DE102010019645A1 - A method of generating 3D medical image data - Google Patents

Abstract

A method for generating medical 3D image data (34) of a treatment target (14) in a patient (8) has the following steps: a) at a first point in time (T1) before or during a treatment of the patient (8) a 3-D image data set (24) of the treatment target (14) is generated, b) at a later second point in time (T2) during the treatment, at least one 2D fluoroscopic image (26) of the treatment target (14), which is a in the 2D fluoroscopic image (26) contains at least partially reproducible object (22) of known geometry, c) between the 3D image data set (24) and the 2D fluoroscopic image (26) a transformation rule (32) is determined, d) on the basis of the The object (22) depicted in the 2D fluoroscopic image (26) is determined its spatial position in the coordinate system of the imaging device (16), and image information representing the position based on the transformation rule (32) is determined correctly with the 3D image ld data set (24) merged into 3D image data (34), e) if a relative position (R) between treatment target (14) and object (22) changes, steps b) to d) are repeated.

Description

The invention relates to a method for generating 3D medical image data.

Medical image data is generated by a patient by mapping at least a portion of the patient's body in the image data, i. H. is illuminated. Image data are z. B. a two-dimensional (2D) X-ray image or a three-dimensional (3D) magnetic resonance image in the form of 3D image data or a 3D image data set. If a medical treatment is performed on the patient, the treatment is usually local; z. B. is to be placed as part of an operation as a treatment on the hip of the patient implant. The hip region of the patient and their environment - also partially outside the patient - then represents the treatment goal in the patient. There, relevant medical measures, such. B. the provision of instruments or implants or the actual intervention in the body of the patient made. The image data are usually generated by the treatment goal, since in particular image information from the patient and implants or tools is desired here. It is often important to know the relative position of tools and patient anatomy. This serves to place the medical tools correctly on the patient to z. B. to install a hole or implant in the patient's place. The anatomy of the patient is generally described completely by 3D image data of the treatment goal.

The correct placement of surgical instruments or implants in open or minimally invasive surgery on patients is still a not completely solved problem. B. by minimizing the trauma intervention the direct visibility on the operation goal only very limited. The direct view of the operation goal takes place - z. B. in the so-called buttonhole surgery - only by an endoscope or its video camera. A fluoroscopy of the patient - usually X-ray imaging - therefore often takes place.

A pure 2D imaging is not sufficient in many cases, z. B. to verify a correct implant position, especially in degenerative or abnormal anatomy of the patient.

It is known to use 2D or 3D X-ray imaging, ultrasound or endoscopy to generate medical image data.

Also known is the use of 2D or 3D navigation with or without imaging to place an implant or instrument at a desired position in or on the patient. Known here is z. Eg the system "CAPPA C-Nav" of the company "CAS innovations".

Also known is an image-based 2D method using additional information. As additional information z. B. geometric information from or about tools, ie instruments or implants used. Such a process is known from products of Surgix.

The object of the present invention is to specify an improved method for generating 3D medical image data.

The object is achieved by a method for generating medical 3D image data of a treatment target in a patient, with the following steps:
In a step a), a 3D image data set of the treatment target is generated at a first time before or during the treatment of a patient. Thus, a pre- or intra-operative 3D image of the treatment or surgical area is made. At a later second point in time, in a step b), during the treatment with the aid of an imaging device, at least one-as a rule also several-2D fluoroscopic image of the treatment target is generated. In other words, therefore real-time or real-time 2D fluoroscopy or projection images are made. This is called fluoroscopy in the case of X-rays. The treatment goal here includes a 2D fluoroscopic image ( 26 ) at least partially imageable object ( 22 ) known geometry.

In a step c), a transformation rule is determined between the 3D image data record and the 2D fluoroscopic image. Thus, the 2D / 3D transformation between the 3D image data set and the real-time 2D projection images is calculated.

In a step d), the spatial position in the coordinate system of the imaging apparatus (FIG. 2) is then determined on the basis of the object imaged in the 2D fluoroscopic image. 16 ) certainly. An image information representing the position is then combined in a location-correct manner with the 3D image data record into 3D image data on the basis of the transformation instruction.

The known geometry makes it possible to determine the 3D location information from the image of the object, since the relative geometric position of characteristic points of the object in the 2D image is recognizable.

In a step e), the steps b) to d) are repeated when the relative position between the treatment target and the object changes. In other words, the above image registration between 2D Radiographic image and 3D image data set each recalculated when the geometries change between the preoperative 3D image and the imaging device, ie between 2D and 3D. This could be z. B. by an object movement, so the movement of the patient or the operation goal happen. Possible movements in or on the patient can in this case in a known manner z. B. by a glued on the patient marker and its location tracking, z. B. in successive recorded 2D X-ray images are detected.

The problem according to the invention can be achieved by the o. G. Technology combined with the o. G. Solve the process very easily. The method according to the invention shows a simple method of obtaining positional information of the object via the 2D fluoroscopic image, for B. the visible in a 2D X-ray image current implant or instrument position. By the 2D-3D transformation then the 3D position of the implant or instrument is known and can, for. B. in relation to the preoperative or intraoperative 3D patient anatomy are made available to the user. The user is thus offered 3D information.

Thus, intra-operatively there is a simple possibility of how to display the positions and orientations of implants and tools in the 3D data set without additional navigation devices or to present them in the preoperative 3D image. Whereby at any time an update of the location information, so an update, by a new 2D fluoroscopic image according to step b) and their o. G. Recycling can be carried out.

The use of a single 2D fluoroscopic image is sufficient z. B. then for the 3D position determination of the object, if it has enough reference points or characteristic in the 2D image locatable points in a known geometric relationship to each other, which allow a 3D location based on a 2D image. Methods for this are, for. B. from image processing or mono-camera navigation systems known.

The image information can be z. For example, an abstract mark such as a crosshair, a symbol of the object, an outline, or the like.

In a preferred embodiment of the method, however, an image of the object is combined with the 3D image data record in the correct location in step d) as image information on the basis of the transformation rule.

If the geometry of the object itself is known in its entirety or in terms of essential structural parts, this information can, for B. can be used together with the known projection geometry of the imaging device. The projection geometry can - if not known - also be determined by known methods. Thus, 3D spatial coordinates of the object or characteristic points of it can actually be determined in the coordinate system of the 3D image data set and thus in the image data. It is then possible, even abstract information about the object, for. B. a marker cross for an implant bore or a virtual projection line of a drill center axis in the 3D image data represent.

In preferred embodiments, the object used is an implant and / or an instrument of known geometry.

If, in an alternative embodiment, the geometry of the instrument and / or implant itself is not known, this can be equipped with markers detectable in the 2D fluoroscopy image, in which case the geometry of the markers or their relative arrangement must be known.

In other words, z. B. on the tools or implants, z. B. attached to K-wires, markers or marker arrangements of known geometries or integrated. These can then be used to calculate the 3D geometries in the coordinate system of the 2D fluoroscopic images, ie projection images. Implants can be z. B. by their 3D design geometries known from design drawings directly for the calculation of 3D positions and orientations in the coordinate system of the 2D projection recording draw.

Thus, finally, the position and orientation of the tools or implants in the coordinate system of the 2D projection recording, and thus in the coordinate system of the imaging device, for. B. an X-ray C-arm, known. Thus, due to the known transformation between the 2D projection imaging apparatus and the 3D image, that is, the actual position of the implant or of the tool can be transferred to the preoperative 3D image or the 3D data set and finally to the generated image data.

In a preferred embodiment, a 3D image containing planning data is used as the 3D image. The inventive approach can be combined with planning data, the z. B. have been obtained from the preoperative 3D data set.

In addition, with such a method then z. B. planned access routes and / or target and Entrypunkte in the second time, so currently obtained 2D-projection images and serve the surgeon as Clue for the procedure. Alternatively or additionally, for. B. the deviation of the current position of an implant from a preoperatively planned implant position calculated and made available to the surgeon.

If the 3D data set, that is to say the 3D image, is acquired intraoperatively at the first time by means of a 3D C-arm, then under certain circumstances the 2D / 3D image registration according to step c) can be performed very simply. The transformation rule is then given by the acquisition geometry or its determination can strictly speaking be omitted if the 2D fluoroscopic images are obtained in a recording geometry known from the 3D image data set.

For a further description of the invention reference is made to the embodiments of the drawing. It shows, in a schematic outline sketch:

1 the treatment of a patient and the generation of medical image data.

1 shows a medical workplace 2 which is essentially an imaging system 4 , here an X-ray system and a patient table 6 includes. On the patient table 6 is a patient 8th stored on which as medical treatment an intervention on the spine 10 to be carried out. A thoracic vertebra 12 and its immediate environment is therefore the target area for the treatment, ie the treatment goal 14 In other words, this is the patient's region of interest (ROI) 8th from which imaging in the form of medical image data is to take place.

The imaging system 4 includes as an imaging device 16 an X-ray C-arm which is capable of producing both 2D and 3D image data. Corresponding necessary for the imaging, as well as necessary for carrying out the method according to the invention computation, processing or other data processing steps found in a computing unit 18 of the imaging system 4 instead of. The generated and other image data are displayed on screens 20 of the imaging system 4 output.

The medical measure is carried out with a needle as a medical instrument or object 22 ,

At a first time T1 is taken by the patient 8th or the treatment goal 14 with the help of the imaging device 16 a 3D image data set 24 generated in the form of an X-ray, which is the thoracic vertebra to be treated 12 three-dimensional images or the corresponding in the treatment goal 14 lying part of the spine 10 shows.

In an alternative embodiment, the 3D image data set is 24 already before the start of treatment on the patient 8th at time T1, z. B. generated with a magnetic resonance tomograph.

During the treatment, at least one 2D fluoroscopy image will now be at a later time T2, ie after the time T1 26 from the treatment goal 14 , or z. B. multiple images in different rotational positions of the X-ray C-arm of the imaging device 16 ie from different points of view on the treatment goal 14 , created. The at least one 2D fluoroscopic image 26 allowed in a coordinate system 28 of the imaging device 16 the spatial determination of the location of the 2D fluoroscopic image 26 imaged objects, in particular the thoracic vertebra 12 and the object 22 ,

The arithmetic unit 18 determined in a calculation step 30 a transformation rule 32 , like - Within the coordinate system 28 - the 2D fluoroscopic image 26 or the objects imaged therein with the 3D image data set 24 are related in the right place.

In the 3D image data set 24 Now an image information is displayed showing the position of the object 22 represents. In the example, this is a real image - in an alternative embodiment also at least a part - of the object 22 , This results in 3D image data 34 ,

In an alternative embodiment of the method is not the image of the object 22 itself, but only an abstract symbol or image 36 of which, namely in the form of 1 shown dashed line shown. The line is a thought image of the object 22 in the sense that this represents the course of the needle when it is advanced in the currently held direction in the direction of its central longitudinal axis in the patient.

In an alternative embodiment, the object 22 as an instrument also a series of with the imaging device 16 detectable, here radiopaque markers 38 in particular in the 2D fluoroscopic images 26 to be imaged with. In this embodiment, in the 2D Druchleuchtungsbild 26 pictured markers 38 the projection geometry of the imaging device 16 determined. Also, the location of the object 22 in the coordinate system of the imaging device 16 not based on the image of the object 22 itself, but the image of the markers 38 determined. The geometry of the object 22 then does not need to be known.

In an alternative embodiment, the object is 22 a dashed line, to be inserted into the patient as part of the treatment implant. This has been made in advance in the 3D image data set 24 in the form of planning data 40 configured, ie its planned position is determined. In the picture data 34 in this case, both in the 2D fluoroscopic images 26 pictured actual implant 22 as well as the position of the implant 22 in the form of planning data 40 displayed simultaneously, causing the viewer of the image data 34 can check a deviation between the planned and actual situation.

Changes a relative position R between fluoroscopic device 16 and treatment goal 14 , T2 new 2D fluoroscopic images will be taken at a later time T2 26 recorded and with these, as explained above, new 3D image data 34 created or refreshed with changed image content.

The relative position R changes z. B. by the movement of the thoracic vertebra 12 , the entire patient 8th or the object 22 ,

LIST OF REFERENCE NUMBERS

2

Workplace

4

imaging system

6

patient table

8th

patient

10

spinal column

12

thoracic vertebra

14

treatment goal

16

imaging device

18

computer unit

20

screen

22

object

24

3D image data set

26

2D X-ray image

28

coordinate system

30

calculation step

32

transformation rule

34

3D image data

36

image

38

marker

40

planning data

T _1,2

time

R

relative position

Claims (5)

Method for generating 3D medical image data ( 34 ) of a treatment goal ( 14 ) in a patient ( 8th ), comprising the following steps: a) at a first time (T1) before or during a treatment of the patient ( 8th ), a 3D image data set ( 24 ) of the treatment goal ( 14 b) at a later, second point in time (T2) during treatment is determined by means of an imaging device ( 16 ) at least one 2D fluoroscopic image ( 26 ) of the treatment goal ( 14 ), the one in the 2D fluoroscopic image ( 26 ) at least partially imageable object ( 22 ) of known geometry, generated, c) between the 3D image data set ( 24 ) and the 2D fluoroscopic image ( 26 ), a transformation rule ( 32 ), d) on the basis of the 2D fluoroscopic image ( 26 ) imaged object ( 22 ) its spatial position in the coordinate system of the imaging device ( 16 ), and an image information representing the position based on the transformation rule ( 32 ) in the correct position with the 3D image data set ( 24 ) to 3D image data ( 34 e) when changing a relative position (R) between the treatment objective ( 14 ) and object ( 22 ), steps b) to d) are repeated. Method according to Claim 1, in which, in step d), as image information, an image ( 36 ) of the object ( 22 ) with the 3D image data set ( 24 ) is merged. Method according to one of Claims 1 to 2, in which as object ( 22 ) an implant and / or instrument of known geometry is used. Method according to one of Claims 1 to 2, in which as object ( 22 ) with markers ( 38 ) known geometry equipped implant and / or instrument is used. Method according to one of the preceding claims, in which as a 3D image data record ( 24 ) a planning data ( 40 ) containing 3D image data set ( 24 ) is used.

Images