KR102020157B1 - Method and system for detecting lesion and diagnosing lesion status based on fluid attenuation inverted recovery - Google Patents

Abstract

Disclosed are a method for detecting a lesion and checking a status based on FLAIR and a system thereof. The method comprises the steps of: dividing a FLAIR image of a patient into a first FLAIR image signal corresponding to a left hemisphere of the brain of the patient and a second FLAIR image signal corresponding to a right hemisphere of the brain of the patient to use the FLAIR image in diagnosing a disease of the patient; estimating the lesion through artificial intelligence in each of the first and second FLAIR image signals; and calculating a ratio with a corresponding region of the other one of the first and second FLAIR image signals as compared to a lesion region estimated as the lesion in any one of the first and second FLAIR image signals, wherein the calculation results are used to diagnose the progress status of the lesion.

Description

FLAIR-based lesion detection and status determination method and system {METHOD AND SYSTEM FOR DETECTING LESION AND DIAGNOSING LESION STATUS BASED ON FLUID ATTENUATION INVERTED RECOVERY}

Embodiment of the present invention relates to a solution for detecting lesion areas based on FLAIR (fluid attenuation inversion recovery), more specifically, correlated with the presence or absence of diffusion weighted imaging (DWI image) The present invention relates to an AI-based lesion detection method and a system and method for detecting lesions based on FLAIR and grasping the progress of lesions.

Recently, images of magnetic resonance imaging (MRI) have been widely used as medical images. MRI uses the magnetic properties of hydrogen particles (H +) in human tissues to capture and image the energy generated after resonance / relaxation phenomena according to tissues in a magnetic field, or a method or an image acquired by the apparatus. Refers to.

In order to detect diseases such as stroke lesions using MRI images, perfusion-weighted imaging (PWI) showing the blood flow of the brain in MRI images, and diffusion-weighted imaging showing the diffusion of water molecules, DWI) is widely used. For example, DWI is used to detect stroke lesion core regions, and PWI is used to detect penumbra regions that extend to the periphery, including stroke lesion core regions. In particular, DWI is the most sensitive technique for diagnosing acute cerebral infarction, with lesions appearing within minutes of development, helping to determine the treatment of acute cerebral infarction.

In addition, FLAIR (fluid attenuated inversion recovery) is a technique that suppresses signals from cerebrospinal fluid with a long reversal time (TI) and echo time (TE) of about 2500 msec. give.

As such, T2-weighted images, DWI images, and FLAIR images are useful for finding lesions and are widely used for the detection and diagnosis of various lesions. Figure 1 shows a variety of spin echo imaging techniques, (a) is a T1 weighted image (TR / TE: 500/10), (b) is a T2 weighted image (TR / TE: 3000/90), and ( c) shows a FLAIR image (TR / TE: 500/150), respectively. TR represents the time interval, or repetition time, between the pulses.

On the other hand, each of the medical images must undergo a different image processing process according to the type, there is a disadvantage that the detection or diagnostic device or its operating system is complicated and large.

In addition, most of the conventional lesion detection techniques using the above medical images require the participation of a user such as a doctor or the like to assist the user in diagnosing a lesion based on the medical image, and automatically diagnose the lesion. Is still limited.

Recently, machine learning or artificial intelligence has been introduced in the medical field, and active research and development is underway. However, the results of research and development for automatically diagnosing a lesion and predicting the prognosis of the lesion in relation to a brain disease such as a stroke such as a stroke are still insignificant.

The present invention was derived to overcome the limitations of the prior art, an object of the present invention is to automatically detect lesions using FLAIR (fluid attenuation inversiont recovery) image directly to determine the status of the lesion progress, etc. The present invention provides a method and apparatus for detecting and detecting the status of FLAIR-based lesions.

Another object of the present invention is to analyze the FLAIR image by artificial intelligence neural network directly detects the quantitative lesion area and volume of patients with acute brain disease, and to provide the progress information of the lesion, FLAIR-based lesion detection and status determination method and To provide a device.

Another object of the present invention is FLAIR, which can detect lesions based on FLAIR images and determine the progress of the lesions, with or without diffusion weighted imaging (DWI imaging) in medical images. To provide a method and apparatus for detecting lesions based on current status.

Another object of the present invention is to provide a computer-readable recording medium recording a program for implementing the above FLAIR-based lesion detection and status determination method.

FLAIR (fluid attenuation inversion recovery) based lesion detection and status determination method according to an aspect of the present invention for solving the technical problem, using a FLAIR (fluid attenuated inversion recovery) image of the patient, to diagnose the lesion of the patient To do this, the method comprises: dividing the first FLAIR image signal corresponding to the left hemisphere of the brain of the patient and the second FLAIR image signal corresponding to the right hemisphere of the brain; Estimating a lesion through artificial intelligence in each of the first and second FLAIR image signals; And calculating a ratio of one of the first and second FLAIR image signals to a corresponding region of the other one of the first and second FLAIR image signals compared to the lesion area estimated as the lesion. The calculation result generated in the calculating step is used to diagnose the progress of the lesion.

In one embodiment, the FLAIR-based lesion detection and grasping method may further include diagnosing the progress of the lesion based on the ratio, the location of the lesion and the size of the lesion area after the calculating. .

In an embodiment, the FLAIR-based lesion detection and status determination method may further include determining whether there is a diffusion weighted imaging (DWI) image corresponding to the FLAIR image before the dividing step, wherein the DWI image is present. If so, the method may further include adjusting a size of the FLAIR image before the dividing step, and after the dividing, matching the DWI image and the FLAIR image.

In one embodiment, the step of adjusting the position information obtained by the signal processing of the DWI image and the ADC (Apparent diffusion coefficient) calculation to obtain positional information as a threshold value, based on the positional information to obtain the DWI image It may be implemented to adjust the size of the image to be grafted to the FLAIR image.

FLAIR-based lesion detection and status determination device according to another aspect of the present invention for solving the above technical problem, including a computing device or mounted on the computing device to extract the lesion based on FLAIR and to determine the current status of the area, And a lesion condition grasping unit connected to an extracting unit and the region extracting unit, wherein the region extracting unit is configured to use an image signal of a FLAIR image of a patient to be input to diagnose a lesion of the patient. Dividing the first FLAIR image signal corresponding to the second FLAIR image signal corresponding to the right hemisphere of the brain and estimating a lesion in each of the first and second FLAIR image signals through artificial intelligence; The lesion state detecting unit may further include one of the first and second FLAIR image signals compared to the lesion region estimated as the lesion in any one of the first and second FLAIR image signals received from the region extractor. The ratio with the corresponding area is calculated, where the result of the calculation is used to diagnose the progress of the lesion.

In one embodiment, the FLAIR-based lesion detection and status determination device further comprises a signal recognition unit for determining whether there is a difference weighted imaging (DWI) image corresponding to the FLAIR image, before the area extractor divides the image signal; And when the signal recognizing unit recognizes that the DWI image exists, the region extracting unit may further include a size adjusting unit adjusting the size of the FLAIR image before the dividing unit, and after the region extracting unit completes the dividing, The apparatus may further include an image matching unit that matches the DWI image and the FLAIR image.

In one embodiment, the size adjusting unit obtains positional information having a threshold value, which is obtained through signal processing of the DWI image and calculating an ADC (Apparent Diffusion Coefficient), and based on the positional information, the size adjustment unit acquires the DWI image. You can resize the image to incorporate it into the FLAIR image.

In one embodiment, the FLAIR-based lesion detection and status determination device further comprises a final diagnostic unit for diagnosing the progress of the lesion based on the ratio, the position of the lesion and the size of the lesion area calculated by the lesion state determiner It may include.

An apparatus according to another aspect of the present invention for solving the above technical problem includes a computer-readable recording medium recording a program for implementing the FLAIR-based lesion detection and status determination method of any one of the above embodiments.

When using the above-described method and apparatus for detecting and presenting a lesion based on FLAIR (fluid attenuation inversion recovery), it is possible to generate quantitative lesion area, volume and progress information by directly analyzing the FLAIR image based on artificial intelligence. As a result, it is possible to quickly perform a status detection on the detection of the lesion and the progress of the lesion, and the effect of simplicity and cost reduction can be obtained.

In addition, according to the present invention, it is possible to reduce or omit manual efforts by automatically diagnosing lesion areas and progression of lesions directly in FLAIR images of patients with acute brain disease, and to diagnose the necessity of treatment at a specialist level. There is this.

Also, in patients with acute intraventricular bleeding such as acute acute hemorrhage, subarachnoid hemorrhage, and diseases including calcification, computed tomography (CT) is superior to MR (magnetic resonance), which is one of the MR imaging FLAIR images. When used in the method or apparatus of the present invention, it is possible to obtain better disease detection performance than CT for the above-described lesions, and also to easily grasp the current status of the lesions.

1 is an exemplary diagram showing various conventional spin echo imaging techniques.
2 is a block diagram showing an apparatus for detecting and detecting a current condition based on FLAIR (fluid attenuation inversion recovery) according to an embodiment of the present invention.
3 is a flow chart for explaining one of the main principles of operation of the apparatus of FIG.
4 is a structural diagram of a convolutional neural network that can be employed in the apparatus of FIG.
5 is a flow chart for a FLAIR-based lesion detection and status diagnosis method according to another embodiment of the present invention.
6 is a block diagram of one embodiment of an apparatus configuration that implements the method of FIG.
FIG. 7 is a diagram for describing information output on a screen of a user terminal coupled or connected to the apparatus of FIG. 6.

Specific structural or functional descriptions of the embodiments according to the inventive concept disclosed herein are provided only for the purpose of describing the embodiments according to the inventive concept. It may be embodied in various forms and is not limited to the embodiments described herein.

Embodiments according to the inventive concept may be variously modified and have various forms, so embodiments are illustrated in the drawings and described in detail herein. However, this is not intended to limit the embodiments in accordance with the concept of the invention to the specific forms disclosed, and includes all changes, equivalents, or substitutes included in the spirit and scope of the present invention.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this specification, terms such as "comprise" or "having" are intended to indicate that there is a feature, number, step, action, component, part, or combination thereof described herein, but one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, parts, or combinations thereof.

The terminology used herein is as follows.

T2-weighted imaging: refers to a technique obtained from a magnetic resonance imaging (MRI) in a specific pulse sequence or an image obtained by this technique, which is mainly a Provide structural information.

FLAIR (Fluid attenuated inversion recovery): Signal acquisition using magnetic imaging devices or images acquired using this technique that attenuate the signal of cerebrospinal fluid with long reversal and echo times, making it easier to find lesions that are easy to miss in T2-weighted images. Refers to. FLAIR may be referred to as fluid attenuation inversion recovery.

DWI (Diffusion weighted imaging): refers to diffusion-weighted images obtained from magnetic resonance imaging, and provides information on the extent and whether the water molecules in the cell tissue diffuse in a specific direction.

Perfusion weighted imaging (PWI): It refers mainly to perfusion-weighted images (simply perfusion images) obtained from magnetic resonance imaging, and provides information on concentration changes over time of injected contrast medium.

Penumbra: A semi-shaded area in the image that is caused by an ischemic event or embolism, which locally reduces oxygen transport function, resulting in hypoxic cell death, or a viable area within a few hours of proper treatment.

Apparent diffusion coefficient (ADC): An apparent diffusion coefficient obtained from magnetic resonance images, which provides information on diffusion barriers of human internal tissues.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the drawings.

FIG. 2 is a block diagram illustrating a device for detecting and presenting a lesion based on fluid attenuation inversion recovery (FLAIR) (hereinafter, simply referred to as a 'diagnosis device').

Referring to FIG. 2, the diagnostic apparatus 100 according to the present exemplary embodiment is used to process a FLAIR image input through an area extractor 10 and a lesion state determiner 20 to detect a lesion and diagnose a lesion progression state. Output the signal, data or information. The diagnostic apparatus 100 may include a computing device or may be mounted on the computing device to extract a lesion based on FLAIR and determine a current state of the lesion.

The main feature of the diagnostic apparatus 100 according to the present embodiment is to estimate a lesion area based on artificial intelligence using a FLAIR image directly, calculate the ratio of the relative hemisphere side area to the estimated lesion area, and estimate the lesion area. Based on the calculated ratio, we can effectively diagnose the progress of the lesion based on artificial intelligence. If the calculated ratio is small, that is, the difference between the two signals to be compared is small, the lesion may be old.

The region extractor 10 corresponds to the first FLAIR image signal corresponding to the left hemisphere of the patient's brain and the right hemisphere of the patient's brain, in order to use the input image signal of the patient's FLAIR image to diagnose the patient's lesion. The second FLAIR image signal may be divided, and the lesion may be estimated from each of the first and second FLAIR image signals through artificial intelligence.

The lesion state determiner 20 may perform the rest of the first and second FLAIR image signals compared to the lesion region estimated as the lesion in any one of the first and second FLAIR image signals received from the region extractor 10. The ratio with one corresponding area can be calculated. The calculation results can be used to diagnose the progression of the lesion through artificial intelligence.

In addition, the diagnostic apparatus 100 or the diagnostic system 100A including the diagnostic apparatus 100 according to the present exemplary embodiment may include the final diagnosis unit 30 in addition to the region extraction unit 10 and the lesion state determining unit 20. It may further include. The final diagnosis unit 30 uses artificial intelligence as the ratio (reference ratio) of the lesion area detected by the region extraction unit 10 and the hemisphere-side corresponding region compared to the lesion area detected by the lesion state determining unit 20. Learning can diagnose the progress of the lesion. The final diagnosis unit 30 may diagnose the progress state of the lesion based on the ratio calculated by the lesion state determining unit 20, the position of the lesion detected by the region detection unit 10, and the size of the lesion region.

The diagnosis result of the final diagnosis unit 30 may be diagnostic information regarding the progress state of the patient's lesion, but is not limited thereto. The diagnosis result may be diagnosed as a diagnosis result delivered to a doctor or predicted the progress state of the lesion. It may have a form of information output that is processed to assist in doing so.

The artificial intelligence 70 employable in the diagnostic apparatus 100 of the present embodiment has a form that is connected to the diagnostic apparatus 100 as a separate preprocessor, recognizer, learner, or a combination of independent functional units or components thereof. It can be provided with the form mounted in a component part. Artificial intelligence 50 may be designed to include machine learning 52 or deep learning 54 in addition to neural networks. Deep learning 54 may include a convolutional neural network or the like. Such an artificial intelligence 70 or an artificial intelligence system including the same may be referred to as.

3 is a flow chart for explaining one of the main principles of operation of the apparatus of FIG.

Referring to FIG. 3, the FLAIR-based lesion detection and status determination method according to the present embodiment (hereinafter, simply referred to as a 'diagnosis method') includes a hemisphere division step (S31), a lesion estimation step (S32), and a ratio calculation step (S33). ).

In the hemispheric dividing step (S31), an image signal of a FLAIR image may be divided into a first FLAIR image signal and a second FLAIR image signal in order to separate the input FLAIR image into a left hemisphere image and a right hemisphere image of the patient's brain. The FLAIR image may be an image of a brain of a patient suffering from an acute stroke, such as cerebral infraction and cerebral hemorrhage, by magnetic resonance imaging (MRI).

In the lesion estimation step S32, the lesion or the lesion region is estimated in the first FLAIR image and the second FLAIR image through artificial intelligence.

Next, in the ratio calculation step S33, the corresponding region of the hemisphere relative to the lesion region of one hemisphere having the lesion region estimated as the lesion is extracted, and the ratio of the corresponding region to the lesion region is calculated.

The ratios calculated directly from the FLAIR images through the above steps are information on lesion detection and the progress of the lesion, and this information can be effectively used as a data for diagnosing the progress of the lesion by artificial intelligence based analysis. have.

4 is a structural diagram of a convolutional neural network that may be employed in the apparatus of FIG. 2.

The FLAIR-based lesion detection and status determination apparatus according to the present embodiment may use a deep learning architecture as an artificial intelligence neural network for lesion detection or status determination. Deep learning architecture can be implemented to learn about the lesion, lesion area, or lesion candidate of each medical image using a convolutional neural network (CNN), deconvolution, skip connection, and the like. Can be.

In one example, the deep learning architecture may have a form including a convolution network and a deconvolution network and a shortcut. As shown in FIG. 6, the deep learning architecture stacks a 3x3 color convolution layer and an activation layer (ReLU) and strides a 2x2 size filter to extract local features of the medical image (X). Apply to (stride) 1 to repeat the calculation of the convolution block to the next lower depth level four times, and then apply a 2x2 deconvolution layer and an activation layer (ReLU) After the connection to the next higher depth level, the operation of the inverse convolution block, which stacks the 3x3 color convolution layer and the activation layer, is repeated four times, where each level of the convolutional network including the operation of the convolution block of each level is performed. Copy and contaten the image of the convolutional block of to the corresponding convolutional level of the inverse convolutional network of the same level. ate) may be configured to perform a convolution operation in each block.

Convolution blocks in the convolutional network and the deconvolutional network may be implemented as a combination of conv-ReLU-conv layers. In addition, the output of the deep learning architecture may be obtained through a classifier connected to a convolution network or a deconvolution network, but is not limited thereto. The classifier can be used to extract local features from an image using a fully connectivity network (FCN) technique.

In addition, the deep learning architecture may be implemented to further use an inception module or multi filter pathway in the convolution block, depending on the implementation. Different filters in the inception module or multi-filter path may include a 1 × 1 filter.

For reference, when the input image has 32 horizontal, 32 vertical, and RGB channels in the deep learning architecture, the size of the input image X may be [32x32x3]. In the deep learning architecture's convolutional neural network (CNN), the convolutional (CONV) layer is associated with some regions of the input image and can be designed to calculate the dot product of these linked regions and their weights. .

Here, the rectified linear unit (ReLU) layer is an activation function applied to each element, such as max (0, x). The ReLU layer does not change the size of the volume. The POOLING layer may output a reduced volume by performing downsampling or subsampling on the dimension represented by (horizontal and vertical).

The fully-connected (FC) layer may calculate the class scores and output a volume having a size of, for example, [1 × 1 × 10]. In this case, the ten numbers correspond to class scores for ten categories. The pre-connection layer is connected to all the elements of the previous volume. There, some layers may have parameters but some layers may not have parameters. CONV / FC layers may include weight and bias as well as input volume as an activation function. Meanwhile, the ReLU / POOLING layers are fixed functions, and the parameters of the CONV / FC layer can be learned in gradient descent so that the class score for each image is the same as the label of the image.

5 is a flow chart of a FLAIR-based lesion detection and status determination method according to another embodiment of the present invention.

Referring to FIG. 5, in the diagnostic method according to the present exemplary embodiment, before the FLAIR image is input to the diagnostic apparatus, after the FLAIR image is input or before the FLAIR image is divided into the hemispherical images by the region extraction unit, the FLAIR image is input to the FLAIR image. It may be determined whether there is a corresponding difference weighted imaging (DWI) image (S51).

As a result of the determination in step S51, if there is a DWI image, the area extractor may adjust the size of the DWI image or the size of the FLAIR image through the size adjusting unit. Image size adjustment is performed by acquiring positional information whose threshold value is obtained through signal processing of the DWI image and calculating an ADC (Apparent diffusion coefficient) (S52), and based on the acquired positional information, the size of the DWI image It may be made to graf the FLAIR image by adjusting (S53). The image graft may include matching the DWI image and the FLAIR image.

Next, the FLAIR image is divided into hemisphere images corresponding to the left hemisphere and the right hemisphere (S54), the lesion is extracted from each hemisphere image, and the lesion region of one hemisphere from which the lesion is extracted and the correspondence of the other hemisphere in contrast to the lesion region The ratio with the area can be calculated (S55).

Then, the progress of the disease or the lesion may be diagnosed through artificial intelligence-based learning based on the information of the previously calculated ratio and the previously detected lesion area.

On the other hand, if there is no corresponding DWI image in the step (S51), in the present embodiment may perform the steps of Figure 3 regardless of the presence or absence of the DWI image to detect the lesion area and determine the progress of the lesion .

6 is a block diagram of one embodiment of an apparatus configuration that implements the method of FIG.

Referring to FIG. 6, the diagnostic apparatus according to the present embodiment may be implemented as a computing device. The computing device may include a memory 40 storing a program and a processor 50 connected to the memory 40 to execute a program, and further include a communication interface 60 according to an implementation.

The memory 40 may store the signal recognizer 10a, the area extractor 10, the ratio calculator 20a, and the final diagnosis unit 30 in the form of a software module, and the software module may be operated by the computing device. The processor 50 may be mounted on the processor 50 to perform a corresponding function or operation.

In the present embodiment, the signal recognizer 10a may include a means or a component for recognizing whether a DWI image corresponding to the FLAIR image exists. The area extractor 10 may include a size adjuster 11, a signal splitter 12, an image matcher 13, and a lesion estimator 14. Basic configurations of the area extractor 10 and the final diagnosis unit 30 may correspond to the corresponding components described above with reference to FIG. 3. In addition, the ratio calculator 20a may correspond to the lesion state determiner described above with reference to FIG. 3.

The size adjusting unit 11 acquires positional information having a threshold value based on the signal processing of the DWI image and calculating an ADC (Apparent Diffusion Coefficient), and integrates the DWI image into the FLAIR image based on the positional information. You can adjust the size of the image.

The image matching unit 13 may operate so that the signal dividing unit 12 matches the DWI image and the FLAIR image after the FLAIR image is divided into hemisphere images.

The area extractor 10 or the final diagnosis unit 30 may include a means or a component for performing a function of artificial intelligence, machine learning, or deep learning.

The processor 50 may correspond to a microprocessor, a processing unit, or a control unit, and may include a central processing unit (CPU) or a core, and execute the program stored in the memory 40 to perform the above-described diagnosis. It can be implemented to execute the method.

The communication interface 60 may include at least one sub-communication system for an intranet, the Internet, a wired network, a wireless network, a satellite network, or a combination thereof. The communication interface 60 may be connected to an MRI device or a database and used to receive a FLAIR image or a DWI image.

FIG. 7 is a diagram for describing information output on a screen of a user terminal coupled to or connected to the apparatus of FIG. 6.

Referring to FIG. 7, the diagnostic apparatus according to the present embodiment may display information based on a result of calculation of a lesion area F2 and a reference ratio of a lesion of a patient with an acute brain disease. 710, 720, and 730.

The lesion area F2 may be displayed in the form overlapping the FLAIR image F in the first display area 710. In addition, the lesion area F2 may be displayed together with its location on the brain map Ref expressing functional role information of the brain area in color or the like.

The automatic diagnosis result by the artificial intelligence may be output to the second display area 720 as information according to the functional role of the brain region. In one example, information may be provided as to whether the detected lesion area is associated with functional role areas of the brain area, ie, motor, sensory, language, visual and auditory areas.

The above-described diagnostic information about the lesion area F2 and the lesion progress state may be changed according to the setting of the user interface 740. For example, the user interface 740 may provide an operation function, a perspective view, a six-sided view, a cross-sectional view, and the like for the enlargement and rotation of the lesion area and the progress state representation thereof. In addition, the user interface 740 may include a function for selecting and executing data output or data output format conversion.

The user terminal may include at least one of a desktop computer, a personal digital assistant (PDA), a mobile terminal, a network terminal, and a server device connected to the diagnostic device.

As described above, according to the present invention, it is possible to directly detect the lesion area and diagnose the progression of the lesion on the basis of artificial intelligence using a FLAIR image, thereby providing a rapid and efficient acute brain disease diagnosis solution. have.

Although the present invention has been described with reference to the embodiments with reference to the drawings, this is merely exemplary, and those skilled in the art will appreciate that various modifications and equivalent other embodiments are possible therefrom. Therefore, the true technical protection scope of the present invention will be defined by the technical spirit of the appended claims.

Claims (9)

Determining whether a signal recognition unit mounted on a processor of the computing device includes a diffusion weighted imaging (DWI) image corresponding to a fluid attenuated inversion recovery (FLAIR) image of the patient;
Adjusting, by the size adjusting unit mounted on the processor, the size of the FLAIR image if the DWI image exists;
A first FLAIR image signal corresponding to the left hemisphere of the brain of the patient and a second FLAIR image corresponding to the right hemisphere of the brain, for use by the signal splitter mounted on the processor to use the FLAIR image to diagnose the lesion of the patient. Dividing into a signal;
An image matching unit mounted on the processor, matching the first FLAIR image corresponding to the first FLAIR image signal and the second FLAIR image corresponding to the second FLAIR image signal with the DWI image;
Estimating a lesion through artificial intelligence in each of the first and second FLAIR image signals based on a matching result of the image matching unit; And
The lesion state acquiring unit mounted in the processor includes a corresponding region of the other one of the first and second FLAIR image signals compared with the lesion region estimated as the lesion in any one of the first and second FLAIR image signals. Calculating the ratio of
The adjusting may include obtaining positional information having a threshold value, which is obtained through signal processing of the DWI image and calculating an ADC (Apparent Diffusion Coefficient), and grafting the DWI image to the FLAIR image based on the positional information. To resize the video so that
The calculation result generated in the calculating step is used to diagnose the progress of the lesion, FLAIR based lesion detection and grasping method. The method according to claim 1,
And diagnosing the progress of the lesion based on the ratio, the location of the lesion, and the size of the lesion region after the calculating. delete delete A device that includes a computing device or is mounted on a computing device and extracts a lesion based on FLAIR (fluid attenuated inversion recovery) and determines the status of the lesion.
A signal recognizer configured to determine whether there is a diffusion weighted imaging (DWI) image corresponding to a fluid attenuated inversion recovery (FLAIR) image of the patient;
A size adjusting unit adjusting the size of the FLAIR image when the DWI image exists;
The image signal of the FLAIR image of the patient is divided into a first FLAIR image signal corresponding to the left hemisphere of the patient's brain and a second FLAIR image signal corresponding to the right hemisphere of the brain for use in diagnosing a lesion of the patient. A signal divider;
An image matching unit for matching the first FLAIR image corresponding to the first FLAIR image signal divided by the signal splitter and the second FLAIR image corresponding to the second FLAIR image signal with the DWI image;
A lesion estimating unit estimating lesions in each of the first and second FLAIR image signals through artificial intelligence based on a matching result of the image matching unit;
Corresponding to the other of the first and second FLAIR image signals compared to the lesion area estimated as the lesion in any one of the first and second FLAIR image signals received from the signal splitter and the lesion estimator Includes; lesion status grasping unit for calculating the ratio with the area;
The size adjusting unit adjusts the size of the image to graf the DWI image to the FLAIR image based on positional information having a threshold value as a threshold value obtained through signal processing of the DWI image and calculating an ADC (Apparent Diffusion Coefficient).
Wherein the calculation result of the lesion state determining unit is used for diagnosing the progress of the lesion, FLAIR-based lesion detection and status determination device. delete delete The method according to claim 5,
And a final diagnosis unit for diagnosing the progress of the lesion based on the ratio calculated by the lesion state determining unit, the position of the lesion, and the size of the lesion area. A computer-readable recording medium having recorded thereon a program for implementing the FLAIR-based lesion detection and status detection method of claim 1.

Images