Reconstruction parameter determination for synthetic magnetic resonance image reconstruction

Abstract

A medical system (100, 300) is disclosed herein, the medical system comprising a memory (110) storing machine executable instructions (120) and an anatomy detection module (122), and a computing system (104). Execution of the machine executable instructions causes the computing system to: receive (200) a set of magnetic resonance images (124) describing a field of view (109) of an object (318) acquired according to a synthetic magnetic resonance imaging protocol; receiving (202) an anomaly indicator (126) from the anomaly detection module in response to inputting at least one magnetic resonance image of the set of magnetic resonance images into the anomaly detection module; determining (204) a set of reconstruction parameters (128) using the anomaly indicator; and reconstructing (206) a composite magnetic resonance image (134) from the set of magnetic resonance images and the set of reconstruction parameters.

Description

Technical Field

[0001] The present invention relates to magnetic resonance imaging, and in particular to synthetic magnetic resonance imaging. Background Art

[0002] Magnetic resonance imaging (MRI) scanners use a large static magnetic field to align the nuclear spins of atoms as part of the process of producing images of the patient's body. This large static magnetic field is called the B0 field or main magnetic field. MRI can be used to spatially measure and image various quantities or properties of an object. One difficulty in performing magnetic resonance imaging is the time required to acquire the k-space data necessary to reconstruct the magnetic resonance image. When acquiring k-space data, the images can be weighted differently using different imaging protocols. For example, magnetic resonance images can be weighted to show spatial variations in proton density, T1, or spatial variations in T2 or T2 stars. These different weightings are often referred to as "contrast" or "contrast types." Doctors and other medical professionals are accustomed to viewing a specific "contrast." In recent years, other imaging techniques have been developed, such as magnetic resonance imaging fingerprinting (MRF) or synthetic magnetic resonance imaging (SyntAC or 3D-QALAS). In these techniques, k-space data is used to acquire images, which can be used to construct synthetic or mapped images that simulate or reproduce a typical or canonical "contrast."

[0003] In synthetic MR imaging, a set of MR images describing (depicting the object's) field of view is received. This set of MR images essentially represents a standardized acquisition (for a particular synthetic MR reconstruction) that provides data that can be used to calculate spatially varying quantitative maps describing the object. These quantitative maps then provide enough information to simulate other MR images using the MR signal equation. Two techniques that achieve this are Magnetic Resonance Fingerprinting (MRF) and Synthetic MR Imaging (SyntAc or 3D-QALAS).

[0004] U.S. patent application publication US2021174937A1 discloses a diagnostic machine for performing medical diagnosis on raw medical imaging data generated by a medical imaging machine, which is contrary to the medical diagnosis performed using medical images reconstructed from raw medical imaging data under normal circumstances. In operation, the raw diagnostic engine includes a medical imaging diagnosis controller that implements a dimensionality reduction preprocessor that is used to select or extract one or more dimensionality reduction feature vectors from the raw medical imaging data, and the controller also implements a raw diagnostic artificial intelligence engine that is used to present a diagnostic evaluation of the raw medical imaging data represented by (one or more) dimensionality reduction feature vectors. The medical imaging diagnosis controller can also control the communication of the diagnostic evaluation of the raw medical imaging data (e.g., display, printing, emailing, text generation, etc.).

[0005] The publication "Autoencoders for Unsupervised Anomaly Segmentation in Brain MR Images A Comparative Study" by Baur et al. (arXiv 2004.03271, https: / / doi.org / 10.48550 / arXiv.2004.03271) discloses a detailed method for using variational autoencoders to segment abnormalities in MRI brain images. Summary of the invention

[0006] The invention provides a medical system, a computer program and a method in the independent claims. Embodiments are given in the dependent claims.

[0007] In various synthetic MRI techniques, one can typically reconstruct various types of synthetic MRI images that are modeled after a real MRI acquisition. This includes the ability to model the selection of various reconstruction parameters, such as the echo time (TR) or repetition time (TR) set when acquiring k-space data, or even the type of pulse sequence used to acquire k-space data. The selection of these parameters affects the image contrast within the image and ultimately affects the type of anatomical structures that are visible in the synthetic MRI image.

[0008] Therefore, in order to enable or optimally see the abnormality in the synthetic magnetic resonance image, repeated trials may be required to correctly set the reconstruction parameters. The abnormality may include anatomical abnormalities or abnormal tissue pathology or abnormal tissue structure or properties (e.g., white matter hyperdensity). An embodiment may provide a means to automatically determine a set of reconstruction parameters for the synthetic magnetic resonance image. This can be achieved by using an anomaly detection module that is configured to receive the set of magnetic resonance images (or a subset of these images) and output an abnormality indicator. The abnormality indicator is then used to determine a set of reconstruction parameters that are optimized or selected to display the detected abnormality in the reconstructed synthetic magnetic resonance image.

[0009] The concept of generating synthetic contrast interactively or by providing a set of parameters for a specific map to the radiologist to review several images for evaluation and identification of pathology results in a time-consuming workflow.

[0010] Examples can have the following benefits: by performing anomaly detection on quantitative sequences and selecting predefined standard synthetic sequences or dedicated parameters for synthetic contrast based on its results, pathologies can be visualized in an optimal manner allowing rapid visual assessment, thereby accelerating the workflow of reading images in the context of synthetic MRI and thereby enhancing the workflow of radiologists.

[0011] The examples may contain one or more of the following subsequent work steps, which are described in more detail:

[0012] 1. Anomaly Detection: AI-based anomaly detection using quantitative SyntAc images can be done by different approaches: (i) anatomically driven anomaly detection, where the symmetry contrast between the left and right hemispheres is analyzed, or other known approaches, such as variational autoencoders. In particular for SyntAc, it is also possible to derive per-voxel anomalies by deviating from normal tissue properties in the brain, such as Figure 1 . All methods produce a map indicating voxel-by-voxel abnormalities and may be post-processed by filtering techniques to remove artifacts in the detected abnormalities (e.g., to remove individual abnormal voxels). For this optional step, standard image processing algorithms (such as thresholding, smoothing, or filtering) may be applied. The generation of symmetric contrast involves: (a) a grid-based image deformation of the segmented scan to a symmetric template; (b) an anatomically symmetric image obtained by L / R flipping; (c) calculating a difference image of the results of (a) and (b); (d) using the (symmetric) difference image and the symmetric image of the brain as input to a convolutional neural network that produces (1) areas indicating residual abnormal contrast changes, and (2) the side on which the damage occurred. Symmetric contrast is the contrast distribution of a symmetric image (that has been made symmetric from its original asymmetric version).

[0013] 2. Derivation of optimal sequence parameters for synthetic contrast generation: Based on the results of step 1, the parameters (TE, TR and TI) for synthetic contrast generation are determined to maximize the grayscale value range in the suspected abnormal area, thereby obtaining a clear abnormal presentation.

[0014] 3. As an option, the parameters may not be optimized for the entire volume, but individually for each suspicious pathology in the image. The final composite image can then be merged via interpolation so that the presentation of each suspicious pathology will be optimal.

[0015] In one aspect, the present invention provides a medical system, the medical system comprising a memory storing machine executable instructions and an anatomical detection module. The medical system also comprises a computing system. Execution of the machine executable instructions causes the computing system to: receive a set of magnetic resonance images describing a field of view of an object acquired according to a synthetic magnetic resonance imaging protocol. The synthetic magnetic resonance imaging protocol may specify acquisition parameters for acquiring the set of magnetic resonance images and / or a reconstruction technique for reconstructing the set of magnetic resonance images.

[0016] Execution of the machine executable instructions further causes the computing system to: receive an anomaly indicator from the anomaly detection module in response to inputting at least one magnetic resonance image in the set of magnetic resonance images into the anomaly detection module. In different examples, the anomaly indicator can take different forms. In one case, the anomaly indicator can simply indicate the presence of an anomaly within the set of magnetic resonance images as a label. For example, it can indicate that the set of magnetic resonance images contains a particular type of tumor or abnormal tissue structure.

[0017] In this case, the anomaly indicator may not provide an indication of where the anomaly is located within the field of view. In other instances, the anomaly indicator may provide a location or bounding box within the field of view to indicate where the anomaly is located. In other examples, the anomaly indicator may provide an indicator or bounding box for the location of the anomaly, and a label identifying the anomaly type or anomaly type.

[0018] Execution of the machine executable instructions further causes the computing system to: determine a set of reconstruction parameters using the abnormality indicator. The set of reconstruction parameters used herein are instructions or details for reconstructing a synthetic magnetic resonance image based on the set of magnetic resonance images. For example, in the well-known SyntAC or 3D-QALAS techniques, known T1 values, T2 values, and proton density values ​​are used to create a synthetic contrast-weighted image. In this case, the reconstruction parameters may be, for example, TE time and TR time. This may, for example, depend on various models used to reconstruct the synthetic magnetic resonance image.

[0019] However, the set of reconstruction parameters may be selected so that the synthetic magnetic resonance image emphasizes or displays the anomaly in a readily achievable manner. The set of reconstruction parameters may be determined in a variety of ways. For example, if the anomaly indicator includes an identification of the type of anomaly, the set of reconstruction parameters may simply be called from a lookup table or database. In other examples, if the anomaly indicator indicates a location, the reconstruction of the synthetic magnetic resonance image may be performed in an iterative manner so that within the identified region where the anomaly is located, various image parameters (e.g., contrast distribution) may be made to match a predetermined range or set of values. That is, the determination of the set of reconstruction parameters may be implemented, for example, by machine executable instructions.

[0020] Execution of the machine executable instructions further causes the computing system to: reconstruct a synthetic magnetic resonance image based on the set of magnetic resonance images and the set of reconstruction parameters. This embodiment may be beneficial because the synthetic magnetic resonance image may have its reconstruction controlled or directed such that anomalies detected by the anomaly detection module may be displayed in a manner that emphasizes or displays the abnormal structure. For example, a region containing an anomaly may have its contrast maximized such that the abnormal structure is presented in detail on a particular display or graphic type.

[0021] In another embodiment, the field of view describes the right and left hemispheres of the subject's brain. The abnormality indicator includes one or more anatomical abnormality locations in the field of view. The abnormality detection module is also configured to spatially locate the one or more abnormalities in the subject's brain using a symmetric contrast between the right and left hemispheres of the subject's brain. It is expected that the anatomical structures of the left and right hemispheres of the subject's brain may have similarities. Abnormal structures such as tumors may cause a large asymmetry between the right and left hemispheres of the subject's brain. This type of abnormality can be identified using a symmetric contrast between the right and left hemispheres of the brain.

[0022] In another embodiment, the abnormality indicator includes one or more anatomical abnormality locations in the field of view. The abnormality detection module includes an autoencoder neural network, which is configured to output an autoencoded image for each of the at least one magnetic resonance image in the set of magnetic resonance images. Receiving the abnormality indicator in response to inputting the at least one magnetic resonance image in the set of magnetic resonance images into the abnormality detection module includes: receiving the autoencoded image in response to inputting the at least one magnetic resonance image in the set of magnetic resonance images into the autoencoder neural network. Receiving the abnormality indicator in response to inputting the at least one magnetic resonance image in the set of magnetic resonance images into the abnormality detection module also includes: determining the one or more anatomical abnormality locations by comparing the autoencoded image of each of the at least one magnetic resonance image in the set of magnetic resonance images with the at least one magnetic resonance image in the set of magnetic resonance images.

[0023] In this embodiment, an autoencoder is used to generate an image in response to receiving at least one magnetic resonance image in the set of magnetic resonance images as input into an autoencoder network. It is expected that the autoencoder takes an image and then generates an identical image. If the autoencoder is trained to use only anatomically correct images, the autoencoder may make errors when encountering abnormalities. Therefore, the comparison between the input image and the output image can provide a means to easily identify anatomically abnormal areas.

[0024] An autoencoder can be any autoencoder architecture that takes an image as input and outputs an autoencoder version of the same image as output. There may be multiple inputs and outputs for each image. Or a series of autoencoders, one for each image in the set of images. For example, the autoencoder can be a fully connected autoencoder, a sparse autoencoder, a deep (fully connected) autoencoder, a convolutional autoencoder, or a variational autoencoder.

[0025] The training data for these autoencoders can be a set of magnetic resonance images that do not contain abnormalities. The images can be input into the autoencoder during training, and the output can be compared to the input images.

[0026] In another embodiment, execution of the machine executable instructions further causes the computing system to: iteratively change the set of reconstruction parameters and reconstruct the synthetic magnetic resonance image to adjust the image contrast of the one or more anatomical abnormality locations in the image relative to their surroundings to have a predetermined image contrast range. In this embodiment, the determination of the set of reconstruction parameters is performed simultaneously with the reconstruction. Once the abnormality locations are determined, predetermined image contrast ranges can be defined for these locations. The reconstruction can start with a seed value for the reconstruction parameters and then change the seed value until the predetermined image contrast range within these regions is obtained.

[0027] In another embodiment, the abnormality indicator indicates a plurality of anatomical abnormality locations. The synthetic magnetic resonance image is reconstructed for each of the plurality of anatomical abnormality locations, thereby producing a set of synthetic magnetic resonance images. In this embodiment, the abnormality indicator indicates or locates a plurality of regions in the field of view where an abnormality may exist. Because there are a plurality of locations, there is a synthetic magnetic resonance image reconstructed for each of these locations. For example, an image having a predetermined image contrast range can be produced for each of these locations.

[0028] In another embodiment, execution of the machine executable instructions further causes the computing system to construct a composite image from a set of synthetic magnetic resonance image locations by: including the anatomical abnormality location from each synthetic magnetic resonance image in the set of synthetic magnetic resonance images; and using the set of synthetic magnetic resonance images to blend pixel values ​​between the anatomical abnormality locations from each synthetic magnetic resonance image in the set of synthetic magnetic resonance images. Displaying the set of synthetic magnetic resonance images may be inconvenient. In this embodiment, the set of synthetic magnetic resonance images is used to make a single composite image. The composite image is formed by including the anatomical abnormality location from each of these images, i.e., taking the abnormal images from the field of view and pasting them together into the composite image. Between these anatomical abnormality locations, the pixel values ​​are then blended. For example, in different examples, this can take different forms. Providing images of multiple anatomical abnormality locations can average or blend their values ​​so that there is a smooth transition from one image contrast range to the next image contrast range.

[0029] In another embodiment, execution of the machine executable instructions further causes the computing system to: display the composite image on a graphical user interface. Execution of the machine executable instructions further causes the computing system to: receive a selection of an anatomical abnormality location within the composite image from the graphical user interface, and then display the synthetic magnetic resonance image including the selected anatomical abnormality location. This embodiment provides a means to select a synthetic magnetic resonance image from the set of synthetic magnetic resonance images. However, the composite image displayed on the graphical user interface may be different or unusual to a physician or other healthcare provider. They are then able to select a specific anatomical abnormality location, which is referred to herein as a selection of an anatomical abnormality location, which is then used to select a specific synthetic magnetic resonance image from a set of synthetic magnetic resonance images. In this way, a physician or other healthcare provider is able to use the composite image to easily select an image for review.

[0030] In another embodiment, the anatomical detection module comprises a convolutional neural network configured to output an abnormality classification as the abnormality indicator in response to receiving the at least one magnetic resonance image in the set of magnetic resonance images. In this embodiment, rather than assigning a location to the set of magnetic resonance images, a label is assigned to the set of magnetic resonance images. This may, for example, indicate a certain abnormality type or a certain tumor type. This knowledge of a particular abnormality type may then be used to select a set of reconstruction parameters. For example, based on previous experience or previous reconstructions, it may be known which reconstruction parameter in the set of reconstruction parameters may be used for a particular abnormality classification.

[0031] The convolutional neural network can use a neural network architecture configured to receive one or more images as input and output labels. For example, a deep convolutional neural network (DCCN) architecture can be used. The convolutional neural network can be trained by inputting a set of neural networks labeled with anomalies.

[0032] In another embodiment, the anatomical detection module comprises a variational autoencoder neural network having a latent space. The anatomical detection module is configured to output an anomaly classification as the anomaly indicator. Receiving the anomaly indicator in response to inputting the at least one magnetic resonance image in the set of magnetic resonance images into the anomaly detection module comprises: inputting the at least one magnetic resonance image in the set of magnetic resonance images into the variational autoencoder, and then determining the anomaly classification using an out-of-distribution detection method on the latent space of the variational autoencoder.

[0033] An autoencoder uses two neural network parts; the first encoding part encodes the input image into a vector in a so-called latent space. The vector in the latent space is then input to the decoder part of the autoencoder, which then produces an identical or very similar image if the variational autoencoder is properly trained. Variational autoencoders use additional constraints on the vectors in the latent space during training. Typically, there is a cost or penalty function assigned to the vectors in the latent space that forces the vector to be towards the center. This has the effect of forcing the positions of the vectors in the latent space to be in order. By monitoring the latent space during use of the variational autoencoder, it is possible to detect when an input image is outside of the training distribution.

[0034] A variational autoencoder configured to output an autoencoded image for each magnetic resonance image in the set of magnetic resonance images or a set of variational autoencoders, one variational autoencoder for each magnetic resonance image in the set of magnetic resonance images, may be used. The variational autoencoder or encoders may be trained to reproduce the magnetic resonance images without anomalies. Out-of-distribution techniques in the latent space of the variational autoencoder may then be used to detect anomalies.

[0035] In another embodiment, the reconstruction parameters are determined by searching a reconstruction parameter lookup table or a reconstruction parameter database using the abnormality classification. In these cases, the label or abnormality classification can be used to retrieve data from the reconstruction parameter lookup table or reconstruction parameter database. The reconstruction parameter lookup table or reconstruction parameter database can be populated by using previous experience of reconstructing images to appropriately display a specific abnormality (e.g., a tumor or other anatomical structure).

[0036] In another embodiment, reconstructing the synthetic magnetic resonance image includes determining a T1-related value (e.g., a T1 map), a T2-related value (e.g., a T2 map, a T2 star map, an R2 map, or an R2 star map), and a proton density for each voxel of the field of view by executing a voxel-by-voxel state of a selected signal intensity equation on the set of magnetic resonance images. Reconstructing the synthetic magnetic resonance image also includes reconstructing the synthetic magnetic resonance image by calculating a signal intensity value for each voxel using a reconstruction signal intensity equation, the reconstruction signal intensity equation taking as input the voxel-by-voxel T1-related value, the voxel-by-voxel T2-related value, the voxel-by-voxel proton density value, and the reconstruction parameters.

[0037] In another embodiment, the reconstruction parameters include echo time, repetition time and preferably inversion delay.

[0038] The above-described embodiments (involving construction of synthetic magnetic resonance images based on T1-related values, T2-related values, and proton density) may relate to synthetic magnetic resonance imaging or SyntAc. Synthetic MR provides a standardized quantitative MR sequence (SyntAc) (to be applied to a magnetic resonance imaging system). One of the claims is to derive a standard contrast (such as a T1-weighted image or a T2-weighted image) based on SyntAc in a post-processing step. However, this is currently done interactively, and finding the best contrast for a specific pathology is time-consuming and non-reproducible. The disclosure of the present invention proposes to automatically find the best derived contrast through an AI-based evaluation of quantitative SyntAc images, so that automatically suspected pathologies can be visually confirmed or rejected with very high confidence.

[0039] Automated detection of brain abnormalities is a high priority topic in MR, with the goal of achieving the “speed” route towards the MR North Star (“a personalized examination of each patient in less than 5 minutes, with fully automated workflow and decision support”).

[0040] Synthetic MR provides standardized quantitative MR sequences (SyntAc) to be applied. One of the benefits is that standard contrasts (such as T1-weighted images or T2-weighted images) can be derived from SyntAc in a post-processing step by freely choosing acquisition-related parameters (e.g., TE, TR, and TI).

[0041] In another embodiment, the set of magnetic resonance images forms a magnetic resonance fingerprint.The synthetic magnetic resonance image is reconstructed from the set of magnetic resonance images according to a magnetic resonance fingerprinting protocol.

[0042] During MRI fingerprinting, a so-called dictionary or MRI fingerprinting dictionary is used to compare with the MRI fingerprint of each voxel. This enables the calculation of various spatially related values ​​(e.g. T1 time, T2 time or proton density). These values ​​can be used to simulate the MRI signal and thus recalculate the synthetic MRI image.

[0043] In another embodiment, the medical system further comprises a magnetic resonance imaging system. The memory further comprises pulse sequence commands configured to control the magnetic resonance imaging system to acquire k-space data according to the synthetic magnetic resonance imaging protocol. Execution of the machine executable instructions further causes the computing system to: acquire the k-space data by controlling the magnetic resonance imaging system using the pulse sequence commands, and reconstruct the set of magnetic resonance images based on the k-space data preferably according to the synthetic magnetic resonance imaging protocol.

[0044] In another aspect, the present invention provides a method of operating a medical system. The method includes receiving a set of magnetic resonance images depicting a field of view of an object acquired according to a synthetic magnetic resonance imaging protocol. The method also includes receiving an anomaly indicator from an anomaly detection module in response to inputting at least one magnetic resonance image in the set of magnetic resonance images into the anomaly detection module. The method also includes determining a set of reconstruction parameters using the anomaly indicator. The method also includes reconstructing a synthetic magnetic resonance image based on the set of magnetic resonance images and the set of reconstruction parameters.

[0045] In another aspect, the invention provides a computer program comprising machine executable instructions for execution by a computing system. The computer program also comprises an anatomical detection module executed by the computing system.

[0046] Execution of the machine executable instructions causes the computing system to: receive a set of magnetic resonance images depicting a field of view of an object acquired according to a synthetic magnetic resonance imaging protocol. Execution of the machine executable instructions further causes the computing system to: receive an anomaly indicator from an anomaly detection module in response to inputting at least one magnetic resonance image from the set of magnetic resonance images into the anomaly detection module. Execution of the machine executable instructions further causes the computing system to: determine a set of reconstruction parameters using the anomaly indicator. Execution of the machine executable instructions further causes the computing system to: reconstruct a synthetic magnetic resonance image based on the set of magnetic resonance images and the set of reconstruction parameters.

[0047] It should be understood that one or more of the aforementioned embodiments of the present invention may be combined as long as the combined embodiments are not mutually exclusive.

[0048] Those skilled in the art will appreciate that aspects of the present invention may be implemented as devices, methods, or computer program products. Therefore, aspects of the present invention may take the form of a complete hardware embodiment, a complete software embodiment (including firmware, resident software, microcode, etc.), or an embodiment that combines software and hardware aspects, all of which are generally referred to herein as "circuits," "modules," or "systems." In addition, aspects of the present invention may take the form of a computer program product implemented in one or more computer-readable media having a computer executable code implemented thereon.

[0049] Any combination of one or more computer-readable media may be utilized. A computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. As used herein, "computer-readable storage medium" encompasses any tangible storage medium that can store instructions that can be executed by a processor or a computing system of a computing device. A computer-readable storage medium may be referred to as a computer-readable non-transient storage medium. A computer-readable storage medium may also be referred to as a tangible computer-readable medium. In some embodiments, a computer-readable storage medium may also store data that can be accessed by a computing system of a computing device. Examples of computer-readable storage media include, but are not limited to, floppy disks, magnetic hard disk drives, solid-state hard disks, flash memories, USB thumb drives, random access memories (RAM), read-only memories (ROM), optical disks, magneto-optical disks, and register files of computing systems. Examples of optical disks include compact disks (CDs) and digital versatile disks (DVDs), such as CD-ROMs, CD-RWs, CD-Rs, DVD-ROMs, DVD-RWs, or DVD-R disks. The term "computer-readable storage medium" also refers to various types of recording media that can be accessed by a computer device via a network or communication link. For example, data may be retrieved on a modem, on the Internet, or on a local area network. Computer executable code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0050] A computer readable signal medium may include, for example, a propagated data signal in which a computer executable code is implemented, in baseband or as part of a carrier wave. Such a propagated signal may take any of a variety of forms, including, but not limited to, electromagnetic, optical, or any suitable combination thereof. A computer readable signal medium may be any computer readable medium that is not a computer readable storage medium and that is capable of conveying, propagating, or transmitting a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0051] "Computer memory" or "memory" is an example of a computer-readable storage medium. Computer memory is any memory that can be directly accessed by a computing system. "Computer storage" or "storage device" is another example of a computer-readable storage medium. A computer storage device is any non-volatile computer-readable storage medium. In some embodiments, a computer storage device may also be computer memory, or vice versa.

[0052] As used herein, "computing system" encompasses an electronic component capable of running a program or machine executable instructions or computer executable code. References to computing systems including examples of "computing systems" should be interpreted as including more than one computing system or processing core. A computing system may, for example, be a multi-core processor. A computing system may also refer to a collection of computing systems within a single computer system or distributed between multiple computer systems. The term "computing system" should also be interpreted as referring to a collection or network of multiple computing devices, each of which includes a processor or computing system. Machine executable code or instructions may be run by multiple computing systems or processors that may be within the same computing device or may even be distributed on multiple computing devices.

[0053] Machine executable instructions or computer executable code may include instructions or programs that cause a processor or other computing system to perform an aspect of the present invention. Computer executable code for performing operations for various aspects of the present invention may be written in any combination of one or more programming languages, including object-oriented programming languages ​​(e.g., Java, Smalltalk, C++, etc.) and conventional program programming languages ​​(e.g., "C" programming language or similar programming languages), and compiled into machine executable instructions. In some instances, the computer executable code may be in the form of a high-level language or in a precompiled form, and may be used in conjunction with an interpreter that generates machine executable instructions in operation. In other instances, machine executable instructions or computer executable code may be in the form of programming a programmable logic gate array.

[0054] The computer executable code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on a remote computer or server. In the latter scenario, the remote computer may be connected to the user's computer through any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., through the Internet using an Internet service provider).

[0055] Various aspects of the present invention are described with reference to flowchart illustrations and / or block diagrams of methods, devices (systems) and computer program products according to embodiments of the present invention. It should be understood that each frame or part of the frame of a flowchart, diagram and / or block diagram can be implemented by a computer program instruction in the form of computer executable code when appropriate. It should also be understood that when mutually non-exclusive, the frames in different flowcharts, diagrams and / or block diagrams can be combined. These computer program instructions can be provided to a computing system of a general-purpose computer, a special-purpose computer or other programmable data processing device to produce a machine, so that the instructions run by the computing system of a computer or other programmable data processing device create a unit for implementing the function / action specified in one or more frames of a flowchart and / or block diagram.

[0056] These machine-executable instructions or computer program instructions may also be stored in a computer-readable medium that can direct a computer, other programmable data processing apparatus, or other device to function in a specific manner so that the instructions stored in the computer-readable medium produce an article of manufacture that includes instructions for implementing the functions / actions specified in the flowchart and / or one or more block diagram blocks.

[0057] Machine-executable instructions or computer program instructions may also be loaded onto a computer, other programmable data processing apparatus, or other device to cause a series of operational steps to be performed on the computer, other programmable apparatus, or other device, thereby producing a computer-implemented process, so that the instructions running on the computer or other programmable apparatus provide a process for implementing the functions / actions specified in the flowchart and / or one or more block diagram blocks.

[0058] "User interface" used herein is an interface that allows a user or operator to interact with a computer or computer system. "User interface" may also be referred to as a "human-machine interface device". A user interface may provide information or data to an operator and / or receive information or data from an operator. A user interface may enable the input from an operator to be received by the computer, and may provide output to a user from the computer. In other words, a user interface may allow an operator to control or manipulate a computer, and an interface may allow a computer to indicate the effect of the control or manipulation of an operator. Displaying data or information on a display or graphical user interface is an example of providing information to an operator. Receiving data by a keyboard, mouse, tracking ball, touch pad, pointing stick, graphic input board, joystick, game controller, webcam, head mounted device, foot pedal, wired gloves, remote controller, and accelerometer is all examples of user interface components that enable receiving information or data from an operator.

[0059] As used herein, "hardware interface" encompasses an interface that enables a computing system of a computer system to interact with and / or control an external computing device and / or apparatus. A hardware interface may allow a computing system to send a control signal or instruction to an external computing device and / or apparatus. A hardware interface may also enable a computing system to exchange data with an external computing device and / or apparatus. Examples of hardware interfaces include, but are not limited to, a universal serial bus, an IEEE 1394 port, a parallel port, an IEEE 1284 port, a serial port, an RS-232 port, an IEEE-488 port, a Bluetooth connection, a wireless LAN connection, a TCP / IP connection, an Ethernet connection, a control voltage interface, a MIDI interface, an analog input interface, and a digital input interface.

[0060] As used herein, a "display" or "display device" encompasses an output device or user interface suitable for displaying images or data. A display may output visual, auditory, and / or tactile data. Examples of displays include, but are not limited to: a computer monitor, a television screen, a touch screen, a tactile electronic display, a Braille screen, a cathode ray tube (CRT), a memory tube, a bi-stable display, an electronic paper, a vector display, a flat panel display, a vacuum fluorescent display (VF), a light emitting diode (LED) display, an electroluminescent display (ELD), a plasma display panel (PDP), a liquid crystal display (LCD), an organic light emitting diode display (OLED), a projector, and a head mounted display.

[0061] K-space data is defined herein as the measurements of radio frequency signals emitted by atomic spins recorded using the antenna of a magnetic resonance apparatus during a magnetic resonance imaging scan. Magnetic resonance data is an example of tomographic medical image data.

[0062] A Magnetic Resonance Imaging (MRI) image or MR image is defined herein as a two-dimensional or three-dimensional visualization reconstructed from anatomical data contained within the magnetic resonance imaging data. Such visualization can be performed using a computer. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Preferred embodiments of the present invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0064] Figure 1 An example of a medical system is illustrated;

[0065] Figure 2 Shows instructions for use Figure 1 A flowchart of a method of a medical system;

[0066] Figure 3 Further examples of medical systems are illustrated;

[0067] Figure 4 Shows instructions for use Figure 3 A flowchart of a method of a medical system;

[0068] Figure 5 illustrates an example of an anomaly detection module; and

[0069] Figure 6 An example of a graphical user interface is illustrated.

[0070] Reference numerals list

[0071] 100 Medical System

[0072] 102 Computer

[0073] 104 Computing Systems

[0074] 106 Hardware Interface

[0075] 108 User Interface

[0076] 110 Memory

[0077] 120 machine executable instructions

[0078] 122 Anatomy Detection Module

[0079] 124 A set of magnetic resonance images

[0080] 126 Exception indicator

[0081] 128 A set of reconstruction parameters

[0082] 130 Lookup table or database

[0083] 132 Reconstruction Algorithm

[0084] 134 Synthetic MRI

[0085] 200 receiving a set of magnetic resonance images describing a field of view of an object acquired according to a synthetic magnetic resonance imaging protocol

[0086] 202 receiving an anomaly indicator from the anomaly detection module in response to inputting at least one magnetic resonance image in the set of magnetic resonance images into the anomaly detection module

[0087] 204 Using the anomaly indicator to determine a set of reconstruction parameters

[0088] 206 Reconstructing a synthetic magnetic resonance image according to the set of magnetic resonance images and the set of reconstruction parameters

[0089] 300 Medical System

[0090] 302 Magnetic Resonance Imaging System

[0091] 304 Magnet

[0092] 306 Magnet bore

[0093] 308 Imaging Area

[0094] 309 Field of View

[0095] 310 Magnetic Field Gradient Coil

[0096] 312 Magnetic Field Gradient Coil Power Supply

[0097] 314 RF Coil

[0098] 316 Transceiver

[0099] 318 objects

[0100] 320 Object Support

[0101] 330 Pulse sequence command

[0102] 332 k-space data

[0103] 334 A set of initial reconstruction parameters

[0104] 336 Experimental MRI

[0105] 338 Abnormal Location

[0106] 340 Calculated image contrast range

[0107] 342 Predetermined image comparison range

[0108] 400 Acquiring k-space data by controlling a magnetic resonance imaging system using pulse sequence commands

[0109] 402 Reconstructing a set of magnetic resonance images based on k-space data

[0110] 404 Setting initial reconstruction parameters

[0111] 406 Reconstructing a test magnetic resonance image using the initial reconstruction parameters

[0112] 408 Is the image contrast within the (one or more) abnormal locations within a predetermined image contrast range?

[0113] 410 providing a test magnetic resonance image as a synthetic magnetic resonance image

[0114] 412 Modify initial reconstruction parameters

[0115] 600 Graphical User Interface

[0116] 602 Composite Image DETAILED DESCRIPTION

[0117] In these drawings, elements with the same number are either equivalent elements or perform the same function. If the function is equivalent, an element that has been discussed previously will not necessarily be discussed again in the following drawings.

[0118] Figure 1 An example of a medical system 100 is illustrated. Medical system 100 is shown to include a computer 102. Computer 102 is intended to represent one or more computing devices located in one or more locations. The computer includes a computing system 104, which is also intended to represent one or more computing systems or computing cores located within one or more computer systems 102. Computing system 104 is shown in communication with optional hardware interface 106, user interface 108, and memory 110. Hardware interface 106 can enable computing system 104 to exchange commands and data with other components, if present.

[0119] Thus, the hardware interface 106 may enable the computing system 104 to control other components (e.g., a magnetic resonance imaging system). The user interface 108, if present, may be used to provide a means for interacting with an operator or user and displaying data to it. Likewise, the medical system 100 may receive control commands from the user interface 108. The memory 110 is intended to represent various types of memory accessible to the computing system 104, and may be, for example, a non-transitory storage medium.

[0120] The memory 110 is shown as containing machine executable instructions 120. The machine executable instructions can enable the computing system 104 to perform tasks such as controlling other components, performing digital tasks, and performing basic image processing tasks. The memory 110 is also shown as containing an anatomical detection module 122, which can receive at least one magnetic resonance image in a set of magnetic resonance images 124 and output an abnormality indicator 126 in response. The anatomical detection module 122 can be implemented, for example, using a machine learning module or an algorithm module, or a combination of the two. The memory 110 is also shown as containing an abnormality indicator 126 output by the anatomical detection module 122 in response to receiving at least one magnetic resonance image in a set of magnetic resonance images 124.

[0121] The memory 110 is shown as optionally including a lookup table 130 or a database 130. For example, given a particular abnormality indicator 126, it can be used as a query to the lookup table or database 130; they can then output a set of reconstruction parameters 128. The memory 110 is also shown as including a reconstruction algorithm 132 for reconstructing a synthetic magnetic resonance image 134, which is also stored in the memory. In some instances, the reconstruction algorithm 132 can be iterative and can be used to generate the set of reconstruction parameters 128 while it reconstructs the synthetic magnetic resonance image 134. For example, the abnormality indicator 126 can represent an area within the field of view of the set of magnetic resonance images 124 that should have a predetermined contrast range. The reconstruction algorithm 132 can then iteratively change the reconstruction parameters 128 so that within the area, the synthetic magnetic resonance image 134 has a predetermined image contrast range.

[0122] Figure 2 Shows the instructions for operation Figure 1 Flowchart of a method of a medical system 100 of the present invention. First, in step 200, the set of magnetic resonance images 124 is received. These images 124 describe the field of view of the object acquired according to the synthetic magnetic resonance imaging protocol. This may mean that the images constituting the set of magnetic resonance images are all for the same field of view and have different contrasts. The contrast of the magnetic resonance images means a specific set of acquisition parameters and / or reconstruction parameters.

[0123] Next, in step 202, an abnormality indicator 126 is received from the anatomy detection module 122 in response to inputting at least one magnetic resonance image in the set of magnetic resonance images 124. Next, in step 204, the set of reconstruction parameters 128 is determined using the abnormality indicator 126, and this may be performed, for example, using a lookup table or database 130 and using an iterative reconstruction algorithm 132 as described above.

[0124] Figure 3 A further example of a medical system 300 is illustrated. Figure 3 The medical system 300 depicted in FIG. Figure 1 The medical system 100 is different in that it additionally includes a magnetic resonance imaging system 302 controlled by the computing system 104 .

[0125] The magnetic resonance imaging system 302 comprises a magnet 304. The magnet 304 is a superconducting cylindrical magnet having a bore 306 passing therethrough. It is also possible to use different types of magnets; for example, split cylindrical magnets and so-called open magnets can also be used. A split cylindrical magnet is similar to a standard cylindrical magnet, except that the cryostat has been divided into two sections to allow access to the isoplane of the magnet, such a magnet can be used, for example, in conjunction with charged particle beam therapy. An open magnet has two magnet sections, one above the other, with a space large enough in between to accommodate the subject: the arrangement of the two section areas is similar to a Helmholtz coil. Open magnets are popular because the subject is less restricted. Inside the cryostat of the cylindrical magnet, there is a collection of superconducting coils.

[0126] Within the bore 306 of the cylindrical magnet 304 there is an imaging zone 308 in which the magnetic field is sufficiently strong and uniform to enable magnetic resonance imaging to be performed. A field of view 309 is shown within the imaging zone 308. The acquired k-space data is typically acquired for the field of view 309. The region of interest may be identical to the field of view 309, or the region of interest may be a sub-volume of the field of view 309. An object 318 is shown supported by an object support 320 such that at least a portion of the object 318 is within the imaging zone 308 and the field of view 309.

[0127] There is also a set of magnetic field gradient coils 310 in the bore 306 of the magnet, which is used to collect preliminary k-space data to spatially encode magnetic spins in the imaging zone 308 of the magnet 304. The magnetic field gradient coils 310 are connected to a magnetic field gradient coil power supply 312. The magnetic field gradient coils 310 are intended to be representative. Typically, the magnetic field gradient coils 310 include three separate sets of coils, which are used to perform spatial encoding in three orthogonal spatial directions. The magnetic field gradient power supply supplies current to the magnetic field gradient coils. The current supplied to the magnetic field gradient coils 310 is controlled according to time and can be ramp-varied or pulse-varied.

[0128] Adjacent to the imaging zone 308 is a radio frequency coil 314 for manipulating the orientation of magnetic spins within the imaging zone 308 and for receiving radio transmissions from spins within the imaging zone 308. The radio frequency antenna may include multiple coil elements. The radio frequency antenna may also be referred to as a channel or antenna. The radio frequency coil 314 is connected to a radio frequency transceiver 316. The radio frequency coil 314 and the radio frequency transceiver 316 may be replaced by separate transmit and receive coils and separate transmitters and receivers. It should be understood that the radio frequency coil 314 and the radio frequency transceiver 316 are representative. The radio frequency coil 314 is also intended to represent a dedicated transmit antenna and a dedicated receive antenna. Similarly, the transceiver 316 may also represent a separate transmitter and receiver. The radio frequency coil 314 may also have multiple receive / transmit elements, and the radio frequency transceiver 316 may have multiple receive / transmit channels. The transceiver 316 and the gradient controller 312 are shown as being connected to the hardware interface 106 of the computer system 102.

[0129] The memory 110 is also shown as containing pulse sequence commands 330 that enable the computing system 104 to control the magnetic resonance imaging system 302 to acquire k-space data 332 according to a synthetic magnetic resonance imaging protocol. The pulse sequence commands 330 are commands or data that can be converted into commands that control the timing and function of various components of the magnetic resonance imaging system 302. The memory 110 is also shown as containing k-space data 332 that is acquired by controlling the magnetic resonance imaging system 302 using the pulse sequence commands 330. In this example, the set of reconstruction parameters 128 and the synthetic magnetic resonance image 134 are determined iteratively.

[0130] The memory 110 is also shown as containing a set of preliminary reconstruction parameters 334 used during reconstruction of a test magnetic resonance image 336. An abnormality location 338 is specified within the test magnetic resonance image 336. Also in the memory 110 is a calculated image contrast range 340 within the anatomical abnormality location 338 of the test magnetic resonance image 336. The calculated image contrast range 340 is compared to a predetermined image contrast range 342. If the calculated image contrast range 340 differs from the predetermined image contrast range 342, then the set of preliminary reconstruction parameters 334 can be modified and the test magnetic resonance image 336 can be recalculated.

[0131] Figure 4 Shows the instructions for operation Figure 3 Flowchart of a method of a medical system 300 of the present invention. First, in step 400, k-space data 332 is acquired by controlling the magnetic resonance imaging system 302 using pulse sequence commands 330. Next, in step 402, a set of magnetic resonance images 124 is reconstructed based on the acquired k-space data 332. The method then proceeds to perform the following steps: Figure 2Steps 200 and 202 are shown. Steps 404, 406, 408, 410 and 412 may be replaced by Figure 2 In this example, after step 202 is completed, the method then proceeds to step 404. In this case, a set of preliminary reconstruction parameters 334 is set or determined. Next, in step 406, a trial magnetic resonance image 336 is reconstructed based on the set of magnetic resonance images 124 and the set of preliminary reconstruction parameters 334.

[0132] The method then proceeds to step 408, which is a question box, and the question is "Is the calculated image contrast range 340 the same as the predetermined image contrast range 342?" If the answer is yes, then the method proceeds to step 410 and provides the trial magnetic resonance image 336 as the synthetic magnetic resonance image 134. If the answer is no, then the method proceeds to step 412, where the set of tentative reconstruction parameters 334 is modified, and then the method returns to step 406, where the trial magnetic resonance image 336 is recalculated. The cycle is then repeated until the calculated image contrast range 340 matches the predetermined image contrast range 342.

[0133] Figure 5 An example of an anatomy detection module 122 is illustrated. The anatomy detection module 122 receives as input the set of magnetic resonance images 124 and then outputs an abnormality indicator 126. As described above, the anatomy detection module 122 may be algorithmically implemented as a trained machine learning module or a combination of both.

[0134] Figure 6 An example of a graphical user interface 600 showing a composite image 602 is illustrated. The composite image in this example 602 is formed from a set of synthetic magnetic resonance images. Each synthetic magnetic resonance image has its own abnormality 338. Within the composite image 602, the abnormality 338 from each image is displayed, and the space of the composite image 602 between these images is formed by blending the individual synthetic magnetic resonance images. In this user interface, if the user selects one of the anatomical abnormality locations 338, then the image from which the anatomical abnormality 338 was obtained is displayed. Thus, the graphical user interface 600 can be used to quickly display the various anatomical abnormality locations 338 and then select the more clinically relevant image for the doctor or healthcare provider to review.

[0135] While the invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary and not restrictive; the invention is not limited to the disclosed embodiments.

[0136] Those skilled in the art can understand and implement other variations of the disclosed embodiments when practicing the claimed invention by studying the drawings, the disclosure and the claims. In the claims, the word "comprising" does not exclude other elements or steps, and the words "one" or "an" do not exclude multiple. A single processor or other unit can implement the functions of several items recorded in the claims. Although certain measures are recorded in mutually different dependent claims, this does not indicate that a combination of these measures cannot be used to advantage. The computer program may be stored / distributed on a suitable medium, for example, an optical storage medium or solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, for example, via the Internet or other wired or wireless telecommunications systems. Any reference numerals in the claims should not be interpreted as limiting the scope.

Claims

1. A medical system (100, 300), comprising: a memory (110) storing machine executable instructions (120) and an anomaly detection module (122); as well as A computing system (104), wherein execution of the machine-executable instructions causes the computing system to: receiving (200) a set of magnetic resonance images (124) depicting a field of view (309) of a subject (318) acquired according to a synthetic magnetic resonance imaging protocol; receiving (202) an anomaly indicator (126) from the anomaly detection module in response to inputting at least one magnetic resonance image of the set of magnetic resonance images into the anomaly detection module; using the anomaly indicator to determine (204) a set of reconstruction parameters (128); and A synthetic magnetic resonance image (134) is reconstructed (206) based on the set of magnetic resonance images and the set of reconstruction parameters.

2. The medical system according to claim 1, wherein: The field of view describes a right hemisphere and a left hemisphere of the subject's brain, wherein the anomaly indicator includes one or more anomaly locations in the field of view, and wherein the algorithmic anomaly detection module is further configured to spatially locate the one or more anomalies in the subject's brain using a symmetric contrast between the right hemisphere and the left hemisphere of the subject's brain.

3. The medical system according to claim 1, wherein: The anomaly indicator comprises one or more anomaly locations in the field of view, wherein the anomaly detection module comprises an autoencoder neural network configured to output an autoencoded image for each of the at least one magnetic resonance image in the set of magnetic resonance images, wherein receiving the anomaly indicator in response to inputting the at least one magnetic resonance image in the set of magnetic resonance images into the anomaly detection module comprises: receiving the autoencoded image in response to inputting the at least one magnetic resonance image of the set of magnetic resonance images into the autoencoder neural network; The one or more abnormality locations are determined by comparing the autoencoded image of each of the at least one magnetic resonance image of the set of magnetic resonance images with the at least one magnetic resonance image of the set of magnetic resonance images.

4. The medical system according to claim 2 or 3, wherein: Execution of the machine executable instructions further causes the computing system to iteratively (406, 408, 412) vary the set of reconstruction parameters and reconstruct the synthetic magnetic resonance image to adjust image contrast of the one or more abnormal locations in the image relative to their surroundings to have a predetermined image contrast range (342).

5. The medical system according to any one of claims 2 to 4, wherein: The anomaly indicator indicates a plurality of anomaly locations, wherein the synthetic magnetic resonance image is reconstructed for each of the plurality of anomaly locations, thereby generating a set of synthetic magnetic resonance images.

6. The medical system according to claim 5, wherein: Execution of the machine executable instructions further causes the computing system to construct a composite image from a set of synthetic magnetic resonance image locations by: including the location of the abnormality from each synthetic magnetic resonance image of the set of synthetic magnetic resonance images; and The set of synthetic magnetic resonance images is used to blend pixel values ​​between the abnormality locations from each synthetic magnetic resonance image in the set of synthetic magnetic resonance images.

7. The medical system according to claim 6, wherein: Execution of the machine-executable instructions further causes the computing system to: displaying the composite image on a graphical user interface (600); receiving, from the graphical user interface, a selection of an anomaly location within the composite image; and The synthesized magnetic resonance image including the selected abnormality location is displayed.

8. The medical system according to claim 1, wherein: The anatomical detection module includes a convolutional neural network configured to output an abnormality classification as the abnormality indicator in response to receiving the at least one magnetic resonance image in the set of magnetic resonance images.

9. The medical system according to claim 1, wherein: The anatomical detection module comprises a variational autoencoder neural network having a latent space, wherein the anomaly detection module is configured to output an anomaly classification as the anomaly indicator, wherein receiving the anomaly indicator in response to inputting the at least one magnetic resonance image of the set of magnetic resonance images into the anomaly detection module comprises: inputting the at least one magnetic resonance image in the set of magnetic resonance images into the variational autoencoder; and The anomaly classification is determined using an out-of-distribution detection method on the latent space of the variational autoencoder.

10. The medical system according to claim 8 or 9, wherein: The reconstruction parameters are determined by searching a reconstruction parameter look-up table or a reconstruction parameter database using the anomaly classification.

11. The medical system according to any one of the preceding claims, wherein: Reconstructing the synthetic magnetic resonance image comprises: determining a T1-related value, a T2-related value, and a proton density for each voxel of the field of view by performing a voxel-by-voxel fit of a selected signal intensity equation to the set of magnetic resonance images; and The synthetic magnetic resonance image is reconstructed by calculating a signal intensity value for each voxel using a reconstruction signal intensity equation that takes as input the voxel-by-voxel T1 correlation value, the voxel-by-voxel T2 correlation value, the voxel-by-voxel proton density value, and the reconstruction parameters.

12. The medical system according to any one of claims 1 to 10, wherein: The set of magnetic resonance images forms a magnetic resonance fingerprint, and wherein the synthetic magnetic resonance image is reconstructed from the set of magnetic resonance images according to a magnetic resonance fingerprinting protocol.

13. The medical system according to any one of the preceding claims, wherein: The medical system further comprises a magnetic resonance imaging system (302), wherein the memory further comprises pulse sequence commands (330) configured to control the magnetic resonance imaging system to acquire k-space data (332) according to the synthetic magnetic resonance imaging protocol, wherein execution of the machine executable instructions further causes the computing system to: acquiring (400) the k-space data by controlling the magnetic resonance imaging system using the pulse sequence commands; and The set of magnetic resonance images is reconstructed (402) based on the k-space data.

14. A method of operating a medical system (100, 200), wherein: The method comprises: receiving (200) a set of magnetic resonance images (124) depicting a field of view (309) of a subject (318) acquired according to a synthetic magnetic resonance imaging protocol; receiving (202) an anomaly indicator (126) from the anomaly detection module in response to inputting at least one magnetic resonance image of the set of magnetic resonance images into the anomaly detection module (122); using the anomaly indicator to determine (204) a set of reconstruction parameters (128); and A synthetic magnetic resonance image (134) is reconstructed (206) based on the set of magnetic resonance images and the set of reconstruction parameters.

15. A computer program comprising machine executable instructions (120) for execution by a computing system (104), wherein: The computer program also includes an anatomical detection module (122) for execution by the computing system, wherein execution of the machine-executable instructions causes the computing system to: receiving (200) a set of magnetic resonance images (124) depicting a field of view (309) of a subject (318) acquired according to a synthetic magnetic resonance imaging protocol; receiving (202) an anomaly indicator (126) from the anomaly detection module in response to inputting at least one magnetic resonance image of the set of magnetic resonance images into the anomaly detection module; using the anomaly indicator to determine (204) a set of reconstruction parameters (128); and A synthetic magnetic resonance image (134) is reconstructed (206) based on the set of magnetic resonance images and the set of reconstruction parameters.

Images