than the average signal intensity among the volumes slightly larger than the candidate volume. 13 Average value of voxels with lower than average signal intensity among candidate volumes 14 Standard Deviation of Voxel with Lower Value than Average Signal Intensity among Candidate Volumes 15 Average value of the voxel with a value lower than the average signal intensity among the volumes slightly larger than the candidate volume 16 Standard deviation of voxels with a value lower than the average signal intensity among the volumes slightly larger than the candidate volume 17 Number of voxels with zero signal intensity among candidate volumes 18 Number of voxels with a signal intensity of 0 among those slightly larger than the candidate volume 19 X-axis / y-axis ratio of candidate volume 20 Y-axis / z-axis ratio of candidate volume 21 Z-axis / x-axis ratio of candidate volumes

As described above, when the feature for the tumor candidate is extracted, the tumor determination unit 150 may determine whether the tumor candidate is tumor tissue or normal tissue using the extracted feature (S250).

If it is determined that the tumor tissue, the controller 160 may display on the original image to identify the tumor candidate (S260).

Meanwhile, in the process of extracting a tumor candidate, a method of displaying the extracted tumor candidate on an original image may be described through an image.

5 is a diagram illustrating a process of extracting a tumor candidate group through an image.

Figure 5 (a) shows the brain image extracted through the preprocessing process (S220). 5 (b) shows an NCCC map generated using an NCCC correlation coefficient through application of a template collation technique. In addition, 5 (c) shows an image in which a plurality of tumor candidates are formed into one group through a grouping process in the tumor candidate detection step (S220). FIG. 5 (d) shows an example in which the tumor candidate 510 determined as tumor tissue in the extracted tumor candidate (grouped candidate) is displayed to be identified in the original image.

As described above, the magnetic resonance image analysis method and apparatus according to an embodiment of the present invention can significantly reduce false positives by using magnetic resonance black blood imaging, thereby increasing accuracy in extracting tumor candidates. .

The above-described magnetic resonance image analysis method may be implemented in the form of program instructions that can be executed by various computer means and recorded in a computer-readable recording medium. In this case, the computer-readable recording medium may include program instructions, data files, data structures, and the like, alone or in combination. On the other hand, the program instructions recorded on the recording medium may be those specially designed and configured for the present invention or may be available to those skilled in the art of computer software.

The computer-readable recording medium includes a magnetic recording medium such as a magnetic medium such as a hard disk, a floppy disk and a magnetic tape, an optical medium such as a CD-ROM and a DVD, a magnetic disk such as a floppy disk, A magneto-optical media, and a hardware device specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like.

The recording medium may be a transmission medium such as an optical or metal wire, a waveguide, or the like including a carrier wave for transmitting a signal specifying a program command, a data structure, or the like.

In addition, program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the present invention, and vice versa.

The magnetic resonance image analysis method and apparatus described above are not limited to the configuration and method of the above-described embodiments, but the embodiments may be modified in whole or in part to enable various modifications. It may alternatively be configured in combination.

100: magnetic resonance image analysis device
110:
120:
130: tumor candidate extracting unit
140: feature extraction unit
150: tumor determination unit
160:

Claims (23)

  1. Obtaining an image by applying a magnetic resonance black blood flow imaging technique;
    Extracting an object to be inspected from the obtained image;
    Extracting a tumor candidate from the extracted object by applying a template collating technique;
    Extracting features of the extracted tumor candidate; And
    Magnetic resonance image analysis method comprising the step of determining whether the tumor candidate is a tumor tissue using the extracted features.
  2. According to claim 1, The magnetic resonance image analysis method
    Magnetic resonance image analysis method further comprises the step of correcting the signal of the obtained image.
  3. The method of claim 2, wherein the correcting the signal of the obtained image
    Magnetic resonance image analysis method, characterized in that for using the signal intensity distribution for the entire signal image obtained.
  4. The method of claim 1, wherein the object to be inspected
    Magnetic resonance imaging analysis method comprising the brain.
  5. The method of claim 1, wherein the tumor candidate extraction step
    Obtaining a correlation using a contrast between a specific template prepared in advance and an image;
    Adjusting the brightness of the image using the obtained correlation; And
    And extracting only voxels having the obtained correlation greater than or equal to a predetermined threshold as a tumor candidate.
  6. The method of claim 5, wherein the tumor candidate extraction step
    When there are a plurality of extracted tumor candidates and the first tumor candidate and the second candidate candidate are located within a predetermined distance,
    Magnetic resonance image analysis method further comprises the step of grouping the first tumor candidate and the second tumor candidate in one group.
  7. The method of claim 6, wherein the predetermined distance is
    Magnetic resonance image analysis method, characterized in that determined in consideration of the size of the specific template.
  8. According to claim 1, The magnetic resonance image analysis method
    And if the extracted tumor candidate is determined to be a tumor tissue, further comprising displaying the extracted tumor candidate on the original MR image so as to be identified.
  9. The method of claim 1, wherein the feature extraction step of the tumor candidate is
    Magnetic resonance image analysis method, characterized in that performed in consideration of only the signal strength for the tumor candidate.
  10. An image acquisition unit which acquires an image by applying a magnetic resonance black blood flow imaging technique;
    A preprocessor extracting an object to be inspected from the acquired image;
    A tumor candidate extracting unit extracting a tumor candidate from the extracted object by applying a template collating technique;
    A feature extraction unit for extracting features of the extracted tumor candidates;
    A tumor determination unit that determines whether the tumor candidate is tumor tissue using the extracted features; And
    And a controller for controlling the image acquisition unit, the preprocessor, the tumor candidate extractor, the feature extractor, and the tumor determiner.
  11. The method of claim 10, wherein the pretreatment unit
    Magnetic resonance image analysis device, characterized in that for correcting the signal of the obtained image.
  12. The method of claim 11, wherein the signal correction performed by the preprocessor is performed.
    Magnetic resonance image analysis apparatus, characterized in that performed using the signal intensity distribution for the entire signal image obtained.
  13. The method of claim 10, wherein the object to be inspected
    Magnetic resonance imaging analysis device comprising a brain.
  14. The method of claim 10, wherein the tumor candidate extracting unit
    Obtaining a correlation using a contrast between a specific template prepared in advance and an image, adjusting the brightness of the image using the obtained correlation, and extracting only voxels having the acquired correlation greater than or equal to a predetermined threshold as a tumor candidate Magnetic resonance image analysis device, characterized in that.
  15. The method of claim 14, wherein the tumor candidate extracting unit
    When there are a plurality of extracted tumor candidates and the first tumor candidate and the second candidate candidate are located within a predetermined distance,
    Magnetic resonance image analysis apparatus, characterized in that for grouping the first tumor candidate and the second tumor candidate in a group.
  16. The method of claim 15, wherein the predetermined distance is
    Magnetic resonance image analysis device, characterized in that determined in consideration of the size of the specific template.
  17. 11. The apparatus of claim 10, wherein the control unit
    When the extracted tumor candidate is determined to be a tumor tissue, the magnetic resonance image analysis apparatus, characterized in that the extracted tumor candidate is displayed on the original MRI to identify.
  18. The method of claim 10, wherein the feature extraction unit
    Magnetic resonance imaging apparatus for extracting the characteristics of the tumor candidate in consideration of the signal strength of the tumor candidate only.
  19. Obtaining an image by applying a magnetic resonance black blood flow imaging technique;
    Extracting an object to be inspected from the obtained image;
    Extracting a tumor candidate from the extracted object by applying a template collating technique;
    Extracting features of the extracted tumor candidate; And
    And a program for executing the step of determining whether the tumor candidate is tumor tissue by using the extracted feature.
  20. 20. The method of claim 19, wherein extracting the tumor candidate is
    Obtaining a correlation using a contrast between a specific template prepared in advance and an image;
    Adjusting the brightness of the image using the obtained correlation; And
    And extracting only voxels having the obtained correlation greater than or equal to a predetermined threshold as a tumor candidate.
  21. The method of claim 20, wherein the tumor candidate extraction step is
    When there are a plurality of extracted tumor candidates and the first tumor candidate and the second candidate candidate are located within a predetermined distance,
    And grouping the first tumor candidate and the second tumor candidate into one group.
  22. The method of claim 21, wherein the predetermined distance is
    The recording medium, characterized in that determined in consideration of the size of the specific template.
  23. 20. The method of claim 19, wherein extracting the feature of the tumor candidate
    The recording medium, characterized in that performed only in consideration of the signal strength for the tumor candidate.
KR1020110109808A 2011-10-26 2011-10-26 Method and apparatus for analyzing magnetic resonance imaging, and recording medium for executing the method KR101284388B1 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
KR1020110109808A KR101284388B1 (en) 2011-10-26 2011-10-26 Method and apparatus for analyzing magnetic resonance imaging, and recording medium for executing the method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
KR1020110109808A KR101284388B1 (en) 2011-10-26 2011-10-26 Method and apparatus for analyzing magnetic resonance imaging, and recording medium for executing the method

Publications (2)

Publication Number Publication Date
KR20130045544A true KR20130045544A (en) 2013-05-06
KR101284388B1 KR101284388B1 (en) 2013-07-09

Family

ID=48657544

Family Applications (1)

Application Number Title Priority Date Filing Date
KR1020110109808A KR101284388B1 (en) 2011-10-26 2011-10-26 Method and apparatus for analyzing magnetic resonance imaging, and recording medium for executing the method

Country Status (1)

Country Link
KR (1) KR101284388B1 (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20150026115A (en) * 2013-08-30 2015-03-11 삼성전자주식회사 Method for controlling stereo convergence and stereo image processor adopting the same
KR101659578B1 (en) * 2015-09-01 2016-09-23 삼성전자주식회사 Method and apparatus for processing magnetic resonance imaging
KR101950815B1 (en) * 2017-08-25 2019-02-21 뉴로핏 주식회사 Patch guide method and program
WO2019039708A1 (en) * 2017-08-25 2019-02-28 Samsung Electronics Co., Ltd. Magnetic resonance imaging apparatus and method of reconstructing mr image by using neural network
KR20190081656A (en) * 2017-12-29 2019-07-09 한국과학기술원 Method and apparatus for correction of a distortion in MR image

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101004342B1 (en) * 2007-10-23 2010-12-27 (주)토탈소프트뱅크 Method for detecting liver region and hepatoma in computer tomography images
KR100903589B1 (en) * 2007-10-26 2009-06-23 한국기초과학지원연구원 Apparatus for analyzing image of Magnetic Resonance Imager and method for the same
KR100968244B1 (en) * 2008-01-28 2010-07-06 고려대학교 산학협력단 System for recognizing a specific subject in image and Method thereof
EP2348990A4 (en) * 2008-10-31 2014-08-20 Method and apparatus using magnetic resonance imaging for cancer identification

Cited By (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR20150026115A (en) * 2013-08-30 2015-03-11 삼성전자주식회사 Method for controlling stereo convergence and stereo image processor adopting the same
KR101659578B1 (en) * 2015-09-01 2016-09-23 삼성전자주식회사 Method and apparatus for processing magnetic resonance imaging
US10198799B2 (en) 2015-09-01 2019-02-05 Samsung Electronics Co., Ltd. Method and apparatus for processing magnetic resonance image
KR101950815B1 (en) * 2017-08-25 2019-02-21 뉴로핏 주식회사 Patch guide method and program
WO2019039708A1 (en) * 2017-08-25 2019-02-28 Samsung Electronics Co., Ltd. Magnetic resonance imaging apparatus and method of reconstructing mr image by using neural network
US10426373B2 (en) 2017-08-25 2019-10-01 Samsung Electronics Co., Ltd. Magnetic resonance imaging apparatus and method of reconstructing MR image by using neural network
KR20190081656A (en) * 2017-12-29 2019-07-09 한국과학기술원 Method and apparatus for correction of a distortion in MR image

Also Published As

Publication number Publication date
KR101284388B1 (en) 2013-07-09

Similar Documents

Publication Publication Date Title
Zhang et al. Detecting anatomical landmarks for fast Alzheimer’s disease diagnosis
EP3046478B1 (en) Image analysis techniques for diagnosing diseases
Rekik et al. Medical image analysis methods in MR/CT-imaged acute-subacute ischemic stroke lesion: Segmentation, prediction and insights into dynamic evolution simulation models. A critical appraisal
Wang et al. Computer aided-diagnosis of prostate cancer on multiparametric MRI: a technical review of current research
Seehaus et al. Histological validation of high-resolution DTI in human post mortem tissue
Trebeschi et al. Deep learning for fully-automated localization and segmentation of rectal cancer on multiparametric MR
CN106056595B (en) Based on the pernicious assistant diagnosis system of depth convolutional neural networks automatic identification Benign Thyroid Nodules
Zhan et al. Deformable segmentation of 3-D ultrasound prostate images using statistical texture matching method
ES2595210T3 (en) Method and apparatus for processing medical images
EP1941453B1 (en) Method of registering images, algorithm for carrying out the method of registering images, a program for registering images using the said algorithm and a method of treating biomedical images to reduce imaging artefacts caused by object movement
US20190183429A1 (en) Deep-learning-based cancer classification using a hierarchical classification framework
EP2374092B1 (en) A method, apparatus, and computer program product for acquiring medical image data
US8761473B2 (en) Tree structure creation apparatus, method and program
Lee et al. Unsupervised connectivity-based thresholding segmentation of midsagittal brain MR images
US20070081712A1 (en) System and method for whole body landmark detection, segmentation and change quantification in digital images
ES2578706T3 (en) Improvement of magnetic resonance imaging in relation to the correction of chemical displacement artifacts and intensity inhomogeneity
Pichat et al. A survey of methods for 3D histology reconstruction
Jannin et al. Validation in medical image processing.
Zheng et al. Machine learning based vesselness measurement for coronary artery segmentation in cardiac CT volumes
Valindria et al. Reverse classification accuracy: predicting segmentation performance in the absence of ground truth
US20060182321A1 (en) Method and apparatus for extracting third ventricle information
US8600135B2 (en) System and method for automatically generating sample points from a series of medical images and identifying a significant region
Diciotti et al. 3-D segmentation algorithm of small lung nodules in spiral CT images
US20030053669A1 (en) Magnetic resonance angiography method and apparatus
Wang et al. Cerebral micro-bleed detection based on the convolution neural network with rank based average pooling

Legal Events

Date Code Title Description
A201 Request for examination
E902 Notification of reason for refusal
E701 Decision to grant or registration of patent right
GRNT Written decision to grant
FPAY Annual fee payment

Payment date: 20160704

Year of fee payment: 4

FPAY Annual fee payment

Payment date: 20170712

Year of fee payment: 5

FPAY Annual fee payment

Payment date: 20180627

Year of fee payment: 6