KR20150047408A - Method of Medical Imaging Apparatus for Image Reconstruction and Medical Imaging Apparatus Thereof - Google Patents

Abstract

Provided are a medical image reconstruction method for providing a medical image having high spatial resolution and high time-based resolution, and low artifact, and a medical imaging apparatus thereof. The method for reconstructing a tomographic image comprises the steps of: obtaining a reference image imaging an object and a reconstruction target image imaging the object; determining at least one reference area corresponding to a unit area in the reconstruction target image in the reference image; and updating data in the unit area based on data included in at least one reference area.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image reconstruction method for a medical imaging apparatus,

The present invention relates to a method for reconstructing a medical image by a medical imaging apparatus and a medical imaging apparatus. More particularly, to a method for reconstructing a medical image such that artifacts are reduced.

Medical images such as computed tomography (CT) images are widely used for diagnosis of diseases. High resolution is required to identify a small tissue in a human tissue. For example, in order to identify small tissues such as coronary artery vessels, cardiac computed tomography (CT) should be applied not only to a sub-millimeter high spatial resolution, High temporal resolution is also required for unaffected images.

An embodiment of the present invention provides a medical image reconstruction method and medical image apparatus for providing a medical image having a high spatial resolution and a high temporal resolution and also having a small artifact.

As a technical means for achieving the above technical object, a method of reconstructing a tomographic image of a medical imaging apparatus includes a step of acquiring a reference image of a target object and a reconstruction target image of the target object, Determining at least one reference area corresponding to the unit area in the reference image in the reference image and updating the data in the unit area based on the data included in the at least one reference area.

According to another embodiment, the at least one reference region has characteristics similar to those of the unit region.

According to another embodiment, a method of reconstructing a tomographic image further comprises acquiring electrocardiography (ECG) data, and the step of acquiring a reference image includes acquiring a reference image from a subject of a diastolic period determined based on electrocardiogram data Acquiring a reference image based on diastolic data that is data on a subject to be reconstructed, wherein acquiring the reconstructed object image acquires a reconstructed object image based on systolic data, which is data on a systolic object determined based on the electrocardiogram data .

According to another embodiment, the step of updating data in the unit area may include minimizing a cost function using a Patch-based Low Rank Regularization, And reconstructs the image.

According to still another embodiment, updating data in the unit area includes generating a matrix based on data included in the unit area and at least one reference area, and generating a matrix based on the matrix, determining an eigenvalue, and updating data in the unit area based on the determined eigenvalue.

According to still another embodiment, the step of updating the data in the unit area may be characterized by using the patch-based low-rank normalization based on the following equation.

는 후버 함수(Huber function)일 수 있다.Where v _p is the generated matrix, 隆_k (v _p ) is the kth largest singular value of v _p ,

May be a Huber function.

In yet a cost function against a accordance with a further embodiment, the step of updating the data in the unit area based on a result of minimizing a cost function with respect to phase, and w _p that minimizes the cost function with respect to w _p, x Minimizing < / RTI >

Also, In accordance with a further embodiment, the tomographic image reconstruction method based on a result of minimizing a cost function with respect to phase, and the w _p that minimizes the cost function with respect to w _p that minimizes the cost function with respect to the x And repeating the steps of FIG.

According to another embodiment, the step of updating data in the unit area may be characterized by minimizing a cost function using a shrinkage operator.

In addition, the computer-readable recording medium according to an embodiment may be characterized in that a program for executing the above-described method is recorded.

The medical imaging apparatus according to an embodiment of the present invention acquires a reference image of a target object and a reconstruction target image of the target object, determines at least one reference region corresponding to a unit region in the reconstruction target image, An image processing unit configured to update data in the unit area based on data included in one reference area, and a display unit configured to display the reconstructed image according to the update result.

According to another embodiment, the at least one reference region has characteristics similar to those of the unit region.

According to another embodiment of the present invention, the medical imaging apparatus further includes an electrocardiogram data acquiring unit for acquiring electrocardiogram data, and the image processing unit may acquire the electrocardiogram data based on the diastolic data, which is data on the object of the diastolic period determined based on the electrocardiogram data, And obtains the reconstructing object image based on the systolic data, which is data on the object of the systolic object determined based on the electrocardiogram data.

According to another embodiment of the present invention, the image processing unit reconstructs the reconstruction target image by minimizing a cost function using a patch-based Low Rank Regularization .

According to still another embodiment, the image processing unit generates a matrix based on the data included in the unit area and the at least one reference area, determines an eigen value based on the matrix, And update data in the unit area based on the determined eigenvalue.

According to still another embodiment, the image processing unit may use the patch-based low-rank normalization based on the following equation.

는 후버 함수(Huber function)일 수 있다.Where v _p is the generated matrix, 隆_k (v _p ) is the kth largest singular value of v _p ,

May be a Huber function.

According to another embodiment of the present invention, the image processing unit may be characterized by minimizing a cost function using a shrinkage operator.

1 is a conceptual diagram showing an example of a medical imaging apparatus.
2 is a block diagram illustrating the structure of a medical imaging apparatus according to an embodiment.
3 is a flow chart illustrating a process for reconstructing a medical image in accordance with one embodiment.
4 is a conceptual diagram showing a configuration in which the medical imaging apparatus acquires image data based on electrocardiographic data.
5 is a conceptual diagram illustrating a reference area corresponding to a unit area according to an embodiment.
6 is a diagram illustrating an example in which a medical imaging device determines a matrix.
FIG. 7 is a schematic diagram illustrating a structure of a medical imaging apparatus according to an embodiment of the present invention. Referring to FIG.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings, which will be readily apparent to those skilled in the art. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly illustrate the present invention, parts not related to the description are omitted, and similar parts are denoted by like reference characters throughout the specification.

For convenience of explanation, the medical warrant apparatus exemplifies a computer tomography apparatus.

The terms used in this specification will be briefly described and the present invention will be described in detail.

While the present invention has been described in connection with what is presently considered to be the most practical and preferred embodiment, it is to be understood that the invention is not limited to the disclosed embodiments. Also, in certain cases, there may be a term selected arbitrarily by the applicant, in which case the meaning thereof will be described in detail in the description of the corresponding invention. Therefore, the term used in the present invention should be defined based on the meaning of the term, not on the name of a simple term, but on the entire contents of the present invention.

When an element is referred to as "including" an element throughout the specification, it is to be understood that the element may include other elements as well, without departing from the spirit or scope of the present invention. Also, as used herein, the term "part " refers to a hardware component such as software, FPGA or ASIC, and" part " However, 'minus' is not limited to software or hardware. The " part " may be configured to be in an addressable storage medium and configured to play back one or more processors. Thus, by way of example, and not limitation, "part (s) " refers to components such as software components, object oriented software components, class components and task components, and processes, Subroutines, segments of program code, drivers, firmware, microcode, circuitry, data, databases, data structures, tables, arrays and variables. The functions provided in the components and "parts " may be combined into a smaller number of components and" parts " or further separated into additional components and "parts ".

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily carry out the present invention. The present invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein. In order to clearly explain the present invention in the drawings, parts not related to the description will be omitted.

As used herein, the term "image" may refer to multi-dimensional data composed of discrete image elements (e.g., pixels in a two-dimensional image and voxels in a three-dimensional image). For example, the image may include a medical image or the like of the object obtained by the CT photographing apparatus.

In the present specification, a "CT (Computed Tomography) image" may mean a composite image of a plurality of x-ray images obtained by photographing an object while rotating around at least one axis of the object.

As used herein, an " object "may include a person or an animal, or a portion of a person or an animal. For example, the subject may include a liver, a heart, a uterus, a brain, a breast, an organ such as the abdomen, or a blood vessel. The "object" may also include a phantom. A phantom is a material that has a volume that is very close to the density of the organism and the effective atomic number, and can include a spheric phantom that has body-like properties.

As used herein, the term "user" may be a doctor, a nurse, a clinical pathologist, a medical imaging specialist, or the like, and may be a technician repairing a medical device.

The CT system can provide a cross-sectional image for a target object, which is advantageous in that the internal structure of the target object (for example, an organ such as a kidney, a lung, etc.) can be expressed without overlapping with a general X-ray imaging apparatus.

The CT system can obtain a relatively accurate sectional image with respect to a target object by, for example, acquiring image data of 2 mm or less in thickness at several tens or hundreds of times per second. Conventionally, there has been a problem that only the horizontal section of the object is expressed, but it has been overcome by the appearance of the following various image reconstruction techniques. Three-dimensional reconstruction imaging techniques include the following techniques.

- SSD (Shade surface display): A technique to represent only voxels having a certain HU value by the initial 3D image technique.

- MIP (maximum intensity projection) / MinIP (minimum intensity projection): A 3D technique that represents only those with the highest or lowest HU value among the voxels that make up the image.

- VR (volume rendering): A technique that can control the color and transparency of voxels constituting an image according to a region of interest.

- Virtual endoscopy: A technique capable of endoscopic observation on reconstructed 3-D images using VR or SSD techniques.

- MPR (multi planar reformation): An image technique that reconstructs into other sectional images. It is possible to reconstruct the free direction in the direction desired by the user.

- Editing: Several techniques for organizing surrounding voxels to more easily observe the region of interest in the VR.

- VOI (voxel of interest): A technique that expresses only the selected region by VR.

According to one embodiment of the present invention, the medical warrant apparatus 100 may be a computed tomography (CT) system. The medical imaging device 100 according to one embodiment may be described with reference to FIG. The CT system 100 according to an embodiment of the present invention may include various types of devices.

1 is a conceptual diagram showing an example of a medical imaging apparatus 100. As shown in FIG. 1, the medical imaging apparatus 100 may include a gantry 102, a table 105, an X-ray generating unit 106, and an X-ray detecting unit 108.

The gantry 102 may include an X-ray generating unit 106 and an X-ray detecting unit 108.

The object 10 can be placed on the table 105. [

The table 105 can be moved in a predetermined direction (e.g., at least one of up, down, left, and right) in the CT photographing process. In addition, the table 105 may be tilted or rotated by a predetermined angle in a predetermined direction.

Also, the gantry 102 may be inclined by a predetermined angle in a predetermined direction.

2 is a block diagram illustrating the structure of a medical imaging apparatus according to an embodiment.

The medical imaging apparatus 100 according to an embodiment of the present invention includes a gantry 102, a table 105, a control unit 118, a storage unit 124, an image processing unit 126, an input unit 128, a display unit 130, and a communication unit 132.

As described above, the object 10 may be located on the table 105. [ The table 105 according to an embodiment of the present invention can be moved in a predetermined direction (e.g., at least one of up, down, left, and right) and the movement can be controlled by the control unit 118. [

The gantry 102 according to an embodiment of the present invention includes a rotating frame 104, an X-ray generating unit 106, an X-ray detecting unit 108, a rotation driving unit 110, a data acquiring circuit 116, And may include a transmitter 120.

The gantry 102 according to an embodiment of the present invention may include a rotating frame 104 in the form of a ring that is rotatable based on a predetermined rotation axis RA. The rotating frame 104 may also be in the form of a disk.

The rotating frame 104 may include an X-ray generating unit 106 and an X-ray detecting unit 108 arranged to face each other with a predetermined field of view (FOV). In addition, the rotating frame 104 may include an anti-scatter grid 114. Scattering prevention grid 114 may be positioned between the X-ray generation unit 106 and the X-ray detection unit 108.

 In medical imaging systems, X-ray radiation reaching the detector (or photosensitive film) includes attenuated primary radiation, which forms useful images, as well as scattered radiation, which degrades the quality of the image . An anti-scatter grid may be placed between the patient and the detector (or photosensitive film) to transmit the majority of the radiation and attenuate the scattered radiation.

For example, the anti-scatter grid may include strips of lead foil, solid polymer material, solid polymer and fiber composite material, etc. And the interspace material of the interlayer material may be alternately stacked. However, the form of the scattering prevention grid is not necessarily limited thereto.

The rotation frame 104 receives a drive signal from the rotation drive unit 110 and can rotate the X-ray generation unit 106 and the X-ray detection unit 108 at a predetermined rotation speed. The rotation frame 104 can receive a driving signal and power from the rotation driving unit 110 through a slip ring (not shown) in a contact manner. In addition, the rotation frame 104 can receive a driving signal and power from the rotation driving unit 110 through wireless communication.

 The X-ray generator 106 receives voltage and current from a power distribution unit (PDU) (not shown) through a slip ring (not shown) through a high voltage generator (not shown) . When the high voltage generating section applies a predetermined voltage (hereinafter referred to as a tube voltage), the X-ray generating section 106 can generate X-rays having a plurality of energy spectra corresponding to this predetermined tube voltage .

The X-ray generated by the X-ray generator 106 may be emitted in a predetermined form by a collimator 112.

The X-ray detector 108 may be positioned to face the X-ray generator 106. The X-ray detecting unit 108 may include a plurality of X-ray detecting elements. The single x-ray detection element may form a single channel, but is not necessarily limited thereto.

The X-ray detector 108 senses X-rays generated from the X-ray generator 106 and transmitted through the object 10, and can generate an electric signal corresponding to the intensity of the sensed X-rays.

The X-ray detector 108 may include an indirect method for detecting and converting radiation into light, and a direct method detector for detecting and converting the radiation directly into electric charge. An indirect X-ray detector can use a Scintillator. In addition, a direct-type X-ray detector can use a photon counting detector. A Data Acquisition System (DAS) 116 may be coupled to the X-ray detector 108. The electrical signal generated by the X-ray detector 108 may be collected at the DAS 116. The electric signal generated by the X-ray detecting unit 108 may be collected in the DAS 116 by wire or wirelessly. The electric signal generated by the X-ray detecting unit 108 may be amplified by an amplifier (not shown) To an analog / digital converter (not shown).

Only some data collected from the X-ray detector 108 may be provided to the image processor 126 or only some data may be selected by the image processor 126 depending on the slice thickness or the number of slices.

The digital signal may be provided to the image processing unit 126 through the data transmission unit 120. The digital signal may be transmitted to the image processing unit 126 via the data transmission unit 120 in a wired or wireless manner.

The controller 118 according to an embodiment of the present invention can control the operation of each module of the medical imaging apparatus 100. [ For example, the control unit 118 includes a table 105, a rotation driving unit 110, a collimator 112, a DAS 116, a storage unit 124, an image processing unit 126, an input unit 128, 130, the communication unit 132, and the like.

The image processing unit 126 may receive data obtained from the DAS 116 (for example, pure data before processing) through the data transmission unit 120 and perform pre-processing.

The preprocessing may include, for example, a process of non-uniformity of sensitivity correction between channels, a sharp decrease in signal intensity or a process of correcting loss of signal due to an X-ray absorber such as a metal.

The output data of the image processing unit 126 may be referred to as raw data or projection data. Such projection data can be stored in the storage unit 124 together with shooting conditions (e.g., tube voltage, shooting angle, etc.) at the time of data acquisition.

The projection data may be a set of data values corresponding to the intensity of the X-ray passing through the object roll. For convenience of explanation, a set of projection data simultaneously obtained with the same shooting angle for all the channels is referred to as a projection data set.

The storage unit 124 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (SD, XD memory, etc.), a RAM (Random Access Memory) SRAM (Static Random Access Memory), ROM (Read Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory) And may include at least one type of storage medium.

Also, the image processing unit 126 can reconstruct the cross-sectional image of the object using the acquired projection data set. Such a cross-sectional image may be a three-dimensional image. In other words, the image processing unit 126 may generate a three-dimensional image of the object using a cone beam reconstruction method based on the acquired projection data set.

External inputs for X-ray tomography conditions, image processing conditions and the like can be received through the input unit 128. For example, the X-ray tomography conditions include a plurality of tube voltages, energy value setting of a plurality of X rays, selection of a photography protocol, selection of an image reconstruction method, FOV area setting, number of slices, slice thickness, Processing parameter setting, and the like. The image processing condition may include resolution of the image, setting of the attenuation coefficient for the image, and setting of the combination ratio of the image.

The input unit 128 may include a device or the like for receiving a predetermined input from the outside. For example, the input unit 128 may include a microphone, a keyboard, a mouse, a joystick, a touch pad, a touch pan, a voice, and a gesture recognition device.

The display unit 130 may display an X-ray image reconstructed by the image processing unit 126. FIG.

Transmission and reception of data, power, etc. between the above-described elements can be performed using at least one of wired, wireless, and optical communication.

The communication unit 132 can perform communication with an external device, an external medical device, or the like through the server 134 or the like. The communication unit 132 may be connected to the network by wire or wirelessly to perform communication with the server 134, an external medical device (not shown), or an external device (not shown). The communication unit 132 can exchange data with other medical devices in the hospital server or the hospital connected through the PACS (Picture Archiving and Communication System).

In addition, the communication unit 132 may perform data communication with an external device (not shown) according to a DICOM (Digital Imaging and Communications in Medicine) standard.

The communication unit 132 can transmit and receive data related to the diagnosis of the object through the network. Also, the communication unit 132 can transmit and receive medical images and the like acquired from other medical devices 136 such as an MRI apparatus and an X-ray apparatus.

Further, the communication unit 132 may receive the diagnosis history or the treatment schedule of the patient from the server 134, and may utilize it for clinical diagnosis of the patient. In addition, the communication unit 132 may perform data communication with a server 134 or a medical device (not shown) in a hospital, as well as with a portable terminal (not shown) of a user or a patient.

In addition, information on the abnormality of the equipment and the quality management status information can be transmitted to the system administrator or the service person through the network, and the feedback can be received.

3 is a flow chart illustrating a process for reconstructing a medical image in accordance with one embodiment.

는 잡음 전송 실측 데이터

로부터, 수학식 1의 비용 함수(cost function)를 최소화 함으로써, 재구성(reconstruct)될 수 있다.Non-negative image

The noise transmission actual data

Can be reconstructed by minimizing the cost function of Equation (1).

Where L (x) is a negative loglikehood term and Ψo (x) is a patch based low rank regularization term.

- Separable Paraboloidal Surrogates

L (x) can be minimized through separable paraboloidal surrogates. L (x) can be defined from Poisson statistics as: " (2) "

Here, b _i means a blank scan in detector i, and r _i means background noise in detector i.

Since it is difficult to minimize L (x), Hakan Erdogan and Jeffrey A Fessler's "Ordered sunsets algorithms for transmission tomography," Physics in Medicine and biology , vol. 44, no. 11, pp. Separable paraboloidal surrogates proposed in < RTI ID = 0.0 > 2835-2851, < / RTI > First, the quadratic surrogate function is expressed by Equation 3 below.

이고, v_i는

의 곡률이다.Herein, n means repetition number (iteration)

, V _i is

.

Then, a separable surrogate of the quadratic surrogate can be applied as Equation (4) below.

가 이용될 수 있다.Finally, instead of a negative log likelihood term L (x), the equilibrium parabolic replacement function

Can be used.

- Patch-based Low Rank Regularization

The medical imaging apparatus can acquire a reference image of the object and a reconstruction target image of the object (S310).

In general, a reference image is an image obtained based on sufficient data, and may be a relatively low artifact image. Also, the reconstructing target image may be an image obtained based on insufficient data, and may be an image having a relatively large artifact.

According to one embodiment, the medical imaging device may acquire a medical image of a periodically moving object. For example, when a medical imaging device desires to acquire a medical image related to a deep space, it is difficult to obtain a medical image of a high resolution due to a movement of the medical imaging device due to periodic heartbeat. Therefore, in order to photograph a target object at a time when movement is small, a reference image and a reconstruction target image can be acquired based on the electrocardiographic gating (ECG gating) number.

4 is a conceptual diagram showing a configuration in which the medical imaging apparatus acquires image data based on the electrocardiogram data 400. [ The medical imaging device can acquire the electrocardiogram data 400. [ The medical imaging device may acquire the electrocardiogram data 400 by the electrocardiogram data measuring device included in the medical imaging device or may receive the electrocardiogram data 400 generated from the external device. The medical imaging device may determine a systolic 430-1 and a diastolic 430-2 based on the electrocardiographic data 400. The medical imaging apparatus according to one embodiment can acquire the reference image based on the diastolic data 420, which is data on the object of the diastolic 430-2. In addition, the medical imaging apparatus can acquire the reconstruction target image based on the systolic data 410, which is data on the object of the systolic 430-1. The reference image based on the diastolicity data 420 may be a reconstructed image in advance. The reference image based on the diastolic data 420 may be used as a reference image for performing the patch-based low-rank normalization as shown in FIG.

Referring again to FIG. 3, in step S310, the medical imaging apparatus according to an exemplary embodiment acquires a reference image and a reconstruction target image, and then determines at least one reference area corresponding to a unit area in the reconstruction target image in the reference image (S320). In the present specification, the unit area may be referred to as a patch. The reconstruction target image and the reference image may be divided into a plurality of unit areas referred to as a patch.

5, the medical imaging device includes at least one reference area 515-1, 515-2, 515-3 in a diastolic image 520 corresponding to a unit area 515 in a cardiac image 510 of a systole, Can be determined. 5 is a conceptual diagram illustrating a reference area corresponding to a unit area according to an embodiment. According to an exemplary embodiment, the at least one reference region 515-1, 515-2, and 515-3 may have characteristics similar to those of the unit region 515. [

According to one embodiment, the patch-based low-rank normalization of a group of patches that are unit areas can be written as: < EMI ID = 6.0 >

및 적어도 하나의 참조 영역에 포함된 데이터인

를 벡터로 하는 행렬(matrix)이다. 즉,

는 유사 패치 추출 연산자(a similar patch extraction operator)를 의미한다. 또한,

은 참조 영상으로서, 이완기의 영상일 수 있다.Here, v _p is data included in the unit area

And data included in at least one reference area

As a vector. In other words,

Means a similar patch extraction operator. Also,

Can be a diagonal image as a reference image.

For example, referring to FIG. 6, FIG. 6 illustrates an example in which a medical imaging device determines a matrix. The medical imaging device may determine the patch 615 from the reconstruction object image and select at least one patch 625-1, 625-2, 625-3 from the reference image. The medical imaging device may generate a matrix 610 based on the selected patches 615, 625-1, 625-2, and 625-3.

를 구할 필요가 있다. 이를 위하여, 의료 영상 장치는 재정련 유사도 탐색(re-fineable similarity searches)을 수행할 수 있다. 예를 들어, 의료 영상 장치는 x의 이전 추정(the previous estimation of x)을 이용하여 유사도 매핑(similarity mapping)

을 고정할 수 있다. 이후, 의료 영상 장치는 업데이트된 영상 x로부터 새로운 추정값을 구할 수 있다. 즉, 의료 영상 장치는 단위 영역 및 적어도 하나의 참조 영역 내에 포함된 데이터에 기초하여 행렬을 생성하고, 행렬에 기초하여 고유값(eigen value)를 결정할 수 있다. 의료 영상 장치는 결정된 고유값에 기초하여 단위 영역 내의 데이터를 업데이트할 수 있다.Thereafter, the data in the unit area can be updated based on the data included in at least one reference area (S330). B is the number of pixels (voxels in the case of a three-dimensional image) in the patch, which is a unit area, and Q is the number of selected patches. Np is the number of patch based low rank executions. According to an embodiment, the number of executions may be equal to the number of pixels (voxels in the case of three-dimensional images). In order to construct a group of similar patches,

. For this purpose, the medical imaging device may perform re-fine similarity searches. For example, a medical imaging device may use similarity mapping using the previous estimation of x,

Can be fixed. Thereafter, the medical imaging device may obtain a new estimate from the updated image x. That is, the medical imaging device may generate a matrix based on the data contained in the unit area and the at least one reference area, and may determine an eigen value based on the matrix. The medical imaging device may update the data in the unit area based on the determined eigenvalue.

Here, the rank operator of Equation (6) is not convex. Thus, a nuclear norm can be used in convex relaxation.

이고, σ_k(v_p)는 v_p에서 k번째로 큰 특이값(singular value)를 의미한다. 여기서, 아래 수학식 8의 오목 랭크(concave rank)가 이용될 수 있다.here,

, And σ _k (v _p ) means the k-th singular value at v _p . Here, a concave rank of the following equation (8) can be used.

는 수학식 9의 후버 함수(Huber function)일 수 있다.Here, according to one embodiment,

May be a Huber function of Equation (9).

는 순전히 볼록(convex)하다. 그러므로, 르장드르-펜첼(Legendre-Fenchel) 변환에 따르면 아래 수학식 10에서와 같이

가 존재한다.According to one embodiment, a concave-convex procedure (CCCP) can be used to solve the non-convex penalties. here,

Is purely convex. Therefore, according to the Legendre-Fenchel transformation, as shown in Equation 10 below

Lt; / RTI >

는

일 때 볼록(convex)하며, 아닌 경우 일반적으로 볼록하지 않다. 일 실시 예에 따르면,

가 볼록하지 않고, 닫힌 형태의 표현(close form expression)을 가지지 않더라도, 수학식 11에서와 같이 수축 연산자(shrinkage operator)를 이용하여 수학식 10의 최소화를 위한 닫힌 형태의 해법(closed form solution)이 제공된다.

The

Convex, and generally not convex. According to one embodiment,

Closed form solution for minimizing Equation (10) using a shrinkage operator as in Equation (11), even if the closed form solution is not convex and does not have a closed form expression / RTI >

Therefore, the medical imaging device can perform a patch based low rank regularization as shown in Equation (12). That is, the reconstruction target image can be reconstructed by minimizing the cost function based on the patch based low rank regularization.

- Optimization Framework

가 주어졌을 때, 수학식 13에 따른 최소화 문제(minimization problem)을 해결할 필요가 있다.According to one embodiment, R and < RTI ID = 0.0 &

, It is necessary to solve the minimization problem according to Equation (13).

Due to Ψ (x), the cost function is non-convex. Here, using Equation (5) and Equation (12), it is possible to have the minimization problem as shown in Equation (14).

Equation (14) can be solved by alternately performing minimization for w _p and minimization for x.

1. Minimization to w _p

에 대해 독립적이다. 또한, 최소화 문제는 각 패치들에 대해 분해될 수 있다. 보다 상세하게는, 아래 수학식 15가 주어진다.The minimization problem

Lt; / RTI > In addition, the minimization problem can be resolved for each patch. More specifically, the following expression (15) is given.

Using the shrinkage relationship of Equation (11), the closed form solution of Equation (15) is given by Equation (16) below.

는 요소 특이값 축소 연산자(element singular value shrinkage operator)에 의한 요소를 의미한다.here,

Means an element by an element singular value shrinkage operator.

2. Minimize to x

Given W ^{(n + 1)} , the medical imaging device may obtain a closed form solution for updating x ^{(n + 1)} . The medical imaging device can calculate a fixed point equation of the slope of the cost function for x ⁽ⁿ⁾ . The solution of the closed form is given by Equation (17) below.

FIG. 7 is a schematic diagram illustrating a structure of a medical imaging apparatus according to an embodiment of the present invention. Referring to FIG.

The medical imaging device 700 according to one embodiment may include an image processing unit 726 and a display unit 730. The image processing unit 726 according to an exemplary embodiment may acquire a reference image of a target object and a reconstruction target image of a target object. In addition, the image processing unit 726 may determine at least one reference area corresponding to a unit area in the reconstruction target image in the reference image, and may update the data in the unit area based on the data included in the at least one reference area . The display unit according to one embodiment may display an image reconstructed by the image processing unit 726. [ Here, the reference region may have characteristics similar to those of the unit region.

The medical imaging apparatus 700 according to an embodiment may further include an electrocardiogram data acquisition unit (not shown) for acquiring electrocardiogram data. The image processing unit 726 acquires a reference image based on the diastolic data, which is data on the subject of the diastolic period determined based on the electrocardiogram data, and based on the systolic data, which is data on the subject of the systolic motion determined on the basis of the electrocardiogram data A reconstruction target image can be acquired.

According to one embodiment, the image processing unit 726 can reconstruct the reconstruction target image by minimizing the cost function using the Patch-based Low Rank Regularization. Here, the image processing unit 726 may perform patch-based low-rank normalization based on Equation (12). In this case, the image processing unit 726 may use a shrinkage operator.

According to one embodiment, the image processing unit 726 may generate a matrix based on the data contained in the unit area and at least one reference area. The image processing unit 726 may determine an eigen value based on the matrix and update data in the unit area based on the determined agreement value.

One embodiment of the present invention may also be embodied in the form of a recording medium including instructions executable by a computer, such as program modules, being executed by a computer. Computer readable media can be any available media that can be accessed by a computer and includes both volatile and nonvolatile media such as RAM, removable and non-removable media. In addition, the computer-readable medium may include both computer storage media and communication media. Computer storage media includes both volatile and nonvolatile, removable and non-removable media implemented in any method or technology for storage of information such as computer readable instructions, data structures, program modules or other data. Communication media typically includes any information delivery media, including computer readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave, or other transport mechanism. For example, computer storage media may be embodied in ROM, RAM, flash memory, CD, DVD, magnetic disk, magnetic tape, or the like.

It will be understood by those skilled in the art that the foregoing description of the present invention is for illustrative purposes only and that those of ordinary skill in the art can readily understand that various changes and modifications may be made without departing from the spirit or essential characteristics of the present invention. will be. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. For example, each component described as a single entity may be distributed and implemented, and components described as being distributed may also be implemented in a combined form.

The scope of the present invention is defined by the appended claims rather than the detailed description and all changes or modifications derived from the meaning and scope of the claims and their equivalents are to be construed as being included within the scope of the present invention do.

Claims (17)

A method for reconstructing a tomographic image of a medical imaging device,
Acquiring a reference image of a target object and a reconstruction target image of the target object;
Determining in the reference image at least one reference region corresponding to a unit region in the reconstruction target image; And
And updating data in the unit area based on data included in the at least one reference area.
The method according to claim 1,
Wherein the at least one reference region has characteristics similar to those of the unit region.
The method according to claim 1,
The method of reconstructing a tomographic image comprises:
Further comprising the step of acquiring electrocardiography (ECG) data,
Wherein the acquiring of the reference image comprises:
Acquiring the reference image based on the diastolic data, which is data on a diastolic subject determined on the basis of the electrocardiogram data,
Wherein the step of acquiring the reconstruction target image comprises:
And acquiring the reconstruction object image based on systolic data, which is data on a subject of the systole determined based on the electrocardiogram data.
The method according to claim 1,
Wherein updating the data in the unit area comprises:
And reconstructing the reconstructed image by minimizing a cost function using a patch-based Low Rank Regularization method.
5. The method of claim 4,
Wherein updating the data in the unit area comprises:
Generating a matrix based on data included in the unit area and the at least one reference area;
Determining an eigen value based on the matrix; And
And updating the data in the unit area based on the determined eigenvalue.
6. The method of claim 5,
Wherein updating the data in the unit area comprises:
Wherein the patch-based low-rank normalization is used based on the following equations.

Where v _p is the generated matrix, 隆_k (v _p ) is the kth largest singular value of v _p ,

Is a Huber function.
The method according to claim 6,
Wherein updating the data in the unit area comprises:
minimizing a cost function for w _p ; And
And minimizing a cost function for x based on a result of minimizing the cost function for w _p .
8. The method of claim 7,
The method of reconstructing a tomographic image comprises:
to, tomographic image reconstruction method for performing the step of minimizing a cost function with respect to the x repeatedly based on a result of minimizing a cost function with respect to phase, and the w _p that minimizes the cost function with respect to w _p.
5. The method of claim 4,
Wherein updating the data in the unit area comprises:
Wherein said cost function is minimized by using a shrinkage operator.
A computer-readable recording medium recording a program for causing a computer to execute the method of claim 1.
In a medical imaging device,
A reconstructing target image obtained by photographing a target object and a reconstructing target image obtained by photographing the target object; determining at least one reference area corresponding to a unit area in the reconstructing target image of the reference image; An image processing unit configured to update data in the unit area based on the data; And
And a display unit for displaying the reconstructed image according to the update result.
12. The method of claim 11,
Wherein the at least one reference region has characteristics similar to those of the unit region.
12. The method of claim 11,
Wherein the medical imaging device comprises:
And an electrocardiogram data acquiring section for acquiring electrocardiogram data,
Wherein the image processing unit comprises:
Acquiring the reference image based on the diastolic data which is data on the subject of the diastolic period determined based on the electrocardiogram data, and acquiring the reference image based on the systolic data, which is data on the systolic object determined on the basis of the electrocardiographic data, And acquires an image of the medical image.
12. The method of claim 11,
Wherein the image processing unit comprises:
And reconstructs the reconstruction target image by minimizing a cost function using a patch-based low rank normalization.
15. The method of claim 14,
Wherein the image processing unit comprises:
Generating a matrix based on data included in the unit area and the at least one reference area, determining an eigen value based on the matrix, and determining, based on the determined eigenvalue, And updates the data in the medical imaging device.
16. The method of claim 15,
Wherein the image processing unit comprises:
Characterized in that said patch based low rank normalization is used based on the following equation:

Where v _p is the generated matrix, 隆_k (v _p ) is the kth largest singular value of v _p ,

Is a Huber function.
12. The method of claim 11,
Wherein the image processing unit comprises:
Wherein the cost function is minimized using a shrinkage operator.

Images