CN111821032B - Medical Image Alignment Method - Google Patents

Abstract

A medical image registration method for image registering an ultrasound image and one of a plurality of computed tomography images associated with a plurality of different portions of a patient's bone, the ultrasound image being associated with one of the plurality of different portions of the patient's bone and being implemented by a computer device, comprising: (A) obtaining an ultrasonic target image of the ultrasonic image; (B) for each computed tomography image, obtaining a to-be-compared scanned image of the computed tomography image; (C) for each scanned image to be compared, obtaining a similarity between the scanned image to be compared and the ultrasonic target image; (D) obtaining a corresponding target scanning image with the highest similarity from the scanning images to be compared; (E) and carrying out image alignment on the ultrasonic target image and the target scanning image. The surgeon may be assisted in real time during the procedure to determine the location to be operated on.

Description

Medical Image Alignment Method

technical field

The invention relates to an image alignment method, in particular to an image alignment method suitable for medical images.

Background technique

In the earliest days, the surgeon could only rely on all the relevant information of the patient before the operation (for example, computed tomography images) to determine the location to be operated on. For example: nailing) makes the patient's body move a little, causing the position of the subsequent operation to be shifted, which in turn affects the correctness of the operation.

However, at present, when the surgeon wants to accurately determine the location to be operated on during the operation, he has to take a computed tomography during the operation to determine it. However, this method cannot detect the error caused by the patient's movement in real time, which makes the patient more difficult. Absorb excessive radiation doses.

In view of this, how to provide a method for assisting the surgeon to determine the position to be operated on in real time during the operation without requiring the patient to absorb too much radiation dose is the primary issue to be solved by this creation. .

SUMMARY OF THE INVENTION

The object of the present invention is to provide a method for assisting the surgeon to determine the position to be operated on in real time during the operation without requiring the patient to absorb too much radiation dose.

The medical image alignment method of the present invention is suitable for image alignment of an ultrasonic image with one of a plurality of computed tomography images respectively related to a plurality of different parts of a patient's bone, and the ultrasonic image is a One of the different parts of the patient's skeleton and implemented by a computer device, the medical image alignment method includes a step (A), a step (B), a step ( C), a step (D), and a step (E).

In the step (A), an ultrasonic target image corresponding to a region of interest of the ultrasonic image is obtained by the computer device.

In the step (B), for each computed tomography image, a scan image to be compared corresponding to a region of interest of the computed tomography image is obtained by the computer device.

In the step (C), for each scanned image to be compared, a similarity between the scanned image to be compared and the ultrasonic target image is obtained by the computer device.

The step (D) is to obtain a corresponding target scanned image with the highest similarity from the scanned images to be compared by the computer device.

The step (E) is to perform image alignment between the ultrasonic target image and the target scan image by the computer device.

The medical image alignment method of the present invention further comprises the following steps between the step (B) and the step (C):

(F) determining, by the computer device, whether the size of the ultrasound target image is the same as the size of any scanned image to be compared; and

(G) When the computer device determines that the size of the ultrasonic target image is different from the size of any scanned image to be compared, correct the ultrasonic target image by the computer device according to the size of any scanned image to be compared, so as to Make the size of the corrected ultrasonic target image the same as the size of any scanned image to be compared;

Wherein, in the step (C), the computer device obtains a similarity corresponding to each scanned image to be compared and the corrected ultrasonic target image, and in the step (E), the computer device will Image alignment is performed between the corrected ultrasonic target image and the target scanned image.

The medical image alignment method of the present invention further comprises the following steps between the step (B) and the step (C):

(H) For each scanned image to be compared, use the computer device to perform upper contour feature extraction on the object in the scanned image to be compared, so that the scanned image to be compared after extraction only retains the content contained in the scanned image to be compared the upper outline of the object; and

Wherein, in the step (C), the computer device obtains a similarity corresponding to each extracted scanned image to be compared and the ultrasonic target image, and in the step (D), the computer device is From the scanned images to be compared after extraction, the target scanned image corresponding to the highest similarity is obtained.

In the medical image alignment method of the present invention, in the step (H), the computer device retains the pixel value of the pixel whose first pixel value is not zero in each row of the scanned image to be compared and converts it to The pixel values of other pixels are set to zero to extract the upper contour feature of the object in the scanned image to be compared.

In the medical image alignment method of the present invention, in the step (C), the computer device obtains the similarity between the scanned image to be compared and the ultrasonic target image by using a maximum interactive information method.

The medical image alignment method of the present invention:

In the step (C), the computer device not only obtains the similarity between the scanned image to be compared and the ultrasonic target image, but also obtains the difference between the scanned image to be compared and the ultrasonic target image by using the maximum interactive information method. a set of geometric transformation parameters of the mapping relationship; and

In the step (E), the computer device performs image alignment of the ultrasonic target image and the target scanned image according to a geometric transformation parameter set corresponding to the target scanned image and the ultrasonic target image.

In the medical image alignment method of the present invention, in the step (H), the computer device also sets the pixel value of all non-zero pixels retained in the scanned image to be compared to 255.

The beneficial effect of the present invention is that: the ultrasonic target image corresponding to the ultrasonic image and the to-be-compared scanned image corresponding to each computed tomography image are analyzed for similarity by the computer device, and the highest corresponding image is obtained. The target scan image with the similarity, and finally the image alignment of the ultrasonic target image and the target scan image can assist the surgeon to determine in real time during the operation without requiring the patient to absorb too much radiation dose. The location to be operated on.

Description of drawings

Other features and effects of the present invention will be clearly presented in the embodiments with reference to the drawings, wherein:

FIG. 1 is a block diagram illustrating the electrical connection between a computer device and an ultrasonic image capturing device for implementing an embodiment of the medical image alignment method of the present invention;

2 is a schematic diagram illustrating one of a plurality of computed tomography images stored by the computer device;

3 is a schematic diagram illustrating an ultrasonic image captured by the ultrasonic image capturing device;

4 is a flowchart illustrating steps 51, 52, 53 and 54 of the embodiment of the medical image alignment method of the present invention;

Fig. 5 is a flow chart, which continues Fig. 4 to illustrate step 55, step 56, step 57 and step 58 of the embodiment of the medical image alignment method of the present invention;

Fig. 6 is a flow chart illustrating step 51', step 52', step 53' and step 54' of another embodiment of the medical image alignment method of the present invention;

7 is a schematic diagram illustrating an ultrasonic target image corresponding to the ultrasonic image;

8 is a schematic diagram illustrating a scanned image to be compared corresponding to one of the computed tomography images;

FIG. 9 is a schematic diagram illustrating the image obtained by extracting the upper contour feature of the scanned image to be compared; and

FIG. 10 is a schematic diagram illustrating the result of image alignment between the ultrasonic target image and a target scanning image.

Detailed ways

Before the present invention is described in detail, it should be noted that in the following description, similar components are designated by the same reference numerals.

Referring to FIG. 1 , a computer device 1 according to an embodiment of the medical image alignment method of the present invention is implemented, and an ultrasonic image capturing device 2 is electrically connected to the computer device 1 .

The computer device 1 includes a storage module 11 , an input module 12 , a display module 13 , and a processing module 14 electrically connected to the storage module 11 , the input module 12 and the display module 13 .

The storage module 11 stores a plurality of Computed Tomography (CT) images respectively related to a plurality of different parts of a patient's skeleton. It is worth noting that, for example, the patient's skeleton is a vertebra, and the vertebra includes a plurality of vertebral segments, each vertebral segment can be divided into a plurality of segments, and each computed tomography image is associated with a corresponding segment (see Figure 2). In addition, the spinous process 21 in the segmented portion corresponding to the segment of the vertebra in each computed tomography image is superior in the computed tomographic image, and the segment corresponding to the segment of the vertebra in each computed tomography image The lamina 22 in the segment portion is below in this computed tomography image.

The ultrasonic image capturing device 2 is used to capture an ultrasonic image (Ultrasound) related to one of a plurality of different parts of the patient's bone, and transmit the ultrasonic image to the computer device 1 . It is worth noting that, following the above example, the ultrasonic image capturing device 2 captures the ultrasonic image (see FIG. 3 ) related to a segment of the spine of the patient when the patient is prone to Make the spinous process 31 in the segmented portion corresponding to the segment of the vertebra in the ultrasound image above in the ultrasound image, and the lamina 32 in the segmented portion corresponding to the segment of the spine in the ultrasound image. Bottom in the ultrasound image. Thus, the segmented portion of the spine presented in the ultrasound image faces the same orientation as the segmented portion of the spine presented in the computed tomography image.

In this embodiment, the implementation form of the computer device 1 is, for example, a personal computer, a server or a cloud host, but not limited thereto.

In this embodiment, the implementation of the ultrasonic image capturing device 2 is, for example, an ultrasonic diagnostic instrument, but not limited thereto.

Referring to FIGS. 4 and 5 , the operation details of the computer device 1 and the components of the ultrasonic image capturing device 2 will be described below by using the embodiment of the medical image alignment method of the present invention. The medical image alignment method of the present invention includes a step 51. A step 52, a step 53, a step 54, a step 55, a step 56, a step 57, and a step 58.

In step 51 , after receiving the ultrasonic image, the computer device 1 obtains an ultrasonic target image corresponding to a region of interest in the ultrasonic image (see FIG. 7 ). It is worth noting that the computer device 1 responds to the operation of the surgeon and generates an input signal, and obtains the ultrasonic target image corresponding to the ultrasonic image according to the input signal. It is worth noting that, in other embodiments, the existing image recognition method can also be used to obtain the ultrasonic target image corresponding to the ultrasonic image.

In step 52 , for each computed tomography image, the computer device 1 obtains a scan image to be compared corresponding to a region of interest in the computed tomography image (see FIG. 8 ). It is worth noting that, for each computed tomography image, the computer device 1 responds to the operation of the surgeon and generates another input signal, and according to the other input signal, obtains the standby ratio corresponding to the computed tomography image. on scanned images. It is worth noting that, in other embodiments, the existing image recognition method can also be used to obtain the scanned image to be compared corresponding to the computed tomography image.

In step 53, for each scanned image to be compared, the computer device 1 extracts the upper contour feature of the object in the scanned image to be compared, so that the scanned image to be compared after extraction only retains the features contained in the scanned image to be compared the upper contour of the object (see Figure 9). It is worth noting that, in this embodiment, the computer device 1 first removes the noise in the scanned image to be compared, and then retains the first pixel in each row of the scanned image to be compared The pixel value of the pixel whose value is not zero, and the pixel value of other pixels is set to zero, so as to extract the upper contour feature of the object in the scanned image to be compared. In other embodiments, the computer device 1 can also set the pixel value of all non-zero pixels retained in the scanned image to be compared to 255, so as to extract the object in the scanned image to be compared The upper contour feature of .

Referring to FIG. 6 , in particular, in another embodiment, which is similar to the description of the above-mentioned embodiment, the only difference is that when the step 51 ′ related to the ultrasound image is performed, the step 51 ′ related to the computed tomography image can be compared with that of the computer tomography image. 52' and the step 53' are executed simultaneously, so as to complete the medical image alignment method at a faster speed.

It is worth noting that, when the similarity analysis is performed on the scanned image from which the upper contour feature is extracted in step 53, the lower contour feature does not overlap with the noise of the ultrasonic target image. , which is better than directly using the to-be-compared scanned image without extracting the upper contour feature for similarity analysis, which can further improve the accuracy of image alignment.

In step 54, the computer device 1 determines whether the size of the ultrasonic target image is the same as that of any scanned image to be compared. When the computer device 1 determines that the size of the ultrasonic target image is the same as the size of any of the scanned images to be compared, the process proceeds to step 56; when the computer device 1 determines that the size of the ultrasonic target image is not the same as any of the to-be-compared scanned images When the size of the scanned image is compared, the process proceeds to step 55 .

In step 55, the computer device 1 corrects the ultrasonic target image according to the size of any scanned image to be compared, so that the size of the corrected ultrasonic target image is the same as the size of any scanned image to be compared.

In step 56, for each scanned image to be compared after extraction, the computer device 1 obtains the scanned image to be compared after extraction and the corrected ultrasonic target image (if step 55 is not required, then using a similarity of the ultrasound target image). It is worth noting that, in this embodiment, the computer device 1 uses a maximum interactive information method (Mutual Information, MI) to obtain the extracted scan image to be compared and the corrected ultrasonic target image. Similarity, and a geometric transformation parameter set of the mapping relationship between the scanned image to be compared and the corrected ultrasonic target image after extraction. In other embodiments, the computer device 1 can also use an iterative closest point algorithm (eg, Iterative Closest Points, ICP) to obtain the similarity between the extracted scanned image to be compared and the corrected ultrasonic target image Spend.

In step 57 , the computer device 1 obtains a corresponding target scanned image with the highest similarity from the scanned images to be compared after extraction.

In step 58, the computer device 1 performs image alignment on the corrected ultrasonic target image and the target scanned image, and performs image alignment on the corrected ultrasonic target image and the target scanned image (see Fig. 10) is shown in the display module 13 for the surgeon to refer to during the operation to determine the position to be operated on. It is worth noting that, in this embodiment, the computer device 1 is based on the geometric transformation parameter set corresponding to the target scan image after calibration (if the step 55 is not required, the ultrasonic target image is used) , performing image alignment on the corrected ultrasonic target image and the target scan image. In other embodiments, the computer device 1 can also use an image alignment algorithm (eg, B-spline) to perform image alignment.

To sum up, in the medical image alignment method of the present invention, the computer device 1 uses the ultrasonic target image corresponding to the ultrasonic image and the scanned image to be compared corresponding to each computed tomography image, and utilizes the maximum In the interactive information method, from the scanned images to be compared after extraction, the target scanned image corresponding to the highest similarity is obtained, and the corrected ultrasonic target image and the target scanned image are image-aligned, and then In addition, it can assist the surgeon to determine the position to be operated on in real time during the operation without the need for the patient to absorb too much radiation dose. Therefore, the object of the present invention can be surely achieved.

However, the above are only examples of the present invention, and should not limit the scope of implementation of the present invention. Any simple equivalent changes and modifications made according to the claims of the present invention and the contents of the description are still within the scope of the present invention. within the scope of the invention patent.

Claims (6)

1. A medical image alignment method, which is suitable for performing image alignment on an ultrasonic image with one of a plurality of computed tomography images respectively related to a plurality of different parts of a patient's bone, the ultrasonic image It is related to one of the plurality of different parts of the patient's skeleton, and is implemented by a computer device, wherein the medical image alignment method comprises the following steps: (A) obtaining an ultrasonic target image corresponding to a region of interest of the ultrasonic image by the computer device; (B) for each computed tomography image, obtain a scan image to be compared corresponding to a region of interest of the computed tomography image by the computer device; (C) For each scanned image to be compared, use the computer device to extract the upper contour feature of the object in the scanned image to be compared, so that the scanned image to be compared after extraction only retains the features contained in it. the upper outline of the object; (D) for each scan image to be compared, obtain a similarity between the scan image to be compared and the ultrasonic target image by the computer device; (E) by the computer device, from the scanned images to be compared, obtain a corresponding target scanned image with the highest similarity; and (F) performing image alignment on the ultrasound target image and the target scan image by the computer device; Wherein, in the step (D), the computer device obtains the similarity corresponding to each scanned image to be compared after extraction and the ultrasonic target image, and in the step (E), the computer device automatically After the extraction, among the scanned images to be compared, the target scanned image corresponding to the highest similarity is obtained. 2. The medical image alignment method according to claim 1, wherein between the step (C) and the step (D), the method further comprises the following steps: (G) determining, by the computer device, whether the size of the ultrasound target image is the same as the size of any scanned image to be compared; and (H) when the computer device determines that the size of the ultrasonic target image is different from the size of any scanned image to be compared, correct the ultrasonic target image by the computer device according to the size of any scanned image to be compared, so as to Make the size of the corrected ultrasonic target image the same as the size of any scanned image to be compared; Wherein, in the step (D), the computer device obtains a similarity corresponding to each extracted scanned image to be compared and the corrected ultrasonic target image, and in the step (F), the computer The device performs image alignment between the corrected ultrasonic target image and the target scanned image. 3. The medical image alignment method according to claim 1, wherein in the step (C), the computer device retains the first pixel value in each row of the scanned image to be compared The pixel values of non-zero pixels are set to zero and the pixel values of other pixels are set to zero to extract the upper contour feature of the object in the scanned image to be compared. 4 . The medical image alignment method according to claim 1 , wherein in the step (D), the computer device uses a maximum interactive information method to obtain the extracted scanned image to be compared and the extracted image. 5 . The similarity of the ultrasound target image. 5. The medical image alignment method according to claim 4, wherein: In the step (D), the computer device uses the maximum interactive information method to obtain not only the similarity between the extracted scanned image to be compared and the ultrasonic target image, but also to obtain the extracted scanned image to be compared with the target image. a geometric transformation parameter set of the mapping relationship between the ultrasound target images; and In the step (F), the computer device performs image alignment of the ultrasonic target image and the target scanned image according to a geometric transformation parameter set corresponding to the target scanned image and the ultrasonic target image. 6 . The medical image alignment method according to claim 3 , wherein in the step (C), the computer device also includes all pixels that are not zero in the scanned image to be compared. 7 . The value is set to 255.

Images