Method and system for motion compensation in hybrid PET-MR imaging

Abstract

The present invention is entitled "Method and system for motion compensation in hybrid PET-MR imaging". The present invention discloses a method of controlling and processing data from a hybrid PET-MR imaging system, comprising: acquiring a positron emission tomography (PET) data set over a time period, wherein the PET data set is affected by quasi-periodic motion of a patient; and acquiring magnetic resonance (MR) data during the time period, such that the acquisition time of the MR data relative to the PET data set is known. The characteristics of the patient motion are then determined based on the PET data set, and the MR data are processed based on the characteristics of the patient motion.

Description

[0001] CROSS-REFERENCE TO RELATED APPLICATIONS

[0002] This application is based upon and claims the benefit of priority of U.S. Provisional Patent Application Serial No. 62 / 947,300, filed on December 12, 2019, the disclosure of which is incorporated herein by reference. Background Art

[0003] The present disclosure relates generally to hybrid PET-MR imaging systems and methods, and more particularly to methods and systems for motion compensation in hybrid PET-MR imaging.

[0004] In many cases in medical imaging, motion is unavoidable. For example, in diagnostic PET, the acquisition duration is typically about two minutes per bed position. The patient's respiratory motion during PET acquisition causes the resulting (static) PET image to be blurred. This in turn may lead to lower tumor detectability, inaccurate SUV calculations, and incorrect tumor planning volumes in radiotherapy. The first step in correcting motion is to gate the data to different motion states. Gating is typically based on information from an external device, such as chest position for respiratory motion or ECG signals for cardiac gating. Various external devices are used to obtain gating signals, which are typically devices that must be placed on the patient, such as a respiratory monitor (e.g., a chest strap) and / or an ECG monitor (ECG electrodes). Summary of the invention

[0005] This Summary is provided to introduce a selection of concepts that will be further described below in the Detailed Description. This Summary is not intended to identify key or essential features of the claimed subject matter, nor is it intended to be used to help limit the scope of the claimed subject matter.

[0006] In one embodiment, a method of controlling and processing data from a hybrid PET-MR imaging system includes: acquiring a positron emission tomography (PET) data set over a time period, wherein the PET data set is affected by quasi-periodic motion of a patient; and acquiring magnetic resonance (MR) data during the time period such that the acquisition time of the MR data relative to the PET data set is known. A characteristic of the patient motion is then determined based on the PET data set, and the MR data is processed based on the characteristic of the patient motion.

[0007] One embodiment of a hybrid PET-MR imaging system includes: a PET imaging portion configured to generate PET data to image a patient; and an MR imaging portion configured to generate MR data to image the patient. A controller is configured to: acquire a PET data set from the PET imaging portion for a time period, wherein the PET data set is affected by quasi-periodic motion of the patient; and acquire MR data with the MR imaging portion during the time period, so that the acquisition time of the MR data relative to the PET data set is known. The controller is also configured to determine characteristics of patient motion based on the PET data set, and process the MR data based on the characteristics of the patient motion.

[0008] Various other features, objects and advantages of the invention will become apparent from the following description taken in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The present disclosure is described with reference to the following figures.

[0010] Figure 1 is a diagram of an embodiment of a PET-MR imager and a system for determining gating signals for both PET and MR modalities.

[0011] Figure 2 is a flow chart illustrating one embodiment of a method of gating PET and MR data.

[0012] Figure 3 is a flow chart illustrating another embodiment of a method of gating PET and MR data.

[0013] FIG. 4A to FIG. 4C An embodiment of simultaneous acquisition of PET and MR data is schematically represented.

[0014] Figure 5 An exemplary gating signal is depicted as a respiratory waveform.

[0015] Figure 6 The steps of processing MR data based on a gating signal derived from a PET data set are schematically depicted.

[0016] Figure 7 is a flow chart illustrating another embodiment of motion extraction and compensation for simultaneously acquired PET and MR data. DETAILED DESCRIPTION

[0017] Embodiments of methods and systems for motion detection and compensation in dual-modality PET-MR imagers are disclosed. PET-MR imagers perform both positron emission tomography (PET) and magnetic resonance (MR) imaging modalities in one system. Hybrid PET-MR imaging systems acquire PET and MR data simultaneously. Each of the PET data and the MR data must be motion corrected (such as gated) to compensate for and reduce the effects of quasi-periodic motion of the patient (including respiratory motion and cardiac motion).

[0018] The inventors have recognized that current methods for motion compensation in PET data and MR data from hybrid PET-MR imaging systems are inadequate. Motion compensation is a long-standing problem for nearly all imaging modalities, especially PET and MR modalities. Uncompensated motion adversely affects the quality of both PET and MR images, and therefore many techniques have been derived for "gating" or otherwise motion correcting the data (e.g., to time the data with quasi-periodic motion of the patient) and for other patient motion detection and correction. Various gating and other motion compensation techniques can be used to gate each of the PET and MR data; however, the inventors have recognized that currently available techniques for motion compensating for simultaneously acquired PET and MR data are inadequate.

[0019] Current motion compensation techniques for hybrid PET-MR imaging systems are a) hardware-based, b) MR-data-based, c) based on a combination of MR data and PET data, or d) involve separate gating or motion correction of the MR data and PET data using different techniques. Hardware solutions for gating involve the use of external hardware, such as a respiratory bellows for respiratory gating and / or an ECG monitoring system for cardiac gating. These systems require time-consuming setup and create additional points of failure for the imaging modality. For example, the use of a ventilator bellows and / or ECG monitoring requires additional setup time and may require some effort on the part of the medical professional to acquire physiological data of sufficient quality to perform reliable gating. In addition, the scan may be interrupted if the external physiological monitoring is interrupted for any reason.

[0020] Available data-driven methods for gating involve some form of MR motion detection or involve gating the PET data separately from the MR data or otherwise performing separate motion compensation. Data-driven motion compensation is a class of techniques for identifying respiratory and / or cardiac motion based on data acquired from an imaging modality. In currently available data-driven motion detection for PET-MR systems, the acquired MR data is used to motion correct the MR data set, and / or the acquired PET data is analyzed for motion correction of the PET data set. Thus, separate motion characteristics are used for motion compensation in MR and PET, and such separate motion signals are derived in different ways. Data-driven PET motion correction methods are typically retrospective, meaning that the motion information is decrypted after data acquisition, and correction occurs in post-processing. In contrast, many MR motion detection and correction methods are prospective and occur simultaneously with or prior to MR data acquisition (sometimes referred to as "triggering"), although some retrospective MR gating and other motion compensation methods are currently available.

[0021] The inventors have recognized that methods of providing separate motion correction for PET data and MR data produce uncoordinated PET and MR images and a lack of coordination between the resulting images from the separate modalities, which may hinder the effectiveness of joint imaging. Additionally, detecting and correcting motion separately in both modalities requires a large amount of processing power dedicated to the two separate processes. For example, where data-driven MR gating methods are used, gating typically takes up scan time and thus increases the amount of time required for MR data acquisition, and also requires additional work by the technician operating the system.

[0022] In view of the shortcomings of the prior art solutions and the problems in the related art recognized by the inventors, the disclosed systems and methods have been developed, which utilize only PET data-driven motion analysis methods to perform retrospective motion compensation on both PET data and MR data. That is, one or more characteristics of patient motion are extracted from the PET raw data and applied to process both PET data and MR data. The motion characteristics can be described, for example, by a gating signal or by a motion vector indicating optical flow motion. The disclosed methods and systems provide a device-free method for motion detection and correction of two modalities using a single set of motion characteristics (such as a single gating signal), thereby maintaining image alignment while avoiding the use of external hardware and wasted MR scanning time (which may occur using MR data-driven techniques of navigators or interleaved sequences). Since both PET and MR modalities use the same motion characteristics, image alignment on all gates is ensured, and since the motion information is consistent, so is the consistency between all MR image sequences. The corrected images (including gated MRI and gated PET images) generated as described herein can be processed into a single motion-corrected image with an increased signal-to-noise ratio and lesion detection capability.

[0023] Figure 1 1 is a diagram of a hybrid PET-MR imaging system 100 according to an exemplary embodiment. As is well known in the related art, the PET-MR imaging system 100 includes a patient couch 102 that can be received by two imaging modalities, including a PET imaging portion 104 and an MR imaging portion 116. The couch 102 is configured to occupy multiple couch positions to facilitate imaging. Each imaging portion 104, 116 or imaging modality generates an image data set, wherein the PET imaging portion 104 generates PET data 106 and the MR imaging portion 116 generates MR data 107. A motion correction controller 108 is communicatively coupled to the imaging modalities 104 and 116 and is configured to receive raw data 106 and 107 from each imaging modality and generate corresponding motion-corrected data sets 124 and 125. The motion-corrected data sets 124 and 125 can be displayed on a display device 126 that is communicatively coupled to a computing system that operates as the motion correction controller 108. For example, motion segmented or gated PET data 124 can be displayed as a PET image, and motion gated MR data 125 can be displayed as an MR image, and the images can be displayed in a time-dependent manner. Alternatively or additionally, the gated PET data 124 and the gated MR data 125 can be processed into a single image, which can be displayed on a display 126.

[0024] The motion correction controller 108 is configured to generate at least one motion characteristic 113 describing patient motion and used for motion correction of the PET and MR data to produce corrected PET data 124 and gated MR data 125. As described herein, the motion characteristics 113 are generated based on the PET data, based on any of a variety of data-driven techniques for deriving quasi-periodic motion information based on the acquired PET data. The motion characteristics 113 may describe respiratory motion or cardiac motion, and in some embodiments, two sets of motion characteristics may be generated, one set describing the patient's cardiac motion and the other describing the patient's respiratory motion.

[0025] The motion correction controller 108 may include a pre-processor module 110, a PET motion signal analyzer module 112 that derives motion characteristics from the PET data, and a gate signal 13 ( Figure 5 )) A retrospective gating module 114 that generates motion correction data 124, 125. In other embodiments, the module 114 may alternatively perform a different type of motion correction, such as an optical flow module that retrospectively performs optical flow motion extraction. The motion correction controller 108 also includes a processing system utilizing a processor 118 and a storage system including a memory 120 for storing software (including software modules 110, 112, 114) and data generated thereby. The preprocessor module 110 is configured to receive the PET data set 106 and the MR data set 107, and appropriately associate and / or organize the raw data from the modality. In certain embodiments, the preprocessor module 110 may be configured to correspond the raw data sets 106 and 107 to the bed position of the patient bed 102 based on the acquisition time and / or associate the two data sets 106 and 107 with each other. For example, the PET-MR system may have eight bed positions, and each raw data set 106, 107 may be associated with one of the eight bed positions. In other embodiments, a different number of bed positions may be utilized. The raw PET data 106 and the raw MR data 107 can be acquired using a common reference clock and can therefore be easily associated with the bed position and / or with each other based on the acquisition time. Alternatively, reference timing data can be generated for one or both of the imaging modalities 104, 106 that can allow the raw PET data 106 and the raw MR data 107 to be associated.

[0026] The PET motion signal analyzer module 112 is executed by the processor 118 to extract motion information from the PET data to determine a motion characteristic 113. The retrospective gating module 114 is then executed to process the raw PET data 106 based on the motion characteristic 113 to generate motion-corrected PET data 124, which may be gated PET raw data. The retrospective gating module 114 also processes the raw MR data 107 based on the motion characteristic 113 to generate motion-corrected MR data 125, which may be gated MR raw data. The data is then used to reconstruct PET images and MR images, which may then be visually correlated and / or combined and presented on a display 126 to facilitate diagnosis and / or patient treatment.

[0027] Figure 2 and Figure 3 A flow chart is provided showing an exemplary method 200 for processing data from a hybrid PET-MR system to generate a gated data set. Figure 6 and Figure 7 Additional exemplary embodiments of a method for motion compensation of a hybrid PET-MR system are also disclosed. PET data is acquired at step 204, and MR data is acquired at step 216. The PET and MR data are acquired simultaneously and are time-correlated so that the relative time of the PET raw data and the MR raw data is known. This can be achieved by saving time for the acquired PET data and MR readout using a common reference clock. Steps are then performed at step 212 to extract motion information from the PET raw data in order to generate a gating signal. Various methods for retrospectively extracting motion from PET data can be utilized, which methods are known in the relevant art, examples of which are described herein. The motion characteristic 113 is then used to gate the PET data at step 214a, and the MR data is gated at step 214b. At step 218a, a PET image is reconstructed based on the gated PET data, and at step 218b, an MR image is reconstructed based on the gated MR data.

[0028] In one embodiment, PET data acquisition is performed continuously for a single bed position, while MR data is acquired over multiple MR pulse sequences that are performed at the bed position and are time-correlated with the acquired PET data. FIG. 4 schematically illustrates this concept, where PET acquisition is performed continuously for a single bed position, and MR data is acquired over multiple sequential pulse sequences that are performed at the bed position. In one embodiment, the multiple MR pulse sequences are identical, so that all acquisition settings for each pulse are identical. For example, settings such as sequence, repetition time, echo time, flip angle, echo train length, matrix size, number and thickness of slices, field of view, and other PSD settings can be the same between MR pulses. Figure 4A A simple case of repeating the same MR acquisition multiple times is shown. The sequence is acquired repeatedly to allow retrospective gating of the MR data based on motion determined from the PET data.

[0029] In other embodiments, data processing can be performed for different post-sequences, such as where each range of the MR raw data space (k-space) can be correlated and analyzed across all pulses. For example, the acquisition sequence can also be designed for motion, such as by applying golden angle acquisition. In other embodiments, pseudo-random sampling can be used with a reconstruction algorithm based on compressed sensing. Figure 4B Schematically illustrated are embodiments in which MR data acquisition is performed continuously and simultaneously with PET data acquisition, where continuous pseudo-random sampling of k-space is used for motion-corrected compressed sensing methods. These types of acquisitions are designed to reconstruct subsets of the data (e.g., for gating or other motion-corrected reconstruction methods). Their acquisition times are longer due to the oversampled portions of k-space. As described above, in all such embodiments, the relative timing of the PET acquisition and the MRI raw data must be known to enable retrospective gating across modalities.

[0030] A variety of methods can be used to extract motion from short duration PET data, such as principal component analysis (PCA), independent component analysis (ICA), respiratory phase analysis, amplitude-based gating, static gating, or other methods. Typically, PET data is divided into short durations, each of which is processed and compared to derive motion information. In some examples, the PET raw data is temporally segmented and processed to extract motion information. In other embodiments, motion information can be extracted from image space, where an image volume of short duration is reconstructed, as shown in step 211.

[0031] Each short duration PET data set is analyzed to identify multiple motion states, such as respiratory motion states or cardiac motion states. The motion states can be described or classified in various ways to describe quasi-periodic motion. The motion states are then combined to generate a gating signal that describes quasi-periodic motion over time, such as respiratory motion or cardiac motion. Figure 3 An embodiment of using PCA to identify gating signals based on PET data is shown.

[0032] In one embodiment, PCA is used to generate a component containing motion information describing the patient's cardiac or respiratory motion. An example of gating using PCA to extract motion information from PET data is described in Thielemans, Kris et al., "Device-less gating for PET / CT using PCA.", Proceedings of the IEEE Nuclear Science Symposium, IEEE, 2011, which is hereby incorporated by reference in its entirety.Other PET data-driven gating methods may be used, such as those described in: G. Klein, B. Reutter, E. Botvinick, T. Budinger, and R. Huesman, "Finescale motion detection using intrinsic list mode PET information," Workshop on Mathematical Methods in Biomedical Image Analysis, Los Alamitos, California, USA, 2001, Volume 0, Pages 71-78, IEEE Computer Society; Ralph A. Bundschuh, Axel Martinez-Moeller, Markus Essler, Marfa Jose J. Martinez, Stephan G. Nekolla, Sibylle I. Ziegler, and Markus Schwaiger, "Postacquisition detection of tumor motion in the lung and upper abdomen using list-mode PET data: a feasibility study.", Journal of Nuclear Medicine: Official Publication, Society of Nuclear Medicine, Volume 48, Issue 5, Pages 758-763, May 2007; Florian Blither, Mohammad Dawood, Lars Stegger, Frank Wlibbeling, Michael Schafers, Otrnar Schober; Klaus P. Schlifers, "List Mode Driven Cardiac and Respiratory Gating in PET," Journal of Nuclear Medicine, Vol. 50, No. 5, pp. 674-681, May 2009; and Jianfeng He, Graeme, O'Keefe, Sylvia, Gong, Gareth Jones, Tim Saunder, Andrew M. Scott and Moshi Geso, "A Novel Method for Respiratory Motion Gated With Geometric Sensitivity of the Scanner in 3D PET," IEEE Transactions on Nuclear Science, Vol. 55, No. 5, pp. 2557-2565, October 2008, each of which is incorporated herein by reference in its entirety.Another method for extracting motion date information from short duration PET data is described in US Pat. No. 9,471,976, which is hereby incorporated by reference in its entirety.

[0033] Figure 5 An exemplary gating signal is depicted, which in this example describes the patient's respiratory motion, but in other examples may describe cardiac motion. In one embodiment, the gating signal is a respiratory waveform 13a.

[0034] The respiratory waveform 13a can be used for amplitude-based gating, where the gated MR image data is populated with MR raw data views corresponding to the most recent waveform amplitude. The MR raw data contains many data readouts (sometimes called "views"). These views correspond to the MR raw data space (k-space). For each view in a particular gated segment or bin (which would be an amplitude range in the case of amplitude-based gating), the MR raw data is processed to find a corresponding k-space view from multiple MR acquisitions that has an amplitude range closest to a target bin. For example, one cycle of the respiratory waveform can be divided into a preset number of target bins, such as eight target bins, where one or more of the bins can correspond to the amplitude at peak inspiration, and one or more different bins can correspond to a static period q, which is a period of relative inactivity between the end of exhalation and the beginning of the next inspiration.

[0035] A second exemplary gating signal is also Figure 5 , and the respiratory motion is represented in the respiratory cycle phase, as shown in the exemplary respiratory phase signal 13b. In this example, the respiratory phase gating signal 13b is configured so that 0% is assigned to the peak inspiration and increases to 100% within the respiratory cycle at the last measurement immediately before the next peak. In the depicted example, the respiratory cycle restarts at each peak, with the respiratory phase signal 13b (which is an exemplary gating signal) reset to the 0% phase.

[0036] Phase-based gating using the respiratory phase signal 13b is performed similarly to the amplitude-based gating described above. For each view in the gated target phase or bin, a gating method is performed to locate the corresponding k-space view from the multiple acquisitions that has the closest phase percentage to the target phase for the particular bin. Thus, the respiratory phase between 0% and 100% is divided into target phase bins, such as 12.5% ​​phase segments each representing a respiratory cycle (i.e., the phase is divided into eight bins).

[0037] The concept is Figure 6, where a single target phase or bin of the gated MR data is composed of views from multiple different MR acquisitions acquired on each of the multiple sequential pulse sequences. By identifying the MR acquisition with the closest phase percentage corresponding to the time when one of the views from the multiple MR acquisitions was acquired, each view of the gated MR raw data space is populated with that view. This means that each MR acquisition is likely to contribute in part to the single target phase or bin of the gated MR raw data output by the retrospective gating module 114 (see Figure 1 ).

[0038] As described above, the gated MR raw data 125 is then used by a medical professional to reconstruct an MR image or display and analysis. The PET data is also gated using a gating signal 13 (e.g., a respiratory waveform 13a or a respiratory phase waveform 13b) to generate gated PET raw data. Since the acquired MR data and the acquired PET data are gated using the same gating signal 13, the resulting gated raw data sets will correspond in time, and the resulting reconstructed images will also correspond in time. Additionally, multiple MR pulse sequences can be used during a single PET acquisition. That is, gating or other motion compensation methods can be applied to multiple different MR pulse sequences within the same PET acquisition. This allows MR motion-corrected images to be created at a consistent reference position to align all images. Figure 4C 3. Represents three different MR sequences that are repeatedly acquired for motion correction processing. Each of the sequences will provide a motion corrected image with a common reference position. A common reference frame (or gating) can be applied to allow alignment between multiple different MR sequences. In light of this disclosure, one of ordinary skill in the art will appreciate that cardiac gating can be performed similarly to the respiratory example described herein.

[0039] Figure 7is a flow chart illustrating an alternative motion compensation method for gating. Here, the set of motion characteristics is a set of motion vectors 14 describing patient motion in the PET data. Motion extraction is performed at step 213' to extract optical flow motion from the PET image volume. Optical flow methods define a volumetric summary of motion relative to a reference frame (e.g., a first time point). This is a quantitative measure of motion because it is in the spatial dimension. This is in contrast to a gated signal, which represents a qualitative (or relative) expression of motion. For example, existing optical flow methods can be applied to determine 3D motion vectors over the entire PET image volume as a function of time relative to an initial starting frame. This motion information is then directly incorporated into the PET and MR reconstructions generated at steps 218a and 218b. PET and MR reconstruction with motion vectors are exemplified and described in: Lamare, F. et al., "List-mode-based reconstruction for respiratory motion correction in PET using non-rigid body transformations." Physics in Medicine and Biology, 52.17 (2007): 5187; Chan, Chung et al., "Non-rigid event-by-event continuous respiratory motion compensated list-mode reconstruction for PET." IEEE Transactions on Medical Imaging, 37.2 (2017): 504-515; Batchelor, P. G., Atkinson, D., Irarrazaval, P., Hill, D. L. G., Hajnal, J., & Larkman, D. (2005); Matrix description of general motion correction applied to multishot images. Magnetic Resonance in Medicine, 54(5), 1273-1280; and Zaitsev, M., Maclaren, J., & Herbst, M. (2015). Motion artifacts in MRI: a complex problem with many partial solutions. Journal of Magnetic Resonance Imaging, 42(4), 887-901.

[0040] This written description uses examples to disclose the present invention, including the best mode, and also enables those skilled in the art to perform and use the present invention. Certain terms are used for brevity, clarity and ease of understanding. In addition to the requirements of the prior art, unnecessary restrictions should not be inferred therefrom because such terms are only used for descriptive purposes and are intended to be broadly understood. The patent scope of the present invention is defined by the claims and may include other examples that occur to those skilled in the art. If these other examples have features or structural elements that are not different from the literal language of the claims, or if they include equivalent features or structural elements that are not substantially different from the literal language of the claims, these other examples are intended to be within the scope of the claims.

Claims

1. A method of processing data from a hybrid PET-MR imaging system, the method comprising: Acquiring a positron emission tomography data set within a certain period of time, wherein the positron emission tomography data set is affected by quasi-periodic motion of the patient; acquiring magnetic resonance data during the time period such that an acquisition time of the magnetic resonance data relative to the positron emission tomography data set is known; determining a characteristic of quasi-periodic motion of the patient based on the positron emission tomography data set; as well as The magnetic resonance data is processed based on characteristics of the quasi-periodic motion of the patient.

2. The method of claim 1, wherein the characteristic of the quasi-periodic motion of the patient is a gating signal, wherein the gating signal describes the quasi-periodic motion of the patient evident in the PET dataset.

3. The method of claim 2, wherein the gating signal is determined retrospectively from the positron emission tomography dataset using principal component analysis. The method of claim 2 , wherein the gating signal indicates a phase of the quasi-periodic motion of the patient. The method of claim 2 , wherein the gating signal is indicative of an amplitude of the quasi-periodic motion of the patient. The method of claim 2 , wherein the gating signal describes a respiratory motion of the patient or a cardiac motion of the patient.

7. The method of claim 1 , wherein acquiring the magnetic resonance data comprises acquiring magnetic resonance raw data at each of a plurality of magnetic resonance pulse sequences performed throughout the time period, and wherein the positron emission tomography data set comprises continuously acquiring positron emission tomography data during the time period.

8. The method of claim 7, wherein each magnetic resonance pulse sequence in the plurality of magnetic resonance pulse sequences is identical.

9. The method of claim 2, wherein the PET data set and the MR data correspond to a couch position, and wherein the gating signal corresponds to the couch position.

10. The method of claim 1, wherein a common reference clock is used to acquire the PET data set and the magnetic resonance data.

11. The method of claim 10, wherein the characteristic of the quasi-periodic motion of the patient is a gating signal describing a respiratory phase of the patient derived from the positron emission tomography dataset, and wherein processing the magnetic resonance data based on the gating signal comprises selecting the magnetic resonance data having corresponding phase percentages and acquisition times.

12. The method of claim 1, wherein the characteristic of the quasi-periodic motion of the patient is a set of motion vectors indicative of optical flow motion of the positron emission tomography data set.

13. A hybrid PET-MR imaging system, comprising: a positron emission tomography imaging section configured to generate positron emission tomography data of the patient; a magnetic resonance imaging section configured to generate magnetic resonance data of the patient; A controller, the controller being configured to: acquiring a positron emission tomography data set within a certain period of time from the positron emission tomography imaging portion, wherein the positron emission tomography data set is affected by quasi-periodic motion of the patient; acquiring magnetic resonance data having the magnetic resonance imaging portion during the time period such that an acquisition time of the magnetic resonance data relative to the positron emission tomography data set is known; determining a characteristic of quasi-periodic motion of the patient based on the positron emission tomography data set; and The magnetic resonance data is processed based on characteristics of the quasi-periodic motion of the patient.

14. The system of claim 13, wherein the characteristic of the quasi-periodic motion of the patient is a gating signal, wherein the gating signal describes the quasi-periodic motion of the patient evident in the PET dataset.

15. The system of claim 14, wherein the gating signal is determined retrospectively from the positron emission tomography data set using principal component analysis.

16. The system of claim 14, wherein the gating signal indicates a phase of the quasi-periodic motion of the patient and / or an amplitude of the quasi-periodic motion of the patient.

17. The system of claim 13, wherein acquiring the magnetic resonance data comprises acquiring magnetic resonance raw data at each of a plurality of magnetic resonance pulse sequences performed throughout the time period, and wherein the positron emission tomography data set comprises continuously acquiring positron emission tomography data during the time period.

18. The system of claim 14, further comprising a couch configured to support the patient to be imaged by the PET imaging portion and the MRI imaging portion and further configured to occupy a plurality of couch positions; wherein the positron emission tomography data set and the magnetic resonance data correspond to one bed position of the plurality of bed positions, and wherein the gating signal corresponds to the one bed position of the plurality of bed positions.

19. The system of claim 13, wherein the characteristic of the quasi-periodic motion of the patient is a gating signal describing a respiratory phase of the patient derived from the positron emission tomography dataset, and wherein processing the magnetic resonance data based on the gating signal comprises selecting the magnetic resonance data having corresponding phase percentages and acquisition times.

20. The system of claim 13, wherein the characteristic of the quasi-periodic motion of the patient is a set of motion vectors indicative of optical flow motion of the PET data set.

Images