Method and system for creating a roadmap for medical workflow
By combining multidimensional image dataset and pilot tone signal technology, the problem of real-time motion analysis roadmap in MRI technology is solved, and a dynamic roadmap with high accuracy is achieved, which improves the accuracy and accuracy in the MRI intervention process.
Patent Information
- Application Number
- CN202010999673.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2019-09-25
- Filing Date
- 2020-09-22
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2040-09-22
AI Technical Summary
In the existing MRI technology, the real-time motion analysis roadmap has not been effectively solved, which makes it difficult to provide a highly accurate dynamic roadmap, especially when aiming. There is a problem of dynamic mismatch in small lesions.
By combining multidimensional image datasets and pilot tone signal technology, a real-time motion analysis roadmap is provided, and the pilot tone signal is used to determine the dimensional coordinates of the organ's motion state, select the corresponding image dataset to create a dynamic roadmap to realize the movement of the organ Real-time tracking and display of status.
It realizes the creation of highly accurate dynamic roadmaps in MRI technology, which can track the movement status of the organ in real time, improves the accuracy and accuracy during the intervention process, avoids the limitations of static 3D roadmaps, and is suitable for various contrast and Matching of physiological states.
Smart Images

Figure CN112545482B_ABST
Abstract
Description
Technical Field
[0001] The present invention describes a method and a system for creating a (dynamic) roadmap of a medical workflow, in particular in the field of magnetic resonance imaging (MRI) technology. Background Art
[0002] MRI systems are used for medical examinations to record (image) data of the object under examination by exciting nuclear spins aligned in a strong basic magnetic field. The precession or relaxation of the spins from this excited state to a state with less energy generates an alternating magnetic field ("RF signal") received via an RF antenna in response. The RF frequency is strongly dependent on the basic magnetic field.
[0003] Depending on the pulse sequence used for recording, the measurement of the MRI system takes from a few milliseconds to several seconds. Although longer recording times generally result in minimal noise artifacts, the impact of motion artifacts increases with the duration of the measurement. Although most patients try to remain still in order to avoid motion artifacts, there are unavoidable movements of the patient that cannot be stopped, such as breathing or heartbeat.
[0004] Furthermore, during an intervention, due to different contrasts, situations may arise in which the corresponding organs and devices required for the present invention cannot be seen simultaneously on the recorded images.
[0005] A typical EP workflow includes a so-called 3D “roadmap” that is a core component throughout the entire procedure. More precisely, the roadmap is a 3D acquisition of the organ (e.g., the entire heart) that is used to overlay the current position of the device (e.g., endoscope or catheter), which is actively tracked relative to the organ (e.g., heart) to guide the operator by reformatting 2D slices from the 3D volume according to the current device position and enables segmentation of the organ (e.g., the four chambers of the heart) for further post-processing and visualization options. The 3D dataset is acquired once before the entire workflow and is used to plan and guide the procedure.
[0006] Obviously, this roadmap can only provide limited accuracy because the heart is in continuous motion due to the respiratory and cardiac cycles. Therefore, there is a dynamic mismatch between the tracked device and the roadmap, which is critical when targeting small lesions such as 1 mm in size.
[0007] So far, the problem of live motion resolved roadmaps is currently not solved and applications are usually developed based on static 3D roadmaps with all known limitations.
[0008] Early work was devoted to 2D-3D registration of input image data into the roadmap (Xu, Robert, and Graham A. Wright. "GPU accelerated dynamic respiratory motion model correction for MRI-guided cardiac interventions." Computer methods and programs in biomedicine 136 (2016): 31-43). However, the underlying image registration method must be non-rigid and able to handle many different contrasts in real time. It is difficult to obtain computational workload and robustness. Summary of the invention
[0009] The object of the present invention is to improve known systems, devices and methods and to provide a solution for the (fast) creation of a (dynamic) roadmap of a medical workflow, preferably in order to provide a real-time motion-resolved roadmap of the workflow, in particular in the technical field of MRI.
[0010] This object is achieved by a method according to claim 1, a system according to claim 7, a method for creating a multidimensional image dataset according to claim 10, a multidimensional image dataset according to claim 11, a control device according to claim 12 and a magnetic resonance imaging system according to claim 13.
[0011] The present invention proposes a solution based on a combination of pilot tone technology (for example, see Schroeder, Lea, et al. "A novel method for contact-free cardiac synchronization using the pilot tone navigator." Proceedings of the 24th Annual Meeting of ISMRM, Singapore. 2016) and high-dimensional data sets to provide a real-time motion analysis roadmap.
[0012] The method according to the invention can be used to create a (dynamic) roadmap for a workflow in the technical field of MRI. A roadmap shows the actual state of a certain area of a patient and can be used to locate medical equipment in this area or simply display this area. The roadmap usually includes an actual image of the area, wherein the image does not have to be actually recorded (at least partially recorded), but still represents the actual state of the area.
[0013] The method comprises the following steps, wherein the method is preferably used during examination of a patient or intervention on a patient:
[0014] - providing a multi-dimensional image dataset comprising a plurality of images of a predefined organ in combination with a plurality of state dimensions characterizing a motion state of the moving organ,
[0015] - providing pilot tone data from measurements in continuous pilot tone signal acquisition,
[0016] - determining the coordinates of each state dimension based on the measured pilot tone data,
[0017] - Based on the determined number of coordinates for each state dimension, selecting an image of the multi-dimensional image dataset.
[0018] The selected image can be displayed on a display unit and / or be available in electronic form at the output of a processor, for example as a data file.
[0019] With regard to the step of providing a multidimensional image dataset, it should be noted that the multidimensional image dataset comprises a plurality of (particularly reconstructed) 2D or 3D images. These images are preferably MR images, but may also be or comprise other images, such as CT images, ultrasound images or simulated images. The images in the multidimensional image dataset are connected to coordinates of a plurality of other dimensions ("state dimensions"), particularly time-like.
[0020] These state dimensions represent the state of motion of the moving organ and include definite coordinates, i.e., certain states of motion of the moving organ. The motion state can be characterized by a time point, a time interval, or a signal shape (e.g., of a measured pilot tone signal). It should be noted that the moving organ does not have to be the same or different from the predefined organ. Therefore, in the case where the predefined organ is the heart, it is advantageous for one state dimension to characterize the motion of the heart itself, whereas another state dimension may characterize the motion of the lungs. In this case, the predefined organ is the liver, and it is not necessary to regard the liver as a moving organ. Typically, preferably, the moving organ is the heart (cardiac cycle) and / or the lungs (respiratory cycle). Intestinal motion may also be included, however, wherein the alternating motion cycle is preferably characterized by a state dimension, such as a respiratory cycle or a cardiac cycle. The number of state dimensions is in particular one or two, but may also be three or more. In particular in the trunk region, it is preferred that a 2D image or a 3D image is combined with at least one state dimension of the state of the respiratory cycle and / or the cardiac cycle.
[0021] The coordinates of the state dimension in the data set can be realized by the curve of the recorded pilot tone signal. The coordinates are then the values of the points on the curve or the intervals (bins) of the time intervals of the curve. The curve can be a separated part of the measured pilot tone signal. However, the coordinates can also be defined as a time delay or a part of the curve (for pattern matching) that characterizes a part of the curve (such as a heartbeat curve). However, in the case of alternating motion, it is also preferred to define the coordinates as a phase value or percentage value of one cycle. For example, a cardiac cycle or a respiratory cycle is an alternating motion. When a special point on a cycle of such a curve is predefined as "start", all other points of the cycle can be reached by running 2π of the phase or 0% to 100% of the cycle. Therefore, the coordinates themselves can be simple values of the phase or "percentage", and the measured signal can be compared with a model curve of one cycle of the curve (for example, by pattern matching).
[0022] Preferably, the multi-dimensional image dataset is a 5D dataset comprising a 3D image combined with cardiac cycle states (first state dimension) and respiratory cycle states (second state dimension).
[0023] 5D whole-heart acquisition has been introduced, which provides high-resolution spatial resolution in a way that resolves cardiac and respiratory motion (for example, see Feng, Li, et al. "5D whole-heart sparse MRI." Magnetic resonance inmedicine 79.2 (2018): 826-838). This technology provides basic information on how to obtain 5D data in MRI.
[0024] With respect to the step of providing measured pilot tone data from a continuous pilot tone signal acquisition, it must be noted that the measured pilot tone data is related to the patient's response to a continuous (constant) pilot tone signal applied to the patient from a source. The pilot tone data may be provided by measuring the patient's pilot tone signal when a continuous pilot tone signal is applied during or outside of an MRI acquisition. Using the pilot tone signal, each acquisition independent of dimension, acquisition scheme, or contrast may be associated with a unique fingerprint, which may then be used to match physiological states in different acquisitions. The use of pilot tone signals is known in the art of MRI, where the "tone" does not refer to an audible sound wave, but rather to a continuous RF signal applied from a source (e.g., a coil).
[0025] The location where the continuous pilot tone signal is applied is usually arbitrary, since in an MRI system it can be measured throughout the examination room. However, for best results it is preferred to apply the continuous pilot tone signal to the region of the organ to be examined or the region where intervention should be performed.
[0026] If a model curve is used in a subsequent step of determining coordinates, the measured pilot tone data (or separated pilot tone data) is preferably calibrated or registered in order to match the model curve. For example, the maximum of the separated heartbeat curve of the measured pilot tone data does not necessarily have to coincide with the maximum of the ECG curve, however, the points of the curve of the pilot tone data are still linked to the points of the model curve. Preferably, the correlation of the measured pilot tone data with the model curve for determining the coordinates of the state dimension is defined before determining the coordinates.
[0027] With regard to the step of determining the coordinates of each state dimension based on the measured pilot tone data, it must be noted that the measured pilot tone data includes information about the motion state of an organ (e.g., the heart or lungs). Since the motion state of an organ is a coordinate on the state dimension (see above), the measured pilot tone signal can define the region on the state dimension and the time point on the measured pilot tone signal as a discrete coordinate of the state dimension. In the case where there are two or more state dimensions, the measured pilot tone signal can identify the coordinates on two / all state dimensions because the motion can be separated from a single curve (measured pilot tone signal) due to characteristic behavior, such as by fast Fourier transform or fitting and subtracting a predefined curve shape. Therefore, in the trunk, the motion caused by the respiratory cycle can be separated from the motion caused by the cardiac cycle, and the state of the respiratory cycle and the cardiac cycle can be defined at a given time point, resulting in the coordinates on the state dimension of the breathing being identified on the one hand, and the coordinates on the state dimension of the heart being identified on the other hand.
[0028] Thereafter, since the coordinates are known and the images in the multi-dimensional image dataset are combined with the individual coordinates, it is easy to select the images in the multi-dimensional image dataset based on the number of determined coordinates for each state dimension.
[0029] Thus, the present invention can combine (MR) image acquisition with continuous pilot tone signal acquisition, which can provide real-time data about the respiratory and cardiac cycles without attaching any additional sensors to the patient. With the pilot tone signal used, each acquisition independent of dimension, acquisition scheme or contrast can be linked to a unique fingerprint, which can then be used to match physiological states in different acquisitions.
[0030] The system for creating a roadmap for a medical workflow according to the present invention comprises the following components:
[0031] - a database comprising a multidimensional image dataset comprising a plurality of images of a predefined organ in combination with a plurality of state dimensions characterizing a movement state of the moving organ,
[0032] - a data interface designed to receive pilot tone data from measurements in a continuous pilot tone signal acquisition,
[0033] a determination unit designed to determine the coordinates of each state dimension based on the measured pilot tone data,
[0034] - A selection unit designed to select an image of the multidimensional image dataset based on the number of coordinates of each state dimension.
[0035] The method according to the invention for creating a multidimensional image dataset is designed for use with the method according to the invention and / or the system according to the invention and comprises the following steps:
[0036] -Providing multiple images of a predefined organ, the multiple images showing the predefined organ in different states of motion of the moving organ. Regarding the predefined organ and the moving organ, it should be noted that, depending on the nature of the predefined organ, it is also considered a moving organ or not. If the predefined organ is the heart, a state dimension should also identify the heart as a moving organ. In case the predefined organ is the liver, it is not necessary to consider it as a moving organ. The moving organ is preferably the heart or the lung. A single moving organ can be monitored, or two, three or more moving organs can be monitored, wherein it is advantageous to separate the influence of the two or more moving organs representing the state dimension.
[0037] - determining a motion state of one or more moving organs for each image of the predefined organ. This is done in order to correctly arrange the images of the predefined organ within the multidimensional image dataset, since the motion state is a coordinate of the state dimension in the multidimensional image dataset.
[0038] - Storing each picture in a multi-dimensional image dataset, wherein each picture is associated with a determined motion state (ie, coordinates in the multi-dimensional image dataset).
[0039] In such a multi-dimensional image dataset, for many possible motion states of one or more motion organs, the changing shapes of the predefined organs are depicted in the form of multiple pictures.
[0040] The multidimensional image dataset according to the invention is designed for use with the method according to the invention and / or the system according to the invention and is in particular produced with the method according to the invention. The multidimensional image dataset comprises a plurality of images of a predefined organ showing the predefined organ in different states of motion of the moving organ (or two or more moving organs), wherein each picture is associated with a discrete state of motion of one or more moving organs.
[0041] The control device for controlling a magnetic resonance imaging system according to the invention comprises a system according to the invention (at least one of the two systems described above) and / or is designed to perform a method according to the invention (at least one of the two methods described above). The control device may comprise additional units or devices for controlling components of the magnetic resonance imaging system, such as a sequence control unit for measuring sequence control, a memory, a radio frequency transmission device for generating, amplifying and transmitting RF pulses, a gradient system interface, a radio frequency reception device for acquiring magnetic resonance signals and / or a reconstruction unit for reconstructing magnetic resonance image data.
[0042] The magnetic resonance imaging system according to the invention comprises a control device according to the invention or is at least designed to carry out a method according to the invention (at least one of the two methods described above).
[0043] Some units or modules of the above-mentioned system or control device can be fully or partially implemented as software modules running on the processor of the system or control device. The implementation mainly in the form of software modules can have such advantages that applications already installed on the existing system can be updated with a relatively small burden of installing and running these units of the present application. The purpose of the present invention is also achieved by a computer program product with a computer program, which can be directly loaded into the memory of the device of the system or the control device of the magnetic resonance imaging system, and the computer program product includes a program unit, when the program is executed by the control device or system, the program unit performs the steps of the method of the present invention. In addition to the computer program, this computer program product may also include other parts, such as documents and / or additional components, and hardware components, such as hardware keys (dongles, etc.), to facilitate access to the software.
[0044] Computer readable media such as a memory stick, hard disk or other transportable or permanently installed carrier may be used to transport and / or store executable parts of a computer program product so that these can be read by a processor unit of a control device, MPSU or system. The processor unit may include one or more microprocessors or their equivalents.
[0045] The dependent claims give particularly advantageous embodiments and features of the invention, as revealed in the following description. Features of different claim categories may be combined as appropriate to give further embodiments not described herein.
[0046] With regard to the preferred method, image acquisition is performed in parallel (ie, simultaneously) with continuous pilot tone signal acquisition. The image acquisition is preferably a magnetic resonance (MR) image acquisition and should be optimized to display devices used for intervention (eg, medical procedures using catheters or endoscopes).
[0047] Preferably, the measured pilot tone data are used to match the acquired (MR) images with corresponding images of the multidimensional image dataset. As described above, this can be achieved by selecting images of predefined organs from the multidimensional image dataset that match the actual motion state of the patient (e.g. cardiac and respiratory cycles). The image acquisition is preferably designed so that the device can be tracked within the patient.
[0048] It is advantageous to register the coordinates of the MR image with the image selected from the multi-dimensional image dataset in order to show the correct position of the device relative to the organ.
[0049] It should be noted that measured pilot tone data refers to a measured pilot tone signal where the data is a digital representation of the measured pilot tone signal (or a separated sub-signal of the pilot tone signal). It can be said that the measured pilot tone signal is a curve (e.g., an envelope curve), and the pilot tone data is a digital representation of the curve.
[0050] In this regard, it should be noted that, particularly in the field of MRI, the contrast that is most suitable for depicting organs is usually not the contrast used to depict devices (such as catheters or endoscopes in the body). Therefore, MR images of display devices during intervention are usually not suitable for correctly displaying organs or areas around the device. Because MRI usually requires a non-negligible amount of time, it is preferred to present images of other display organs aligned with the pictures of the display device. This can be achieved by the method according to the present invention, because images are acquired during the operation of the display corresponding device in parallel with the measurement of the continuous pilot tone signal (i.e., pilot tone data). As described above, pilot tone data can be used to select pictures (which should include images of patients performing operations) from a multidimensional image data set. These images depict (or better simulate) each organ in the correct position and shape, regardless of movement (such as movement of the heart or lungs), because the correct picture is selected for each motion state.
[0051] According to a preferred method, the measurements of the pilot tone data are associated with the measurements of the central k-space region of the image acquisition. Alternatively or additionally, preferably, in the process of determining the coordinates for each state dimension, the pilot tone data measured in the central k-space region are used. This information is preferably limited to the central k-space region, because they constitute the majority of the image information.
[0052] According to a preferred method, during the pilot tone signal acquisition process, intervals of measured pilot tone signals (i.e. pilot tone data) corresponding to predefined time periods are stored. With respect to respiratory and cardiac activity, these time periods are typically shorter than a respiratory cycle or a cardiac cycle. Preferably, the coordinates of each state dimension are preferably determined by pattern matching of the stored intervals with a pattern of coordinates according to the state dimension.
[0053] According to a preferred method, the measured pilot tone data (i.e. the measured pilot tone signal) are analyzed to determine the time delays of predefined points of the measured pilot tone data relative to representative motion points in order to mark the motion state of the organ. Preferably, the minimum distance of the determined time delays is used to match the acquired image with the corresponding image of the multidimensional image data set. It is further preferred that the representative physiological heart state is marked by a time delay (ΔtC) relative to the last pilot tone heart signal maximum or minimum, and / or the representative physiological respiratory state is marked by a time delay (ΔtR) relative to the last pilot tone respiratory signal maximum or minimum. This is performed in particular by acquiring a central k-space line. If the k-space center is acquired multiple times per 2D or 3D image (e.g. spiral out), it is preferred to acquire an average delay. This option relies on the maximum detection accuracy of the cardiac and respiratory PT signals. It is also preferred to use two or more time delays (depending on different characterization points on the curve).
[0054] According to a preferred method, for an examination or intervention on a predefined organ, a multidimensional image dataset comprises a plurality of images of the predefined organ in different states of motion, the images being combined with a plurality of state dimensions characterizing the states of motion of the heart and / or lungs, wherein the measured pilot tone data represent a cardiac cycle and / or a respiratory cycle, and wherein the image of the organ is selected from the multidimensional image dataset representing the shape of the organ in motion based on one or more determined coordinates. The fingerprint for such a 5D acquisition (if the predefined image is a 3D image and both the heart and lungs are monitored) will preferably contain a group of many pilot tone signal (data) segments or a set of delays for each 3D data subset (as described above). Due to the measured pilot tone signal, each subsequent acquisition will include a corresponding fingerprint, which can then be used to match an ideal 3D subset of the multidimensional image dataset for providing a dynamic roadmap.
[0055] With respect to the above-described embodiments, in case of recording samples of the measured pilot tone signals, preferably for each 3D subset (image) of the multidimensional image dataset (which is added to the dataset or the coordinates are determined), the sum of squared differences of the signal curves is calculated for all pilot tone signal segments of the relevant k-space region. A specific weighting of the calculated differences across the pilot tone signal segments may be applied to emphasize the contribution of the pilot tone signal segments associated with the central k-space region. A relationship metric may be used to determine similarity instead of the sum of squared differences between the pilot tone signal segments.
[0056] In general, with respect to pattern matching of recorded pilot tone signals, it should be noted that the multi-dimensional image dataset should include a reference curve for each coordinate to facilitate pattern matching.
[0057] With respect to the above embodiments, where the measured time delays of the pilot tone signals are recorded, preferably the minimum distance of the determined delays (ΔtR, ΔtC) is used to match the acquired images to the corresponding 3D subset of the 5D roadmap data.
[0058] In general, with respect to the time delays of the recorded pilot tone signals, it should be noted that the multi-dimensional image dataset should include the time delay for each coordinate to facilitate matching of the time delays.
[0059] According to the preferred method, the time difference between two measurements of pilot tone data is determined and used to optimize the step of determining coordinates. Each cycle, such as a cardiac cycle or a respiratory cycle, has a determined duration. Depending on the actual patient's condition, it may be faster or slower, but in reality, at least for a short period of time, some areas of such a cycle cannot be "reached". Therefore, the second pilot tone data collected shortly after the first pilot tone data (which coordinates have been determined) should be able to be located at the corresponding coordinate on the state dimension (here representing the cycle) after the first coordinate. If the movement cycle is faster or slower, this coordinate may change a little, but not much. Therefore, preferably, by starting from the coordinates determined by the first pilot tone data, measuring the time difference to the acquisition of the second pilot tone data, determining the time interval of the curve of the corresponding movement cycle and / or the coordinate interval of the corresponding state dimension that can be reached in the measured time difference, in particular by involving a predefined threshold value of the acceleration and / or deceleration cycle, based on the determined interval to determine the new coordinates, to achieve optimization. The positive effect of this embodiment is the optimization of calculation time and accuracy, especially if pattern matching is involved.
[0060] A preferred system comprises a magnetic resonance scanner or a data interface designed to control a magnetic resonance scanner, wherein the system is preferably designed to control image acquisition of the magnetic resonance scanner in parallel with pilot tone signal acquisition. Additionally or alternatively, a preferred system comprises a pilot tone generator or a pilot tone control unit controlling the pilot tone generator, wherein the system is preferably designed to control image acquisition of the magnetic resonance scanner in parallel with pilot tone signal acquisition.
[0061] A preferred system comprises a matching unit designed to match the acquired image with a corresponding image of the multi-dimensional image dataset based on the measured pilot tone signal.
[0062] An advantage of the present invention is that a roadmap can be dynamically represented from a high-dimensional dataset in real-time (by continuous pilot tone recording and matching) or by triggering an acquisition by an arbitrary k-space center readout of a pilot tone signal. Furthermore, dynamic roadmap correction of a static 3D dataset can be achieved in real-time by matching the current respiratory state from the pilot tone signal with the lateral correction factor.
[0063] The present invention is compatible with any type of image readout strategy that provides representative respiratory and cardiac pilot tone signal states and does not rely on any specific image contrast. In addition, the method according to the present invention is robust to potential image distortions caused by, for example, acquisition strategies; motion artifacts; susceptibility artifacts, etc., and the method is compatible with non-triggered and non-gated acquisitions, i.e., no other respiratory or ECG signals are included. Instead of selecting a corresponding roadmap, a corrective deformation field can be applied to the input imaging data based on the deformation difference between the shown 3D roadmap and the current best matching roadmap. In addition, the present invention does not require any additional sensors to be attached to the patient, which is a great advantage, especially in interventions that require many other monitoring devices. BRIEF DESCRIPTION OF THE DRAWINGS
[0064] Other objects and features of the present invention will become apparent from the following detailed description considered in conjunction with the accompanying drawings. It should be understood, however, that the accompanying drawings are designed only for illustrative purposes and not as a definition of the limits of the present invention.
[0065] Figure 1 A simplified MRI system is shown with an example of a system according to an embodiment of the invention.
[0066] Figure 2 A block diagram showing the process flow of a preferred method according to the present invention is shown.
[0067] Figure 3 An example of a multidimensional image dataset is shown.
[0068] Figure 4 An example of data collection is outlined.
[0069] Figure 5 An example of determining the coordinates of two status dimensions and creating a roadmap is outlined.
[0070] In the diagrams, identical numbers refer to identical objects. Objects in the diagrams are not necessarily drawn to scale. DETAILED DESCRIPTION
[0071] Figure 1 A schematic diagram of a magnetic resonance imaging system 1 ("MRI system") is shown. The MRI system 1 comprises an actual magnetic resonance scanner 2 having an examination space 3 or a patient channel, in which a patient or an examinee is placed on a moving bed 8, and in which the actual examination object is located.
[0072] The magnetic resonance scanner 2 is usually equipped with a basic field magnet system 4, a gradient system 6 and an RF transmission antenna system 5 and an RF reception antenna system 7. In the exemplary embodiment shown, the RF transmission antenna system 5 is a whole-body coil permanently installed in the magnetic resonance scanner 2, in contrast to which the RF reception antenna system 7 is formed as a plurality of local coils arranged on a patient or test object. In principle, however, a whole-body coil can also be used as an RF reception antenna system and the local coils can be switched to different operating modes respectively.
[0073] In this example, according to the present invention, the RF receiving antenna system 7 is used to measure the pilot tone signal PS (eg Figure 4 ). Below the object O, a generator coil 20 for transmitting a continuous pilot tone PT is located.
[0074] The basic magnetic field magnet system 4 is designed to be able to record a region of interest ("RoI"). Here, it is designed in a typical manner so that it generates a basic magnetic field in the longitudinal direction of the patient, i.e. along the longitudinal axis of the magnetic resonance scanner 2 running in the z direction. The gradient system 6 usually includes individually controllable gradient coils so that gradients can be switched (activated) independently of each other in the x-direction, y-direction or z-direction.
[0075] The MRI system 1 shown here is a full-body system with a patient channel into which the patient can be completely introduced. However, in principle, the invention can also be used for other MRI systems, such as a C-shaped housing with a transverse opening, and smaller magnetic resonance scanners in which only one body part can be located.
[0076] Furthermore, the MRI system 1 has a central control device 13 for controlling the MRI system 1. The central control device 13 comprises a sequence control unit 14 for measuring sequence control. With the sequence control unit 14, a sequence of radio frequency pulses (RF pulses) and gradient pulses can be controlled according to a selected pulse sequence or respectively according to a series of multiple pulse sequences to acquire a magnetic resonance image of the RoI within a measurement session. For example, such a series of pulse sequences can be predetermined in a measurement or control protocol P. Different control protocols P for different measurements or measurement sessions are usually stored in a memory 19 and can be selected by an operator (and possibly modified as required) and then used to realize the measurement.
[0077] In order to output the individual RF pulses of the pulse sequence, the central control device 13 has a radio frequency transmission device 15, which generates and amplifies the RF pulses and feeds them via a suitable interface (not shown in detail) to the RF transmission antenna system 5. In order to control the gradient coils of the gradient system 6, the control device 13 has a gradient system interface 16. The sequence control unit 14 communicates in a suitable manner with the radio frequency transmission device 15 and the gradient system interface 16 to transmit the pulse sequence.
[0078] The control device 13 furthermore has a radio frequency receiving device 17 (likewise communicating in a suitable manner with the sequence control unit 14 ) in order to acquire magnetic resonance signals (i.e. raw data) for the individual measurements, by means of which the magnetic resonance signals are received in a coordinated manner from the RF receiving antenna system 7 within the scope of a pulse sequence.
[0079] The reconstruction unit 18 receives the acquired raw data and reconstructs the magnetic resonance image data for the measurement therefrom. Such reconstruction is usually performed based on the parameters specified in the corresponding measurement or control protocol. For example, the image data can then be stored in the memory 19.
[0080] The operation of the central control device 13 can be carried out via a terminal 11 having an input unit 10 and a display unit 9, and thus the entire MRI system 1 can also be operated by an operator via this terminal. MR images can also be displayed on the display unit 9, and measurements can be planned and started via the input unit (possibly in combination with the display unit 9), and in particular, suitable control protocols can be selected (and possibly modified) with a suitable series of pulse sequences as described above.
[0081] The control device 13 comprises a system 12, which is designed to perform the method according to the invention. The system 12 comprises the following components, which may be partially represented as software modules:
[0082] - A database 21, comprising a multi-dimensional image data set, the multi-dimensional image data set comprising a plurality of images of an organ combined with a plurality of state dimensions characterizing the movement state of the organ. As storage space for the database, the memory 19 can also be used.
[0083] A data interface 22 is designed to receive measured pilot tone data from the continuous pilot tone signal acquisition. The data interface 22 can also be used to receive image data reconstructed by the reconstruction unit 18.
[0084] - a determination unit 23, designed to determine for each state dimension D1, D2 (see e.g. Figure 3 )Determine the coordinates C1 and C2.
[0085] a selection unit 24 designed to select an image A of the multidimensional image data set M based on the coordinates C1, C2 determined for each state dimension D1, D2 11 , A 12 , A 13 , A 21 , A 22 , A 31 .
[0086] A matching unit 25, designed to match the acquired image B with the corresponding image A of the multi-dimensional image dataset M based on the measured pilot tone signal PS 31 (See, for example, Figure 5 The acquired image B has been measured by the magnetic resonance scanner 2 in parallel with the measurement of the pilot tone signal PS.
[0087] The MRI system 1 according to the invention, in particular the control device 13, can have a plurality of additional components which are not shown in detail here but are typically present in such a system, such as a network interface in order to connect the entire system to a network and to be able to exchange raw data and / or image data or parameter maps, respectively, but also additional data (e.g. patient-related data or control protocols).
[0088] Figure 2 A block diagram of the process flow of the preferred method according to the present invention is shown. The steps of the process flow are as follows: Figures 3 to 5 It is assumed that the method is performed during an examination of a patient.
[0089] In step I, a multidimensional image dataset M is provided, in particular in a database 21 (see for example Figure 1 The multidimensional image dataset M includes a plurality of images A of a patient's organ (organ under examination) combined with a plurality of state dimensions D1, D2 representing the motion state of the organ (e.g., one state dimension D2 pertaining to the motion state of the heart and one state dimension D1 pertaining to the motion state of the lung). 11 , A 12 , A 13 , A 21 , A 22 , A 31 .
[0090] In step II, pilot tone data from measurements in a continuous pilot tone signal acquisition are provided. This is done by measuring the pilot tone signal as a response of the patient to the continuous pilot tone signal applied by the HF coil.
[0091] In step II, based on the measured pilot tone data PD, coordinates C1 , C2 are determined for each state dimension D1 , D2 . Figure 5 This step is further explained.
[0092] In step III, an image A of the multidimensional image dataset M is selected based on the determined number of coordinates C1, C2 of each state dimension D1, D2. 31 .
[0093] In step IV, using an MRI system (e.g., Figure 1 ) is performed in parallel with continuous pilot tone signal acquisition. In this example, the acquired image B is a 3D picture (or a stack of 2D pictures) in which the contrast is selected so that the device D (e.g., a catheter or endoscope) can be reconstructed very accurately.
[0094] In step V, the measured pilot tone signal PS is used to compare the acquired image B of the device with the image A of the corresponding multidimensional image dataset M. 31 Match.
[0095] Figure 3 A plurality of images A including organs are shown 11 , A 12 , A 13 , A 21 , A 22 , A 31 An example of a multidimensional image dataset M, where these images A 11 , A 12 , A 13 , A 21 , A 22 , A 31 Combined with two state dimensions D1 and D2. One state dimension D1 represents the motion state of the lungs during the respiratory cycle, and the other state dimension D2 represents the motion state of the cardiac cycle. Image A 11 , A 12 , A 13 , A 21 , A 22 , A 31 An organ O that should be examined in multiple states during many possible combinations of motion states during the respiratory and cardiac cycles is shown, where each image A 11 , A 12 , A 13 , A 21 , A 22 , A 31 With each state dimension D1, D2 special coordinates C1, C2 (see Figure 5 ) are connected, and the coordinates are suitable for obtaining image A 11 , A 12 , A 13 , A 21 , A 22 , A 31 state of motion.
[0096] For better understanding, the picture shows the respiratory cycle in the state dimension D1 at the top, where the arrows show the actual motion state. On the left, the same is true for the cardiac cycle. Each dotted box within the stereo box (multidimensional image dataset M) should represent a set of 2D coordinates in the state dimension. Each coordinate corresponds to a special (3D) image A showing the organ in a very special motion state. 11 , A 12 , A 13 , A 21 , A 22 , A 31 The multidimensional image dataset M contains images A of different motion state combinations. 11 , A 12 , A 13 , A 21 , A 22 , A 31 The more you have, the more accurate the roadmap will be.
[0097] Figure 4 An example of data collection is outlined. In the above-described action route, a continuous pilot tone signal PT is applied to the patient, for example at a frequency of several MHz. The continuous pilot tone signal PT generates a measurable (changed) pilot tone signal in the patient's body in response. Following the arrow in the upper left corner, the measured pilot tone signal PS is shown, wherein it should be noted that the curve shown here is a curve (envelope curve) on the amplitude of the measured pilot tone signal PS. When the applied pilot tone signal has a frequency of approximately several MHz, the peak of the measured curve alternates at a frequency of less than 1 Hz, wherein the sampling of the curve is within the range of the applied frequency of the continuous pilot tone signal. In the shown curve of the measured pilot tone signal PS, the respiratory cycle and the cardiac cycle are monitored in combination.
[0098] In order to find the coordinates C1, C2 of the state dimensions D1, D2, it is advantageous to separate the contributions of the respiratory cycle and the cardiac cycle from the measured pilot tone signal PS. This is shown at the top right, where the data of the separated signal are here the pilot tone data PD. The separation can be achieved, for example, by a fast Fourier transform or by fitting a curve (for example for the respiratory cycle) and subtracting the fitted curve from the measured pilot tone signal PS (to obtain the curve for the cardiac cycle).
[0099] In the following course of action, a (real) image B is acquired from the patient, wherein the image is optimized to show a device D, here for example a catheter, which is inside the patient during an examination or intervention. With regard to the k-space k of this image B, it is known that the center k line comprises the most important information of the image. Therefore, in this example, the part in which the pilot tone signal PS measured during the center k line of the image B is recorded is marked (arrow), wherein this marking is also transmitted to the pilot tone data PD.
[0100] Figure 5 An example of determining the coordinates C1, C2 of the two state dimensions D1, D2 and creating a roadmap is outlined. As described above, the measured pilot tone signal PS is separated into data related to the respiratory cycle and data related to the cardiac cycle to form the pilot tone data PD. The dashed box shows the time of acquiring the center k line of the MR image B (see, for example Figure 4 ), where the dashed line marks the center of the dashed box and the time points defined by it.
[0101] Now, each curve in the pilot tone data is compared with the corresponding curve of the corresponding state dimension D1, D2. This example shows two possible techniques.
[0102] The curve associated with the respiratory cycle is recorded at time intervals around the dashed line. By using pattern matching, the recorded part is compared with the part of the reference curve representing the state dimension (see the upper curve of the multidimensional image dataset M). The dashed line is determined to be located at a clear coordinate C1 of a state dimension D1 (dashed circle).
[0103] The curve for the cardiac cycle is defined by recording the time distance Δt to a characteristic point of the cardiac cycle (here the center or R peak of the QRS complex). By comparing this time distance Δt with the time distance stored with respect to another state dimension D2 (see the arrow of the curve next to the left part of the multidimensional image data set M), it is determined that the dashed line is located at the position of the clear coordinate C2 of one state dimension D2 (dashed circle).
[0104] Image A of the multidimensional image dataset M 31 It is defined by the determined coordinates C1, C2 and is selected for the roadmap. Here, image A 31 Show in Figure 1 The MRI system 1 is shown on a monitor 9 of a computer.
[0105] In combination with this method, in particular according to Figure 4, the device D can now also be displayed on the monitor 9. Thus, the problem that organs are usually not optimally visible on the image showing the device D is solved, and by using the multidimensional image dataset M, the organs can be displayed in a good relationship with respect to the device D without any errors regarding breathing or heartbeat.
[0106] Although the present invention has been disclosed in the form of preferred embodiments and variations thereof, it will be appreciated that many additional modifications and variations may be made thereto without departing from the scope of the present invention. For the sake of clarity, it will be appreciated that "a" or "an" as used in this application does not exclude a plurality, and "comprising" does not exclude other steps or elements. Reference to a "unit" or "device" does not exclude the use of more than one unit or device.
Claims
1. A method for creating a roadmap for a medical workflow, comprising the following steps: - providing a multidimensional image dataset (M) comprising a plurality of images (A) of a predefined organ (O) in combination with a plurality of state dimensions (D1, D2) characterizing a motion state of the moving organ; 11 , A 12 , A 13 , A 21 , A 22 , A 31 ), - providing pilot tone data (PD) from measurements in continuous pilot tone (PT) signal acquisition, - based on the measured pilot tone data (PD), determining the coordinates (C1, C2) for each state dimension (D1, D2), - based on the number of determined coordinates (C1, C2) of each state dimension (D1, D2), selecting an image (A) of the multidimensional image dataset (M) 11 , A 12 , A 13 , A 21 , A 22 , A 31 ); wherein image acquisition is performed in parallel with continuous pilot tone (PT) signal acquisition, and The measured pilot tone data (PD) is used to compare the acquired image (B) with the corresponding predefined image (A) of the multidimensional image dataset (M). 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) matches.
2. The method according to claim 1, wherein: The image acquisition is designed to enable the device (D) to be tracked within the patient's body.
3. The method according to claim 2, wherein: Measurements of pilot tone data (PD) are associated with measurements of a central k-space region of the image acquisition and / or wherein pilot tone data (PD) measured in the central k-space region are used during determination of coordinates (C1, C2) for each state dimension (D1, D2).
4. The method according to claim 1, wherein: During pilot tone signal acquisition, intervals of measured pilot tone data (PD) corresponding to predefined time periods are stored, wherein coordinates (C1, C2) of each state dimension (D1, D2) are determined by pattern matching the stored intervals with a pattern of coordinates (C1, C2) according to the state dimensions (D1, D2).
5. The method according to claim 4, wherein: The time difference between two acquisitions of pilot tone data is measured and used to optimize the coordinate determination.
6. The method according to claim 1, wherein: analyzing the measured pilot tone data (PD) to determine a time delay (Δt) of a predefined point of the measured pilot tone data (PD) relative to a representative motion point in order to mark a motion state of a predefined organ (O), The minimum distance of the determined time delay (Δt) is used to compare the acquired image (B) with the corresponding image (A) of the multidimensional image dataset (M). 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) matches.
7. The method according to claim 6, wherein: A representative physiological cardiac state is marked by a time delay (Δt) relative to the last pilot tone cardiac signal maximum or minimum, and / or a representative physiological respiratory state is marked by a time delay (Δt) relative to the last pilot tone respiratory signal maximum or minimum.
8. The method according to any one of the preceding claims, wherein: For an examination or intervention on a predefined organ (O), a multidimensional image dataset (M) comprises a plurality of images (A) of the predefined organ (O) in different motion states. 11 , A 12 , A 13 , A 21 , A 22 , A 31 ), the plurality of images (A 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) is combined with a plurality of state dimensions (D1, D2) characterizing a motion state of a heart and / or a lung as a moving organ, wherein the measured pilot tone data (PD) represent a cardiac cycle and / or a respiratory cycle, wherein an image (A) of a predefined organ (O) is selected from a multidimensional image dataset (M) representing a shape of the predefined organ (O) in a motion state based on one or more determined coordinates (C1, C2). 11 , A 12 , A 13 , A 21 , A 22 , A 31 ).
9. The method according to claim 1, wherein: The multidimensional image dataset (M) is created by the following steps: - Provides multiple images (A) of a predefined organ (O) 11 , A 12 , A 13 , A 21 , A 22 , A 31 ), multiple images (A 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) shows a predefined organ (O) in different motion states of the motion organ, - for each image (A) of a predetermined organ (O) 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) determining the movement state of the movement organ, - For each image (A 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) is stored in a multidimensional image dataset (M), where each image (A 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) is associated with a certain state of motion.
10. A system (12) for creating a roadmap for a medical workflow, comprising: - a database (21) comprising a multidimensional image dataset (M) comprising a plurality of images (A) of a predefined organ (O) in combination with a plurality of state dimensions (D1, D2) characterizing a movement state of the moving organ 11 , A 12 , A 13 , A 21 , A 22 , A 31 ), - a data interface (22) designed to collect received measured pilot tone data (PD) from a continuous pilot tone (PT) signal, a determination unit (23) designed to determine the coordinates (C1, C2) of each state dimension (D1, D2) based on the measured pilot tone data (PD), a selection unit (24) designed to select an image (A) of the multidimensional image data set (M) based on a determined number of coordinates (C1, C2) of each state dimension (D1, D2); 11 , A 12 , A 13 , A 21 , A 22 , A 31 ),as well as a matching unit (25) designed to match an image (B) acquired in parallel with the acquisition of a continuous pilot tone (PT) signal with a corresponding image (A) of the multidimensional image dataset (M) based on the measured pilot tone signal (PS); 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) to match.
11. The system according to claim 10, comprising a magnetic resonance scanner (1) or a data interface designed to control a magnetic resonance scanner (1), wherein: The system (12) is designed to control the image acquisition of the magnetic resonance scanner (1) in parallel with the pilot tone signal acquisition and / or The system (12) comprises a pilot tone generator (20) or a pilot tone control unit controlling the pilot tone generator (20), wherein the system (12) is designed to control image acquisition of a magnetic resonance scanner (1) in parallel with pilot tone signal acquisition.
12. The system according to claim 10, wherein: The plurality of images (A 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) shows a predefined organ (O) in different motion states of the moving organ, wherein each image (A 11 , A 12 , A 13 , A 21 , A 22 , A 31 ) is associated with discrete movement states of the motor organ.
13. A control device (13) for controlling a magnetic resonance imaging system (1), comprising a system (12) according to one of claims 10 to 12 and / or being designed to carry out a method according to any of claims 1 to 9.
14. A magnetic resonance imaging system comprising a control device (13) according to claim 13.
15. A computer program product comprising a computer program, which can be directly loaded into a system (12) or a control device (13) for a magnetic resonance imaging system, the computer program product comprising program elements for performing the steps of the method according to any one of claims 1 to 9 when the computer program is executed by the system (12) or the control device (13).
16. A computer-readable medium having a program element stored thereon, which, when the program element is executed by a computer unit, can be read and executed by the computer unit in order to perform the steps of the method according to any one of claims 1 to 9.
