WO2009050676A1

WO2009050676A1 - Pathology-related magnetic resonance imaging

Info

Publication number: WO2009050676A1
Application number: PCT/IB2008/054271
Authority: WO
Inventors: Peter Boernert; Stewart Young; Daniel Bystrov; Thomas Netsch; Vladimir Pekar; Arianne Van Muiswinkel; Christian Adrian Cocosco; Sebastian Peter Michael Dries; Jürgen Heinrich GIESEKE; Ronaldus Frederik Johannes Holthuizen; Arjan Willem Simonetti
Original assignee: Koninklijke Philips Electronics N.V.; Philips Intellectual Property & Standards Gmbh
Priority date: 2007-10-17
Filing date: 2008-10-17
Publication date: 2009-04-23

Abstract

A magnetic resonance imaging method includes acquiring a first image (16, 20, 20') of a subject, electronically identifying information related to a suspected pathology based on the first image, electronically planning a magnetic resonance imaging session (40) relating to the suspected pathology based on the identified information relating to the suspected pathology, and executing a selected one of the magnetic resonance imaging sessions to acquire one or more magnetic resonance images of the subject probative of the suspected pathology. A suitable magnetic resonance imaging system may include a magnetic resonance scanner (10), a models database (34, 36) including pathology models (36) including information pertaining to pathologies derived from previously acquired magnetic resonance images of subjects having said pathologies, and a session planner (30) configured to construct a pathology- specific and subject- specific magnetic resonance imaging session plan (40) based on a magnetic resonance scout image (16) and the pathology models (36).

Description

Pathology-Related Magnetic Resonance Imaging

FIELD OF THE INVENTION

The following relates to the diagnostic imaging arts, and more particularly to magnetic resonance imaging for the purpose of identifying, screening, diagnosing, monitoring, assessing, or otherwise analyzing a pathology, and is described with particular reference thereto.

BACKGROUND

Magnetic resonance imaging is a well-known technique for diagnosis and monitoring of various pathologies that provide manifestations discernable in magnetic resonance images. For example, tumors or other growths, metastatic cancers, cardiac pathologies, and so forth are amenable to diagnosis and monitoring by magnetic resonance imaging. In diagnostic applications, magnetic resonance imaging is used to identify a pathology and characteristics thereof, such as tumor size and position, extent of metastasis, and so forth. In diagnostic applications, the characteristics of the pathology are generally not known a priori. For example, a particular type of cancer may be suspected based on blood test results or by virtue of being the target of a magnetic resonance imaging screening, but the existence, location, and extent of any tumor or metastatic incursions of cancerous tissue are not known prior to the magnetic resonance imaging. Indeed, in some diagnostic applications such as screening, the existence of the suspected pathology may be uncertain prior to the magnetic resonance imaging. In monitoring applications, the magnetic resonance imaging is used to determine the progress or regression of the pathology.

Conventionally, a subject to be diagnosed for a suspected pathology is loaded into a magnetic resonance scanner. A scout image of the relevant general anatomical area is acquired. Interactive geometric planning is sometimes performed using the scout image based on anatomical landmarks identified automatically or by the radiologist or other medical professional operating a graphical user interface. Additionally, the radiologist or other medical professional examines the scout image, looking for features likely to be associated with the suspected pathology. If such features are identified by the radiologist, a diagnostic magnetic resonance imaging session relating to the suspected pathology is designed by the radiologist based on the information gleaned by the radiologist from the scout image and based on the radiologist's expertise, possibly guided by standard protocols practiced by the magnetic resonance imaging facility. The diagnostic magnetic resonance imaging session is executed, and the resulting magnetic resonance image or images are used for diagnosis or assessment.

The conventional approach has certain disadvantages. The geometric planning employs landmarks designed for or extracted from normal patients. These anatomical landmarks may be distorted or nonexistent in the case of certain pathologies. Similarly, other anatomically significant features may be introduced by the pathology, which may be fundamentally different from normal features - for example, a cancerous region may not have the well-defined border of a healthy organ, but rather may manifest in the image as shadowing or other gradually changing grayscale intensity.

Additionally, the identification of pathology-related information and the planning of the diagnostic magnetic resonance imaging session are performed manually, based on the radiologist's expertise and possibly guided by standard protocols. This reliance on the radiologist introduces uncertainty and variation into the process, and the results can vary widely depending upon the expertise of the radiologist respective to diagnostic magnetic resonance imaging directed toward the particular suspected pathology. Moreover, manual study and planning by the radiologist is time-consuming and generally results in longer magnetic resonance imaging sessions.

The following provides new and improved apparatuses and methods which overcome the above -referenced problems and others.

SUMMARY OF THE INVENTION

In accordance with one aspect, a magnetic resonance imaging method is disclosed, comprising: acquiring a first image of a subject; electronically identifying information related to a suspected pathology based on the first image; electronically planning a magnetic resonance imaging session relating to the suspected pathology based on the identified information relating to the suspected pathology; and executing a selected one of the magnetic resonance imaging sessions to acquire one or more magnetic resonance images of the subject probative of the suspected pathology.

In accordance with another aspect, a magnetic resonance imaging system is disclosed, comprising: a magnetic resonance scanner; a models database including pathology models for different pathologies, each pathology model including information pertaining to a corresponding pathology derived from previously acquired magnetic resonance images of one or more subjects having the corresponding pathology; and a session planner configured to construct a pathology specific and subject specific magnetic resonance imaging session plan based on a magnetic resonance scout image of a subject acquired using the magnetic resonance scanner and a pathology model corresponding to a suspected pathology retrieved from the models database.

In accordance with another aspect, a magnetic resonance imaging system comprising: a magnetic resonance scanner configured to acquire magnetic resonance medical images; at least one other-modality imager configured to acquire non-magnetic resonance medical images using an imaging modality other than magnetic resonance; a picture archiving and communications system (PACS) communicating with the magnetic resonance scanner and the at least one other-modality imager to archive the acquired magnetic resonance and non-magnetic resonance medical images; and a session planner configured to construct a pathology specific and subject specific magnetic resonance imaging session plan for execution by the magnetic resonance scanner based on a magnetic resonance scout image of a subject acquired by the magnetic resonance scanner and at least one non magnetic resonance medical image of the subject acquired by the at least one other-modality imager and archived by the PACS. One advantage resides in more efficient medical magnetic resonance imaging workflow.

Another advantage resides in reduced likelihood of human error in medical magnetic resonance imaging.

Another advantage resides in leveraging pathology-related information acquired from previous subjects having a given pathology to improve magnetic resonance imaging of features of the pathology in a current subject.

Still further advantages of the method and device disclosed herein will be appreciated by those of ordinary skill in the art upon reading and understanding the following detailed description.

The drawings are only for purposes of illustrating embodiments of the device and method disclosed herein, and are not to be construed as limiting the invention. FIGURE 1 diagrammatically shows a medical imaging system. FIGURE 2 diagrammatically shows processing making use of clinical query information obtained via a user interface in scouting and planning a magnetic resonance medical imaging system performed using the system of FIGURE 1.

FIGURE 3 diagrammatically shows planning of a magnetic resonance medical imaging system by the system of FIGURE 1.

DETAILED DESCRIPTION

With reference to FIGURE 1, a medical imaging system includes a magnetic resonance scanner 10 and optionally includes one or more other- modality imagers such as a positron emission tomography (PET) scanner 12 diagrammatically illustrated in FIGURE 1 as an example. The optional one or more other-modality imagers may include one or more PET scanners, one or more computed tomography (CT) scanners, one or more gamma cameras, one or more x-ray fluoroscopes, some combination of these aforementioned instrumentalities, or so forth. Although not illustrated, the medical imaging system may also include more than the one illustrated magnetic resonance scanner 10, for example may include two or more magnetic resonance scanners.

The magnetic resonance scanner 10 is used to perform medical magnetic resonance imaging of a subject suspected to have a pathology that is to be the target of the medical magnetic resonance imaging. The term "suspected" is intended herein to encompass certain knowledge as well; for example, the subject may be known to have the suspected pathology that is the target of the medical magnetic resonance imaging. It is also to be noted that though the methods and apparatuses described herein are with reference to suspected pathology, it is possible to apply the described methods to normal tissue as well.

A magnetic resonance (MR) scout image planner 14 selects an initial geometry for acquiring at least one magnetic resonance scout image 16 that is used to assess the precise geometry of the subject and to detect and initially assess any features visible in the magnetic resonance image that may be indicative of the suspected pathology. Alternatively or additionally, the scout image 16 may be used to assess features indicative of normal tissues or areas of interest as well. For example, in imaging a breast suspected of containing a lesion or tumor, the scout image may first be used to assess normal breast tissue and lymph nodes, while excluding air, heart, lungs silicone implants, and the suspected pathology as well. Optionally, the magnetic resonance scout image planner 14 may access information useful for planning the scout image acquisition, such as one or more "other-modality" images 20 acquired by an "other-modality" imager such as the illustrated PET scanner 12 and stored in a picture archiving and communications system (PACS) 22, or clinical query information 24 received from a radiologist or other medical professional via a user interface 26. For example, the magnetic resonance scout image planner 14 may retrieve the other-modality image 20 from the PACS 22 and display it to the radiologist on the user interface 26, whereupon the radiologist recognizes a tumorous feature associated with a cancerous pathology of interest in the displayed other-modality image and interacts with the magnetic resonance scout image planner 14 via the user interface 26 to select an appropriate slice for the magnetic resonance scout image acquisition and potentially also for the high resolution scans to be performed for final diagnosis, that is likely to intersect the tumorous feature. The identification of the tumor or other pathology feature can also be automatic or semi-automatic using a suitable algorithm for identifying the pathology-related feature. Alternatively, based on the clinical query information 24 indicating the suspected cancerous pathology mentioned here as an example, or based on a referring physician having marked a suspected lesion in the "other-modality" image or images 20, the magnetic resonance scout image planner 14 may automatically select the appropriate MR scanning location and plane orientation.

A magnetic resonance session planner 30 receives the magnetic resonance scout image 16 and the optional other-modality image or images 20. The other-modality image or images 20, if available, are preferably spatially registered with the magnetic resonance scout image 16 by a spatial registration processor 32 to facilitate comparison of corresponding features in the magnetic resonance and other-modality images 16, 20. Some suitable spatial registration techniques are described, for example, in the International patent application WO 2005/057495 by Pekar et al, published June 23, 2005, and in the International patent application WO 2007/054864 by Carlsen et al., published May 18, 2007. The spatial registration technique or techniques employed by the spatial registration processor 32 can be either a rigid or a non-rigid technique.

The session planner 30 compares the scout image 16 and, if available, other- modality image or images 20, with a stored normal model 34 of a "normal" subject that does not have the pathology of interest and is representative of a typical healthy subject. The normal model 34 may be a singular model of a representative subject, or may be a model selected from a plurality of normal models of subjects of different genders, sizes, body shapes, or so forth. The normal subject model 34 is suitably generated from magnetic resonance images of representative normal subjects previously acquired using the magnetic resonance scanner 10 or another scanner. Some suitable landmark-based techniques for deriving the normal subject model 34 from such previous imaging and for correlating scout images with such landmark-based models are described, for example, in the International patent applications WO 2007/096804 by Bystrov et al, published August 30, 2007, WO 2006/013499 by Springorum et al., published February 9, 2006, WO 2007/057816 by Netsch et al., published May 24, 2007, and WO 2005/102160 by Bystrov et al., published November 3, 2005. In some such techniques, the normal subject model 34 defines certain anatomical landmarks used to geometrically orient the scout image 16, from which the geometrical orientation of the subject may be inferred. In addition to landmarks, other image aspects can be used in the analysis, such as grayscale characteristics. For example, if a (normalized) grayscale intensity is typically around 50 + 5 for one type of tissue in a normal subject, and is typically around 40 + 5 for another type of tissue, then these grayscale values can be stored in the normal subject model 34 and used to distinguish the two different tissue types from one another. The magnetic resonance imaging session geometry can be selected with reference to this subject geometry to enable acquisition of image slices or volumes of selected location and orientation, e.g., sagittal slices through the lungs, axial slices through a selected region of the brain, or so forth.

In similar fashion, a suitable pathology model is selected from a database of stored pathology models 36. The stored pathology model 36 suitably stores landmarks associated with the suspected pathology, and/or stores a distortion or deviation due to the suspected pathology of the normal landmarks of the normal subject model 34. Such pathology-specific landmarks or landmark adjustments enable more accurate geometrical planning than is typically achieved using the normal subject model alone. Additionally, the stored pathology model 36 optionally stores non-landmark based pathology information, such as grayscale intensity characteristics of tumors or other features associated or sometimes associated with the suspected pathology. Intensity-related features or characteristics are sometimes well-representative of the pathology-related features, and can be used for example to identify the nature and extent of a cancerous tumor, a region encroached upon by metastatic cancerous fibrous tissue, or other pathology related features. For example, if a (normalized) grayscale intensity in a particular region is typically around 50 + 5 for a normal subject, then any areas of that region having a normalized grayscale intensity substantially deviating from this normal value (for example, deviating by more than two standard deviations) is suitably identified as a potential pathology-related feature. The pathology model 36 is suitably derived from previous magnetic resonance imaging of representative subjects determined to have the pathology of interest. Generation and use of landmarks in the pathology model database 36 is substantially similar to generation and use of landmarks in the normal subject database 34, except that magnetic resonance data from normal subjects are used in generating the landmarks of the normal subject database 34 whereas magnetic resonance data from one or more subjects (or the present subject) having the pathology of interest are used in generating the landmarks or landmark deviation or distortion information stored in the pathology database 34.

Based on the modeling, the magnetic resonance session planner 30 creates a magnetic resonance session plan 40 to be executed by the magnetic resonance scanner 10. In addition to the modeling, the session planner 30 optionally takes into account the clinical query information 24 obtained from the radiologist or other medical personnel via the user interface 26 or PACS 22. In some embodiments, the session planner 30 causes the user interface 24 to display indicia of a plurality of magnetic resonance imaging session options chosen from a parameterized pathology-specific session plans database 42 based on the clinical query information 24. The chosen magnetic resonance imaging session options are chosen to be useful in imaging the suspected pathology. For example, the chosen magnetic resonance imaging session options may include a suitable pulse sequence or a plurality of suitable pulse sequences, a slice orientation or plurality of slice orientations, or so forth. The radiologist selects from the displayed indicia via the user interface 26 one or more of the session plans or plan parameters from among the presented magnetic resonance imaging session plans or plan parameters relating to the suspected pathology.

The clinical query information 24 can be provided by the radiologist or other medical personnel via the user interface 26 in various ways, such as based on the imaging request data or picked from a list generated on a display of the scout image 16, optionally after some pre-processing. The clinical query information 24 may include, for example, the organ to be examined and the suspected pathology. This information can be provided, for example, by having the user interface 26 display a representation of the patient and having the radiologist use a mouse pointer, touch screen, or the like to choose the organ, or can be extracted from an electronically filed report of a referring physician that is stored in the patients hospital database. Selection of the organ can cause display of a list of pathologies related to that organ from which the radiologist can select the suspected pathology. In some embodiments, more than one suspected pathology can be selected. With the pathology selected, the acquisition parameter presets are suitably retrieved from the database 42. In some embodiments, the presets are taken from magnetic resonance imaging of the suspected pathology previously performed on other subjects. In some embodiments, an automated analysis of the scout image 16 is performed based on comparison with the pathology model

36, and the most likely pathology is highlighted or listed first in the list of pathology choices.

With continuing reference to FIGURE 1 and further reference to FIGURE 2, an illustrative example of use of the clinical query information 24 in the session planning is described, which utilizes the database 42 of acquisition parameter presets relational to organs and pathologies. The database 42 optionally includes logical links between diagnostically useful combinations. The selection of parameters for the magnetic resonance session plan 40 is achieved by interaction with the user interface 26 by which the radiologist or other medical professional provides the magnetic resonance imaging system with the clinical query information 24 about the patient, such as the organ to scan and clinical questions pertaining to the pathology of interest. In the illustrative example of FIGURE 2, the organ is specified as the lumbar spine, by selection of a lumbar spinal region representation 50 on a diagrammatic patient representation 52. Instead of the illustrated diagrammatic patient representation 52, video camera images of the patient, an initial scout scan, a non-graphical organ selection menu, or so forth could also be used. The pathology of interest is selected for example, as discitis, the selection for example being made from a list (not shown) of possible lumbar spinal pathologies retrieved from the database 42.

After input of the clinical query information 24, the scout image 16 is acquired automatically or responsive to a request made by the radiologist via the user interface 26. The acquired scout image 16 is input to the session planner 30 for automated geometry scan planning performed using the normal subject and pathology models 34, 36, and optional automated pathology detection performed using the pathology model 34 optionally including selection of suggested magnetic resonance imaging session parameters from the database 42. At the user interface 26, suggestions of scan geometry and scan parameters are presented for the magnetic resonance session plan 40. The scan geometry and the choice of magnetic resonance sequence or sequences to use, for example defined by presets for the pathology stored in the database 42, can be adjusted manually by the radiologist via the user interface 26, or can be adjusted automatically based on results of an automated pathology analysis.

The final magnetic resonance session plan 40 is executed by the magnetic resonance scanner 10 to generate clinical magnetic resonance images 54 for study by a physician or other medical personnel. Preferably, a report 56 is also generated containing information including the clinical query information 24, the scan settings, optional clinical information automatically derived from the images 54, or so forth. To facilitate development of the pathology model 36 and the pathology- specific session plans database 42, aspects of the final magnetic resonance session plan 40 are optionally fed back into the database 42 to facilitate learning of site-specific preferences, and/or the scout image 16 annotated by information input by or approved by the radiologist is used to update the pathology model 36.

While only a single scout image 16 is illustrated, it is to be understood that multiple scout images may be acquired to facilitate construction of the session plan 42. Typically, the first scout image is of low resolution while subsequent scout images (if acquired) are of higher resolution to confirm the automated pathology detection results with image processing of the subsequent higher resolution images. Alternatively or additionally, more than one low resolution scout image may be acquired with different parameters to improve the reliability and validity of automated pathology detection.

With continuing reference to FIGURES 1 and 2, for illustration purposes the example provided in the flow chart of FIGURE 2 is further described. The patient is referred to the MR facility with the clinical question of a discitis of the lumbar spine. With the patient inside the magnetic resonance scanner 16, at first the radiologist at the user interface 26 clicks on the lumbar spine region 50 of the diagrammatic patient representation 52 using a mouse or other point- and-select input device. The scout image 16 is acquired and input into the magnetic resonance session planner 30 for analysis, and the radiologist is presented with a list of clinical questions selected based on the automated analysis of the scout image 16 leading to selection of lumbar spine examinations previously known and stored in the database 42, for which specific acquisition parameter, presets exist in the database 42. In the illustrated example of pathologies pertaining to the lumbar spinal region, such examinations may comprise "fracture", "spinal stenosis", "disc degeneration", "scoliosis" and "discitis". If one or more of these is detected by analysis of the scout image 16 by the session planner 30, the detected pathology or pathologies are highlighted and presented to the user as first choices. In the case of an exclusion diagnosis with regard to the clinical question this may aid in focusing the differential diagnosis. It is left to the radiologist to make the final selection and configuration of the magnetic resonance session plan 40 to execute, but the planner 30 provides recommendations in the form of default values, textual recommendations, highlighting of detected pathology-related features in a display of the scout image 16, or so forth. Subsequent scout images, if acquired, are suitably input into the session planner 30 to further tailor the acquisition parameters. It is also contemplated for the clinical magnetic resonance images 54 to be fed back into the session planner 30, as indicated by dashed arrows in FIGURE 2, to further refine subsequent portions of the session plan 40. When the last acquisition is finished and the operator marks or selects the examination as complete, the clinical magnetic resonance images 54 are stored and may also be input back to the database 42 or used to further refine the stored pathology model 36. As an example, for updating the pathology model 36 a deformable mesh fitting procedure may be used to determine distortion or deformation of landmarks in the current patient having the pathology being studied, and the distortion or deformation of these landmarks in the current patient may be used (weighted to reflect a per-subject contribution of the current patient to the model 36) to update the landmark distortion or deformation information contained in the stored pathology model 36. The updating of the pathology model 36 and the database 42 can be automatic or semi-automatic - for example, the radiologist may be asked whether the update should be performed, and the radiologist may elect not to update the stored information 36, 42 if the radiologist has reason to believe the current subject is abnormal or otherwise not well-representative of persons having the pathology under study.

With continuing reference to FIGURE 1 and further reference to FIGURE 3, use by the magnetic resonance session planner 30 of the normal subject and pathology models 34, 36 and the other-modality image or images 20 (if available) is further described. The planning includes two components: geometric planning 60 and pathology features delineation 62. However, the two components 60, 62 are interrelated - for example, the geometry is generally selected to effectively image the pathology features. The geometric planning 60 employs the scout image 16 compared with normal landmarks 64 obtained from the normal subject model 34 and pathology-related landmarks or landmark deviations or distortions 66 obtained from the pathology model 36. Additionally, the geometric planning 60 may make use of the clinical query information 24 - for example, examination of a pathology of the lumbar spinal region may preferably entail a geometry acquiring sagittal slices.

The pathology features delineation 62 also may utilize the scout image 16 and the normal and pathology-related landmarks information 64, 66. Additionally, the pathology features delineation 62 may utilize information from the processed other-modality image 20' (where the prime on the reference numeral here indicates the other-modality image after storage in and retrieval from the PACS 22 and spatial registration by the registration processor 32). Still further, the pathology features delineation 62 may utilize pathology intensity features 70 stored in the pathology model 36. The pathology intensity features 70 advantageously can provide information in situations where the pathology does not distort or shift landmarks, but rather changes the grayscale intensity in a region under examination. For example, metastatic cancer may cause the grayscale intensity level (possibly represented by coloration or so forth) to change without shifting or distorting lines, edges, or corners used in defining landmarks.

If available, spatially registered other-modality image 20' can be used to compensate for information that is ambiguous or missing in the magnetic resonance scout image 16. For example, if the radiologist or other medical professional has identified and marked a feature in a PET image that is suspected of being a lesion, then by spatially registering the PET image with a magnetic resonance scout image, the location of, or at least the general area of, the suspected lesion in the magnetic resonance scout image can be determined. This additional information can aid in automated or semi-automated analysis of the magnetic resonance scout image, for example by biasing toward a conclusion that a magnetic resonance image intensity anomaly or other feature at the location corresponding to the suspected lesion in the PET image is indicative of a lesion of interest.

Based on the geometric planning 60 and the pathology-related (or suspected pathology-related) features identified by the pathology features delineation 62, the session planner 30 performs session plan construction 72 to generate the magnetic resonance session plan 40. The session plan construction 72 suitably starts with a magnetic resonance imaging pulse sequence or set of pulse sequences identified in the database 42 as being particularly suitable for addressing the clinical question presented in the clinical query information 24. The plan construction 72 further includes adjusting parameters of the selected imaging sequence or sequences to comport with the slice orientation and field-of-view (FOV) determined in the geometrical planning 60. Still further, the plan construction 72 may entail adjusting other pulse sequence parameters to optimize contrast, resolution, or other imaging characteristics respective to the pathology features identified for study by the pathology features delineation 62, adjustment of the FOV and/or slice orientation such that the entire extent of the detected pathology (e.g. tumor) is encompassed within the field-of-view of the planned geometry, or so forth.

The adjustment of the initially proposed plan (selected, for example, from the database 42 based on the pathology analysis 62) is in some embodiments computed such that the deviation between the initially proposed plan and the adjusted plan is minimal, while still providing adequate slice orientation, FOV, and other image characteristics. Other details of the scan plan, such as the pulse sequence selection, may also usefully be updated - for example, a FLAIR scan might be included if the presence of multiple sclerosis lesions is indicated. In addition to detection and delineation of pathology, the methods described herein may also be used to provide the basis for image quality improvements in general. For example, it is known that for various anatomies, image -based shimming techniques provide improvements in image quality for MRI scanners with higher-order shim capabilities. In these shimming techniques, a good 3D segmentation (or delineation) of the anatomy of interest is needed to perform a robust phase -unwrapping and to calculate the shim settings accurately. However, unlike in current 3D segmentation routines, the methods described herein do not require the user to perform segmentation of the area of interest on a slice-by- slice basis or to provide a start-point (on a per-slice or per-volume basis) for the phase- unwrapping. Thus, by using the methods discussed herein, the user no longer needs to enter this information manually, thereby giving an implementation that is robust, and without additional user-interaction. Additional information on image-based shimming techniques may be obtained from Schar M, et al., "Cardiac SSFP Imaging at 3 Tesla", Magnetic Resonance in Medicine 2004, volume 51, pages 799-806. For example, in imaging a breast suspected of containing a lesion or tumor, it would be possible to shim more accurately on an area of interest by segmenting the breasts and lymph nodes using the techniques described herein, while excluding air, heart, lungs and silicone implants, as the latter significantly adversely affect the phase unwrapping algorithm and/or the calculation of shim settings. The calculated shim settings can then be applied automatically to all future scans that are performed on the same subject and area of interest.

It should be noted that the above-mentioned embodiments illustrate rather than limit the invention, and that those skilled in the art will be able to design many alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps other than those listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The disclosed method can be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer. In the system claims enumerating several means, several of these means can be embodied by one and the same item of computer readable software or hardware. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

Claims

CLAIMS:

1. A magnetic resonance imaging method comprising: acquiring a first image (16, 20, 20') of a subject; electronically identifying information related to a suspected pathology based on the first image; electronically planning a magnetic resonance imaging session (40) relating to the suspected pathology based on the identified information relating to the suspected pathology; and executing a selected one of the magnetic resonance imaging sessions to acquire one or more magnetic resonance images of the subject probative of the suspected pathology.

2. The magnetic resonance imaging method as set forth in claim 1, wherein: the acquiring includes acquiring a scout magnetic resonance image (16) of the subject; and the identifying includes electronically identifying an abnormal image feature indicative of the suspected pathology in the scout magnetic resonance image.

3. The magnetic resonance imaging method as set forth in claim 2, wherein the identifying includes: electronically identifying an abnormal intensity level indicative of the suspected pathology.

4. The magnetic resonance imaging method as set forth in claim 2, wherein the planning includes: electronically identifying one or more landmarks in the scout magnetic resonance image (16) that are associated with or distorted by the suspected pathology; and electronically performing or adjusting geometric planning (60) based on the identified landmarks.

5. The magnetic resonance imaging method as set forth in claim 4, wherein the performing or adjusting of geometric planning (60) includes: electronically comparing the identified one or more landmarks with a pathology model (36) derived from one or more subjects having the suspected pathology.

6. The magnetic resonance imaging method as set forth in claim 1, wherein the acquiring includes acquiring the first image (20, 20') using a modality (12) other than magnetic resonance, the method further comprising: acquiring a scout magnetic resonance image (16) of the subject; and spatially registering the first image (20) and the scout magnetic resonance image to generate a spatially registered first image (20').

7. The magnetic resonance imaging method as set forth in claim 6, wherein the acquiring includes retrieving the first image (20) from a picture archiving and communications system (PACS) (22) or retrieving an electronic report articulating information drawn from the first image.

8. The magnetic resonance imaging method as set forth in claim 6, wherein the planning comprises: electronically delineating (62) an abnormal image feature indicative of the suspected pathology based at least in part on comparison of the scout magnetic resonance image (16) and the spatially registered first image (20').

9. The magnetic resonance imaging method as set forth in claim 1, wherein the acquiring includes acquiring a first magnetic resonance image (16) of the subject, and the method further comprises: electronically selecting an anatomical region for the acquiring of the first magnetic resonance image based on received clinical query information (24).

10. The magnetic resonance imaging method as set forth in claim 9, wherein the electronically selecting comprises: displaying indicia (52) of a plurality of magnetic resonance imaging session options based on at least the received clinical query information (24); and receiving a selection (50) of one or more magnetic resonance imaging session options relating to the suspected pathology from the displayed indicia.

11. The magnetic resonance imaging method as set forth in claim 10, further comprising: electronically analyzing the first magnetic resonance image (16, 20, 20') based on the received clinical query information (24); and selecting the plurality of magnetic resonance imaging session options for display based on the analyzing.

12. The magnetic resonance method as set forth in claim 1, wherein the planning comprises electronically delineating (62) one or more image features of interest, and the executing includes a shimming routine designed to improve magnetic field homogeneity based on the delineated one or more image features of interest.

13. A magnetic resonance imaging system comprising: a magnetic resonance scanner (10); a models database (34, 36) including pathology models (36) for different pathologies, each pathology model including information pertaining to a corresponding pathology derived from previously acquired magnetic resonance images of one or more subjects having the corresponding pathology; and a session planner (30) configured to construct a pathology- specific and subject- specific magnetic resonance imaging session plan (40) based on a magnetic resonance scout image (16) of a subject acquired using the magnetic resonance scanner and a pathology model corresponding to a suspected pathology retrieved from the models database.

14. The magnetic resonance imaging system as set forth in claim 13, further comprising: a spatial registration processor (32); and a picture archiving and communications system (PACS) (22), the session planner (30) further configured to construct the pathology-specific and subject- specific magnetic resonance imaging session plan (40) based on at least one non-magnetic resonance image (20) of the subject retrieved from the PACS and spatially registered by the spatial registration processor with a magnetic resonance scout image (16).

15. The magnetic resonance imaging system as set forth in claim 14, further comprising: a second modality imager (12) configured to acquire the at least one non-magnetic resonance image (20) that is stored in the PACS (22).