KR102072476B1 - Apparatus and method for detecting tumor for image-guided surgery - Google Patents

Abstract

A tumor detection apparatus and method for image guided surgery are provided. Tumor detection method for image guided surgery according to an embodiment of the present invention comprises the steps of acquiring an image of a patient taken for surgery planning; Extracting a parameter for clustering image processing from the acquired image; Transmitting an image of a patient photographed during an image guided surgery in a video frame format; Receiving a video frame and searching for a tumor by clustering image processing applying a parameter for each frame; And outputting to indicate the location and size of the tumor detected in the current frame.

Description

Apparatus and method for detecting tumor for image-guided surgery

The present invention relates to a tumor detection apparatus and a method thereof, and more particularly, to a tumor detection apparatus and method for image-guided surgery to search for tumors in the image taken during image-guided surgery.

Recently, image-guided surgery using laparoscopic or endoscopy has been performed to increase the accuracy of tumor resection. Such image-guided surgery is a minimally invasive surgery, and the patient's recovery speed is fast and the surgical site can be optimized to operate accurately, which has attracted much attention. In particular, image-guided surgery can increase the survival rate and quality of life after surgery by minimizing damage to the body and increasing the accuracy and safety of the surgery.

At this time, the image-guided surgery uses the medical image taken before the operation as a reference auxiliary image and proceeds with the operation while checking the laparoscopy or endoscope image during the operation.

However, laparoscopic or endoscopic images do not provide accurate information such as the location or size of the entire lesion by providing only partial information of the lesion. Therefore, an improvement of a method of providing an image of the lesion during surgery is needed.

KR 10-1701043 B1

In order to solve the problems of the prior art as described above, an embodiment of the present invention is a tumor detection apparatus for image guided surgery that can display the image information on the surgical site of the patient in real time during the image-guided surgery and its To provide a method.

According to an aspect of the present invention for solving the above problems, a parameter extraction unit for obtaining an image of a patient photographed for surgery planning, and extracting a parameter for clustering image processing from the obtained image; And receiving the image of the patient taken during the image guided operation by the image capturing unit in a video frame format, searching for a tumor by the clustering image processing applying the parameter for each frame, and detecting the location of the tumor in the current frame. Provided is a tumor detection device for image-guided surgery comprising a; tumor search unit to output to the display unit to display the size.

In one embodiment, the parameter extracting unit is selected from the image of the tumor is observed in the image of the patient, detects the number of clusters in the selected image, and search for the cluster to which the tumor belongs from the detected population the number of the population and the tumor This cluster information may be transmitted to the tumor search unit.

In one embodiment, the parameter extractor calculates an SUM (Sum of Squares error) according to the change in the number of clusters in the selected image, converts the calculated error square sum into a graph, and outputs the result to the screen, wherein the output In the graph, it is possible to identify the number of clusters at the time when the distortion increases the fastest.

In one embodiment, the parameter extractor may apply the calculated number of clusters in the selected image, receive a position of a tumor, and search for a cluster to which the input tumor position belongs.

In one embodiment, the tumor search unit may search for a tumor by the k-means algorithm based on the parameter for each frame.

In one embodiment, the tumor search unit may display the position of the tumor detected in the current frame as an outline, and output to the display unit to calculate and display the size of the tumor.

In one embodiment, the acquired image of the patient may be any one of an MRI image, CT image, X-ray image and PET image.

In one embodiment, the explored tumor may be a brain tumor.

According to another aspect of the invention, the step of extracting the parameter image of the patient taken for surgery planning from the imaging unit; Extracting, by the parameter extracting unit, a parameter for clustering image processing from the obtained image; Transmitting, by the image capturing unit, an image of a patient photographed during an image guided surgery to a tumor search unit in a video frame format; Receiving the video frame by the tumor search unit and searching for a tumor by the clustering image processing to which the parameter is applied for each frame; And outputting the tumor search unit to a display unit to display the position and size of the tumor searched in the current frame.

In an embodiment, the extracting step may include selecting a region where a tumor is observed in an image of the patient; Detecting the number of clusters in the selected image; Searching for a population to which a tumor belongs among the detected populations; And transmitting the cluster number and cluster information to which the tumor belongs to the tumor search unit.

The detecting may include calculating a sum of squares error (SSE) according to a change in the number of clusters in the selected image; Converting the calculated error square sum into a graph and outputting the result to a screen; And identifying the number of clusters at the time when the distortion is most rapidly increased in the output graph.

In one embodiment, the step of searching for the cluster to which the tumor belongs may apply the calculated number of clusters in the selected image, search for a cluster to which the location of the input tumor belongs by receiving the location of the tumor.

In one embodiment, the step of searching for the tumor may search for the tumor by the k-means algorithm based on the parameter for each frame.

In one embodiment, the outputting step includes the step of contouring the location of the detected tumor in the current frame; And calculating and displaying the size of the tumor.

Tumor detection device and method for image guided surgery according to an embodiment of the present invention extracts the parameters for clustering from the image taken in the operation planning stage and search for tumor by applying the parameter to the image taken during the operation By displaying the location and size of the lesion in real time, it is possible to accurately track the surgical site and at the same time shorten the operation time, thereby minimizing errors that may occur during medical surgery.

In addition, the present invention by applying the parameters extracted in the operation planning stage in real time processing the image streaming during the operation, it is applied to the equipment utilizing a variety of medical images and diagnostic devices without restriction on the type of medical image or diagnostic device Because it can be applied to existing equipment, it is possible to build a system at low cost.

1 is a block diagram schematically showing a tumor detection apparatus for image guided surgery according to an embodiment of the present invention;
FIG. 2 is a diagram illustrating an image of a patient acquired in FIG. 1;
3 is a diagram illustrating clustering results in FIG. 1;
Figure 4 is a view showing the tumor location searched in Figure 1,
5 is a flowchart illustrating a tumor detection method for image guided surgery according to an embodiment of the present invention, and
FIG. 6 is a flowchart illustrating a parameter detection process in FIG. 5.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art may easily implement the present invention. As those skilled in the art would realize, the described embodiments may be modified in various different ways, all without departing from the spirit or scope of the present invention. In the drawings, parts irrelevant to the description are omitted to clearly describe the present invention, and like reference numerals designate like elements throughout the specification.

Hereinafter, a tumor detection apparatus for image guided surgery according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings. 1 is a block diagram schematically showing a tumor detection apparatus for image guided surgery according to an embodiment of the present invention.

Referring to FIG. 1, the tumor detection apparatus 100 for image guided surgery according to an embodiment of the present invention includes a parameter extractor 110 and a tumor searcher 120.

The present invention is to display the location and size of the tumor by transmitting the image data in real time streaming from the image capturing unit 10 to the tumor detection device 100 for image guided surgery. Here, in order to apply a clustering image processing algorithm for searching for a tumor to a streamed image, parameters such as cluster number and cluster information are required. The present invention finds an image of a patient who has previously selected such a parameter by a medical staff by applying a clustering image processing algorithm.

More specifically, the present invention distinguishes between the surgery planning stage and the surgery progress stage, and displays the position and size of the tumor from the image taken during the surgery by using parameters extracted from the patient's image acquired in the surgery planning stage. The tumor status during the surgery can be checked.

For example, in the surgery planning step, a parameter for clustering such as an optimal number of clusters and a cluster to which a tumor belongs is searched for from an image of a patient. Subsequently, in the surgery progress step, the tumor may be searched for in real time using parameter information obtained in the surgery planning step with respect to the streaming image transmitted from the image capturing unit 10 during surgery, and may be output to the display 130.

As a result, it is possible to easily identify the condition of the tumor to be operated in real time during the operation and implement the system at a low cost.

Here, the tumor detection apparatus 100 for image guided surgery may acquire an image from the image capturing unit 10. In this case, the image capturing unit 10 may include a magnetic resonance imaging apparatus (MRI), a computed tomography (CT) device, an X-ray device, and a positron emission tomography (PET) device. It may be either.

The parameter extractor 110 acquires an image of a patient photographed by the image capturing unit 10 to plan a surgery. Here, the image obtained from the image capturing unit 10 is for planning the image guided surgery. At this time, the surgery can be established by checking the position of the tumor from the patient's image before the surgery.

FIG. 2 is a diagram illustrating an image of a patient acquired in FIG. 1. 2 is an example of an MRI image of a pituitary tumor.

In addition, the parameter extractor 110 extracts a parameter for clustering image processing from an image obtained from the image capturing unit 10. Here, clustering image processing is for searching for a tumor from an image taken during the operation and is a process of clustering a specific portion in the image.

Such clustering is performed in the tumor search unit 120 to be described later. At this time, clustering is performed based on parameters such as the number of clusters and cluster information included in the corresponding image for real-time processing. To this end, the parameter extraction unit 110 may extract a parameter from the image of the patient in the surgery planning step.

More specifically, the parameter extractor 110 may first select a region where a tumor is observed in an image of a patient. In this case, the region where the tumor is observed may be selected by a medical staff.

In addition, the parameter extractor 110 may detect the optimal number of clusters in the image selected by the medical staff. For example, the parameter extractor 110 may extract the optimal number of clusters by an elbow algorithm.

In this case, the parameter extractor 110 may calculate a sum of squares error (SSE) according to a change in the number of clusters in the selected image. That is, while increasing the number k of clusters in the selected image from 1, the error sum sum SSE may be calculated.

In addition, the parameter extractor 110 may convert the calculated error square sum SSE into a graph. That is, the parameter extracting unit 110 may convert the graph into the graph of the error square sum SSE with respect to the number of clusters k and output the graph to the same screen as the display unit.

In addition, the parameter extractor 110 may identify the number k of clusters at the time when the distortion is most rapidly increased in the output graph. At this time, the parameter extraction unit 110 may detect the number k of the corresponding cluster as the optimal number of clusters.

In addition, the parameter extraction unit 110 may search for a cluster to which the tumor belongs among the detected clusters. In this case, the parameter extraction unit 110 may input the location of the tumor confirmed by the medical staff through the input unit 11.

That is, the parameter extractor 110 may apply the optimal number of clusters calculated from the selected image and search for a cluster to which the location of the input tumor belongs.

In addition, the parameter extractor 110 may transmit the extracted optimal number of colonies and cluster information to which the searched tumor belongs to the tumor search unit 120. Here, the parameter extractor 110 may transmit parameter information to the tumor searcher 120 when the tumor is searched by the tumor searcher 120 by capturing an image of the patient during the operation.

The tumor search unit 120 receives an image of a patient photographed during the image guided operation by the image photographing unit 10 in a video frame format. That is, the tumor search unit 120 receives the image of the patient from the image transmitted from the image capturing unit 10 in the video frame format.

In addition, the tumor search unit 120 searches for the tumor for each received frame by applying the parameters extracted by the parameter extraction unit 110. Here, the tumor to be searched may be a brain tumor such as a pituitary tumor or the like, but is not particularly limited to the type of tumor and may be a tumor of various forms.

More specifically, the tumor search unit 120 searches for tumors by clustering image processing for each frame. Here, clustering image processing may be performed by an algorithm capable of classifying a given data set (video streaming image) through a predetermined number of clusters. In one example, clustering image processing may be performed by a k-means algorithm.

That is, the tumor search unit 120 may search for tumors in the image by performing a k-means algorithm based on parameters such as the optimal number of clusters extracted by the parameter extractor 110 and the clusters to which the tumor belongs to each frame. . Here, the clustering image processing is described as being performed by the k-means algorithm, but not limited thereto, and various types of algorithms may be applied.

In this case, the tumor search unit 120 may search for the tumor with respect to a specific video frame transmitted from the image capturing unit 10, and if the tumor is not searched, the tumor search may be performed for the next frame.

In addition, the tumor search unit 120 outputs the position and size of the tumor detected in the current frame to the display unit 130. In this case, the tumor search unit 120 may display the position and size of the tumor and output the same to the display unit 130.

Here, the tumor search unit 120 may output to the display unit 130 to display the location of the tumor searched in the current frame as a contour stone, as shown in FIG. In addition, the tumor search unit 120 may calculate an area corresponding to the corresponding tumor in the current frame, calculate the size of the tumor, and output the same to the display 130.

FIG. 3 is a diagram showing clustering results in FIG. 1, and FIG. 4 is a diagram showing tumor locations searched for in FIG. 1.

3 shows a result of performing the k-means clustering algorithm by applying the tumor search unit 120 to the number of clusters 10 and group number 8. Here, a cluster belonging to a tumor may be searched at the center of the image.

Using this, the tumor search unit 120 may output the tumor to the display unit 130 to display the location of the tumor in a rectangular outline as shown in FIG. 4. In this case, the tumor search unit 120 may calculate a pixel corresponding to the rectangular outline and output the same to the display 130 to display the size of the tumor.

By such a configuration, the tumor detection device 100 for image guided surgery can accurately track the surgical site and at the same time shorten the operation time, thereby minimizing errors that may occur during medical surgery. In addition, the tumor detection device 100 for image-guided surgery can be applied to the equipment using a variety of medical imaging and diagnostic devices can be applied to the existing equipment can be built at a low cost system.

Hereinafter, a tumor detection method for image guided surgery according to an embodiment of the present invention will be described with reference to FIGS. 5 and 6.

5 is a flowchart illustrating a tumor detection method for image guided surgery according to an embodiment of the present invention.

The tumor detection method 100 for image-guided surgery includes the steps of taking an image of a patient (S201), extracting a parameter (S202), taking an image of a surgery in progress (S203), and searching for a tumor for each frame. (S204), and displaying the position and size of the tumor (S205).

In more detail, as shown in FIG. 5, first, an image capturing unit 10 captures an image of a patient for a surgical plan (step S201). Here, the image capturing unit 10 may be any one of a magnetic resonance imaging apparatus (MRI), a computerized tomography apparatus (CT), an X-ray apparatus, and a positron emission tomography apparatus (PET).

At this time, the parameter extraction unit 110 of the tumor detection device 100 for the image guided surgery to obtain the image of the patient taken from the image capturing unit 10. Here, the acquired image is an image provided from the image capturing unit 10 and may be any one of an MRI image, a CT image, an X-ray image, and a PET image. The acquired image may be an image of a brain, but is not limited thereto and may be an image of various body parts such as a heart.

As such, the image obtained from the image capturing unit 10 is for planning an image guided surgery. At this time, the surgery can be established by checking the position of the tumor from the patient's image before the surgery.

Next, the parameter extraction unit 110 extracts a parameter for clustering image processing from the image acquired from the image capturing unit 10 (step S202). Here, the clustering image processing is for searching for a tumor from an image taken during the operation and is a process of clustering a specific portion in the image.

Such clustering is performed in step S204 described later. At this time, clustering is performed based on parameters such as the number of clusters and cluster information included in the corresponding image for real-time processing. For this purpose, parameters can be extracted from the patient's image in the surgical planning stage. Extraction of such parameters will be described in more detail with reference to FIG. 6.

Next, the image capturing unit 10 captures an image of the patient during the image guided surgery (step S203). Here, the image capturing unit 10 transmits the captured image of the patient to the tumor search unit 120 of the tumor detection device 100 for image guided surgery in the form of a video frame.

Next, the tumor searching unit 120 receives the video frame from the image capturing unit 10 and searches for the tumor for each frame by applying the parameters extracted in step S202 (step S204). Here, the tumor to be searched may be a brain tumor such as a pituitary tumor or the like, but is not particularly limited to the type of tumor and may be a tumor of various forms.

More specifically, the tumor search unit 120 searches for tumors by clustering image processing for each frame. Here, clustering image processing may be performed by an algorithm capable of classifying a given data set (video streaming image) through a predetermined number of clusters. In one example, the clustering image processing may be a k-means algorithm.

That is, the tumor search unit 120 may search for tumors in the image by performing a k-means algorithm based on parameters such as the optimal number of clusters extracted in step S202 and the population to which the tumor belongs to each frame. Such clustering image processing is not limited to the k-means algorithm, and various types of algorithms may be applied.

In this case, the tumor search unit 120 may search for the tumor with respect to the video specific frame transmitted from the image capturing unit 10, and if the tumor is not searched, the tumor search may be performed for the next frame.

Next, the tumor search unit 120 displays the position and size of the tumor searched in the current frame (step S205). In this case, the tumor search unit 120 may display the position and size of the tumor and output the same to the display unit 130.

Here, the tumor search unit 120 may output to the display unit 130 to display the location of the tumor searched in the current frame as a contour stone, as shown in FIG. In addition, the tumor search unit 120 may calculate an area corresponding to the corresponding tumor in the current frame, calculate the size of the tumor, and output the same to the display 130.

FIG. 6 is a flowchart illustrating a parameter detection process in FIG. 5.

The parameter detection process 300 includes selecting an image in which a tumor is observed (S301), detecting an optimal number of clusters (S302), searching for a cluster to which a tumor belongs (S303), and selecting an optimal number of tumors and a tumor. Transmitting cluster information belonging to (304).

In more detail, as shown in FIG. 6, first, the parameter extraction unit 110 selects a region where a tumor is observed in an image of a patient (step S301). Here, the region where the tumor is observed may be selected by the medical staff.

Next, the parameter extraction unit 110 detects the optimal number of clusters from the image selected by the medical staff (step S302). For example, the parameter extractor 110 may extract the optimal number of clusters by an elbow algorithm.

In this case, the parameter extractor 110 may calculate a sum of squares error (SSE) according to a change in the number of clusters in the selected image. That is, while increasing the number k of clusters in the selected image from 1, the error sum sum SSE may be calculated.

Next, the parameter extractor 110 may convert the calculated error square sum SSE into a graph. That is, the parameter extracting unit 110 may convert the graph into the graph of the error square sum SSE with respect to the number of clusters k and output the graph to the same screen as the display unit.

Next, the parameter extractor 110 may identify the number k of clusters at the time when the distortion is increased fastest in the output graph. At this time, the parameter extraction unit 110 may detect the number k of the corresponding cluster as the optimal number of clusters.

Next, the parameter extraction unit 110 searches for a cluster to which the tumor belongs among the detected clusters (step S303). In this case, the parameter extraction unit 110 may input the location of the tumor confirmed by the medical staff through the input unit 11.

That is, the parameter extractor 110 may apply the optimal number of clusters calculated from the selected image and search for a cluster to which the location of the input tumor belongs.

Next, the parameter extraction unit 110 transmits the optimal population number and cluster information to which the tumor belongs to the tumor search unit 120 (step S304). Here, the parameter extractor 110 may transmit parameter information to the tumor searcher 120 when the tumor is searched by the tumor searcher 120 by capturing an image of the patient during the operation.

By this method, the present invention can accurately track the surgical site and at the same time shorten the operation time, thereby minimizing errors that may occur during medical surgery. In addition, the present invention can be applied to the equipment utilizing a variety of medical imaging and diagnostic devices can be applied to the existing equipment can be built a system at a low cost.

Such methods may be implemented by the tumor detection device 100 for image guided surgery as shown in FIG. 1, and in particular, may be implemented as a software program that performs these steps, in which case such a program They may be stored in a computer readable recording medium or transmitted by computer data signals combined with carriers in a transmission medium or communication network.

At this time, the computer readable recording medium includes all kinds of recording devices in which data readable by a computer system is stored, and for example, ROM, RAM, CD-ROM, DVD-ROM, DVD-RAM, magnetic tape, Floppy disks, hard disks, optical data storage devices, and the like.

Although one embodiment of the present invention has been described above, the spirit of the present invention is not limited to the embodiments presented herein, and those skilled in the art who understand the spirit of the present invention, within the scope of the same idea, the addition of components Other embodiments may be easily proposed by changing, deleting, adding, etc., but this will also fall within the scope of the present invention.

100: Tumor detection device for image guided surgery
110: parameter extraction unit 120: tumor search unit
130: display unit 10: video recording unit
11 input unit

Claims (16)

A parameter extracting unit obtaining an image of a patient photographed for a surgical plan and extracting a parameter for clustering image processing from the obtained image; And
Receive images of the patient taken during the image-guided surgery by the imaging unit in the form of a video frame, search for tumors by the clustering image processing applying the parameters for each frame, location and size of the tumors found in the current frame A tumor search unit for outputting to a display unit to display a;
Tumor detection device for image-guided surgery comprising a. The method of claim 1,
The parameter extracting unit selects an image in which the tumor is observed from the image of the patient, detects the number of clusters in the selected image, searches for a cluster to which the tumor belongs among the detected clusters, and displays the number of clusters and cluster information to which the tumor belongs. Tumor detection device for image guided surgery to transmit to the tumor search unit. The method of claim 2,
The parameter extractor calculates a sum of squares error (SSE) according to a change in the number of clusters in the selected image, converts the calculated sum of squares into a graph, outputs the result to a screen, and the distortion in the output graph. Tumor detection device for image guided surgery to identify the number of clusters of the fastest growing time point. The method of claim 3,
The parameter extractor applies the calculated number of clusters in the selected image, receives a tumor position, and detects a cluster to which the input tumor position belongs. The method of claim 1,
The tumor detection unit is a tumor detection device for image-guided surgery to search for tumors by the k-means algorithm based on the parameters for each frame. The method of claim 1,
The tumor detecting unit displays the location of the tumor detected in the current frame as an outline, and outputs to the display unit to calculate and display the size of the tumor tumor detection device for image guided surgery. The method of claim 1,
The acquired patient image is any one of MRI image, CT image, X-ray image and PET image of the tumor detection device for image guided surgery. The method of claim 1,
Tumor detection device for the image-guided surgery is the tumor is brain tumor. Obtaining, by the parameter extracting unit, an image of the patient photographed for the surgery plan from the image capturing unit;
Extracting, by the parameter extracting unit, a parameter for clustering image processing from the obtained image;
Transmitting, by the image capturing unit, an image of a patient photographed during an image guided surgery to a tumor search unit in a video frame format;
Receiving the video frame by the tumor search unit and searching for a tumor by the clustering image processing to which the parameter is applied for each frame; And
Outputting to the display unit the tumor search unit to display the position and size of the tumor detected in the current frame;
Tumor detection method for image-guided surgery comprising a. The method of claim 9,
The extracting step,
Selecting an area in which the tumor is observed in the image of the patient;
Detecting the number of clusters in the selected image;
Searching for a cluster to which a tumor belongs among the detected clusters; And
And transmitting the cluster number and the cluster information to which the tumor belongs to the tumor search unit. The method of claim 10,
The detecting step,
Calculating a sum of squares error (SSE) according to a change in the number of clusters in the selected image;
Converting the calculated error square sum into a graph and outputting the result to a screen; And
And identifying the number of clusters at the time when the distortion is most rapidly increased in the output graph. The method of claim 11,
The step of searching for a cluster to which the tumor belongs comprises applying the calculated number of clusters in the selected image, and receiving the location of the tumor to detect the cluster to which the location of the input tumor belongs to the tumor detection method for surgery. The method of claim 9,
The detecting of the tumor may include detecting a tumor by a k-means algorithm based on the parameter for each frame. The method of claim 9,
The output step is
Contouring the location of the detected tumor in the current frame; And
A tumor detection method for image-guided surgery comprising calculating and displaying the size of the tumor. The method of claim 9,
The acquired patient image is any one of the MRI image, CT image, X-ray image and PET image tumor detection method for image-guided surgery. The method of claim 9,
The tumor is a tumor detection method for image-guided surgery is a brain tumor.

Images