(19)
(11)EP 2 289 048 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
27.02.2019 Bulletin 2019/09

(21)Application number: 09746193.3

(22)Date of filing:  04.05.2009
(51)International Patent Classification (IPC): 
G06T 11/00(2006.01)
(86)International application number:
PCT/IB2009/051825
(87)International publication number:
WO 2009/138898 (19.11.2009 Gazette  2009/47)

(54)

USING NON-ATTENUATION CORRECTED PET EMISSION IMAGES TO COMPENSATE FOR INCOMPLETE ANATOMIC IMAGES

VERWENDUNG VON NICHT DÄMPFUNGSKORRIGIERTEN PET-EMISSIONSABBILDUNGEN ZUR KOMPENSATION UNVOLLSTÄNDIGER ANATOMISCHER ABBILDUNGEN

UTILISATION D'IMAGES TOMOGRAPHIQUES PAR ÉMISSION DE POSITONS NON CORRIGÉES EN ATTÉNUATION POUR COMPENSER DES IMAGES ANATOMIQUES TRONQUÉES


(84)Designated Contracting States:
AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO SE SI SK TR

(30)Priority: 15.05.2008 US 53401 P

(43)Date of publication of application:
02.03.2011 Bulletin 2011/09

(73)Proprietor: Koninklijke Philips N.V.
5656 AE Eindhoven (NL)

(72)Inventors:
  • HU, Zhiqiang
    Cleveland, Ohio 44143 (US)
  • GAGNON, Daniel
    Cleveland, Ohio 44143 (US)
  • TUNG, Chia-Hua
    Cleveland, Ohio 44143 (US)

(74)Representative: van Velzen, Maaike Mathilde 
Philips Intellectual Property & Standards High Tech Campus 5
5656 AE Eindhoven
5656 AE Eindhoven (NL)


(56)References cited: : 
US-A1- 2004 260 176
US-B1- 6 490 476
US-A1- 2005 129 295
US-B1- 6 539 103
  
      
    Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


    Description


    [0001] The present application finds particular application in patient imaging systems, particularly involving patient imaging devices such as positron emission tomography (PET) scanners and the like. However, it will be appreciated that the described technique may also find application in spectroscopy systems, other nuclear imaging scenarios, other imaging techniques, and the like.

    [0002] There is a trend in PET imaging to combine a computed tomography (CT) or a magnetic resonance (MR) scanner with a PET scanner. However, such systems are subject to image-truncation problems. For instance, MR and CT images have good quality only within a limited and pre-defined transverse field-of-view (FOV). When scanning a patient or volume of interest (VOI) larger than the CT or MR FOV, the CT or MR image is truncated. The portion of the CT or MR image beyond the CT or MR FOV is badly artifacted, and the portion inside the FOV is moderately artifacted. When one uses the truncated CT or MR image to derive the attenuation map for PET reconstruction, the PET image will be inaccurate and/or artifacts may occur.

    [0003] Moreover, MR is only sensitive to protons in material, and thus there is no direct relationship between the MR image values and linear attenuation co-efficiencies. When trying to derive an attenuation map for PET reconstruction using an MR image, a derived body contour may be smaller than the actual one.

    [0004] Conventional multimodal MR/PET and CT/PET systems acquire the MR or CT data, then acquire PET data, and use the MR or CT data to generate an attenuation map used to compensate for attenuation in the PET data set. However, the CT or MR data is truncated when the subject is larger than a certain size, and does not fit into the FOV of the imaging device. If the CT or MR FOV is increased to accommodate the larger subject, CT or MR reconstruction lags and artifacts occur. Moreover, increasing CT or MR FOV is expensive. Additionally, truncated CT or MR data results in an incomplete attenuation map, which in turn degrades the quality of the PET image for which the incomplete attenuation map is employed during reconstruction of acquired PET data.

    [0005] US 2004/260176 discloses a method of correcting a positron emission tomography (PET) image by developing a compensation of truncated data on the attenuation map.

    [0006] The present application provides new and improved systems and methods for attenuation correction in multi-modal imaging systems, whereby missing image information is compensated for in an anatomical image, and the corrected anatomical image is employed to generate an artifact-free attenuation map for correcting the PET image, which overcome the above-referenced problems and others.

    [0007] In accordance with one aspect, an anatomical imaging system includes a memory that stores truncated anatomical image acquired during a scan of a subject and positron emission tomography (PET) data acquired during a PET scan of the subject. The system further includes a processor that reconstructs a non-attenuation-corrected (NAC) PET image from the acquired PET data, generates an attenuation map from the acquired truncated anatomical image, and compensates for truncated data in the attenuation map using a contour in the NAC PET image.

    [0008] In accordance with another aspect, a method of correcting an attenuation map using a non-attenuation-corrected (NAC) positron emission tomography (PET) image includes generating the NAC PET image, segmenting the NAC PET image to identify a body contour, and identifying a truncated region in the attenuation map. The method further includes correcting truncated image in the truncated region of the attenuation map using the identified contour to infer an appropriate tissue type to use for truncated image, and reconstructing an attenuation-corrected PET image using the corrected attenuation map.

    [0009] The method of correcting a computed tomography (CT) image includes generating a non-attenuation-corrected (NAC) positron emission tomography (PET) image, segmenting the NAC PET image to identify a body contour, and identifying a truncated or attenuated region in the CT image. The method further includes inferring a tissue type consistent with the region of truncated or attenuated region of the CT image using the body contour as a guide, projecting CT image data consistent with the inferred tissue type onto the truncated or attenuated region of the CT image to generate a corrected CT image data set, and backprojecting the corrected CT image data set to correct the CT image.

    [0010] One advantage is that CT or MR field of view is not enlarged.

    [0011] Another advantage resides in using the high-quality volume boundary provided by a non-attenuated corrected (NAC) PET image of a volume of interest.

    [0012] Still further advantages of the subject innovation will be appreciated by those of ordinary skill in the art upon reading and understand the following detailed description.

    [0013] The innovation may take form in various components and arrangements of components, and in various steps and arrangements of steps. The drawings are only for purposes of illustrating various aspects and are not to be construed as limiting the invention.

    FIGURE 1 illustrates a system that uses a non-attenuation corrected PET image to determine a perimeter or outside boundary of a subject or volume of interest.

    FIGURE 2 illustrates a method related to employing a NAC PET image to correct truncated attenuation maps, in accordance with various features.

    FIGURE 3 illustrates an exemplary hospital system with a plurality of imaging devices, such as PET, CT, MRI, or the like, which generate imaging data that are reconstructed by individual or shared reconstruction processors to generate 3D image representations.



    [0014] FIGURE 1 illustrates a system 10 that uses a non-attenuation corrected (NAC) PET image to determine a perimeter or outside boundary of a subject or volume of interest (VOI). When a PET scan is reconstructed without attenuation correction, the resultant image is dark near the center and progressively brighter toward the edges. This enables the PET scan to be used to delineate the periphery of the subject or VOI. Once the periphery is determined, it can be used in various ways. First, it may be used to replace the badly artifacted CT or MR image data between the field of view (FOV) and the patient periphery with nominal soft tissue, and the compensated CT or MR image is then used for attenuation correction in the PET scan. Although CT images are presented by way of example in this section, MR, SPECT-CT, projection x-ray, and the like are also contemplated. Alternately, once the periphery of the subject or VOI is known, this information can be used to reduce artifacting in the truncated CT or MR image. For example, the CT or MR image with the substituted soft tissue can be projected into projection data and re-backprojected. As another example, the portion of the subject outside the FOV can be assigned the attenuation properties of nominal soft tissue and the CT data can be adjusted ray by ray for the attenuation of the nominal soft tissue.

    [0015] The system 10 comprises a user interface 12 (e.g., a workstation or the like) that is coupled to each of a PET scanner 14 and a CT device 16. The user interface 12 comprises a processor 18, which is coupled to a memory 20, both of which are further coupled to a display 22. The memory 20 stores, and the processor generates, analyzes, and/or executes, image data 24 from the PET scanner 14 and/or the CT device 16, reconstruction algorithms 26 for reconstructing PET and/or CT image representations from acquired scan data, CT correction algorithms 28 for correcting CT data and/or images, attenuation correction map(s) 30 generated from CT data to compensate for attenuated PET data, PET correction algorithms 32 for correcting PET data and/or images, etc.

    [0016] In one embodiment, the processor 18 generates a truncated (e.g., uncorrected, or raw) attenuation map from acquired CT data. Truncated portions of the attenuation map are identified using prior knowledge, such as by comparing acquired CT data to expected values, previously acquired CT data for the subject, or a generic atlas of anatomical structures generated from a plurality of reference subjects. The processor also executes a PET reconstruction algorithm to reconstruct a NAC PET image from acquired PET data, and the NAC image is segmented to identify a contour of an anatomical structure, a tumor, or other structure in the subject, that corresponds to a truncated portion of the attenuation map. The processor executes one or more CT correction algorithms 28 to "fill in" portions of the CT attenuation map outside the FOV using the contour identified from the NAC PET image. For instance, if the attenuation map is truncated near an upper surface of a subject's liver, then the NAC PET image is segmented to identify a contour that delineates the upper surface of the subject's liver. The processor analyzes the contour and makes inferences regarding tissue type along or near the contour. Based on such inferences, the processor fills in truncated portions of the attenuation map with appropriate material, such as soft tissue, hepatic tissue, or some other tissue having a similar density to the tissue type inferred for the truncated portion of the attenuation map. The processor then executes the PET correction algorithm(s) 32 to reconstruct an attenuation-corrected PET image representation that is free of artifacts.

    [0017] According to another example, the FOV of the CT device 16 is approximately 60 cm. If a subject is larger than 60 cm in diameter or positioned with a portion of the subject outside the 60 cm FOV, the CT data will be truncated. Reconstructing the PET image without correcting for attenuation provides an image that is bright at the surface and dimmer toward the center. Thus, a high-quality contour showing the surface of the subject can be generated in the NAC PET image. Once the surface is delineated, truncation in the CT image can be corrected. For instance, soft tissue CT data can be interpolated into truncated portions of an MR attenuation map, since soft tissue is likely to be found near the surface (e.g., skin) of the subject.

    [0018] The system can be employed in PET/CT imaging scenarios where a portion of an object or structure close to the body contour is not visible in CT image. Alternatively, the system can be employed in PET/CT imaging scenarios where two modalities are performed sequentially (e.g., not simultaneously), and thus the contour derived from the PET scan can be compared to the CT image to infer potential subject motion between the PET and CT scans. Additionally, the system can be employed in PET imaging scenarios where the contour derived from the NAC PET image is used as emission boundary for scatter correction using single-scatter simulation, in which a tail-fitting procedure utilizes an emission boundary to define pure-scatter tails (e.g., in the absence of true coincidence events).

    [0019] FIGURE 2 illustrates a method related to employing a NAC PET image to correct truncated attenuation maps, in accordance with various features. While the methods are described as a series of acts, it will be appreciated that not all acts may be required to achieve the described goals and/or outcomes, and that some acts may, in accordance with certain aspects, be performed in an order different than the specific orders described.

    [0020] At 50, acquired PET data is reconstructed, without attenuation correction, to generate the NAC PET image. The NAC PET image is segmented, at 52, to identify one or more contours of a volume of interest (e.g., a tumor or anatomical structure, etc.) in a subject. At 54, an attenuation map generated from acquired CT or MR data is analyzed to identify truncated portions of CT or MR data, such as may occur due to attenuation, small CT or MR field of view, etc. At 56, identified truncated portions of CT or MR data are compensated for (e.g., filled in) using a contour corresponding to the anatomical location or position of the truncated CT or MR data as a guide. For instance, since the contour delineates a surface or boundary of the volume of interest, inferences can be made regarding the type and/or characteristics of tissue near the contour, and replacement CT or MR data can be interpolated as a function of such inferences to fill in truncated areas of the CT or MR attenuation map. At 58, an attenuation-corrected PET image is reconstructed using the corrected MR attenuation map.

    [0021] In one embodiment, truncated portions of the attenuation map are identified using known data acquisition algorithms that identify accuracy probabilities for data as it is acquired. The probability of accuracy of acquired data is evaluated at 54, and data with a likelihood of accuracy below a predetermined threshold is identified as truncated data. The truncated data is then replaced with interpolated or replacement CT data at 56.

    [0022] According to an example, data points in the acquired CT data set can be compared to neighboring data points to determine whether their respective values are consistent with the neighboring data points. If a given region has a large number of data points that do not correspond to expected values, then the probability that the data points in the given region are accurate is low. If the probability is below a predetermined threshold level, then the region can be identified as a truncated region, and earmarked for correction or compensation. For instance, a segmented body contour that corresponds to the anatomical location of the truncated CT data region can be generated from the NAC PET image and used as a guide in identifying a body tissue type that is employed when filling in the truncated portion of the CT data.

    [0023] In one embodiment for correcting the truncated CT image, the FOV of CT image is expanded, e.g., into a circle of a larger diameter. The shape of the truncated tissue determined from the NAC PET image. The truncated shape is filled with the attenuation values for appropriate tissue. In one example, the shape is filled with the attenuation value of nominal tissue. In another example in which the patient is off center to one side, the truncated shape can be filled with a mirror image of the attenuation value from the other side of the patient. In another example, a look up table or memory stores untruncated nominal attenuation images of patients. The corresponding image (slice) to the image (slice) being corrected is retrieved, scaled in accordance with the patient outline from the NAC PET image, and the corresponding portion of the scaled nominal image is used to fill the truncated shape.

    [0024] The truncated region filled image is forward projected back into projection data which is backprojected to generate a corrected image. If the corrected image is still artifacted, the truncated shape can be refilled with the same correction attenuation values from the table, of nominal tissue, etc. and the forward projection and backprojection process repeated.

    [0025] In another embodiment, the truncated shape is filled with the correction attenuation values and the contribution of the truncated shape to each ray of the original projection data. For example, after the truncated shape is filled with the correction attenuation values, the rest of the CT image is zeroed. The truncated shape is forward projected to generate correction CT projection data corresponding to the original projection data. Each connection projection is used to adjust the corresponding original projection, e.g., subtracted from it, to create corrected projection data which is backprojected to generate the corrected attenuation map. Optionally, the truncated shape with the synthesized attenuation values can be merged with the corrected attenuation map.

    [0026] With reference to FIGURE 3, an exemplary hospital system may include a plurality of imaging devices 100, such as PET 14, CT 16 (or MR), or the like, which generate imaging data that are reconstructed by individual or shared reconstruction processors 102 to generate 3D image representations. The image representations are communicated over a network 104 to a central memory 106 or a local memory 108.

    [0027] At a station 110 connected with the network, an operator uses user interface 12 to move a selected 3D patient CT or MR attenuation map to or between the central memory 106 and the local memory 108. A video processor 116 displays the selected attenuation map in a first viewport 1181, of the display 20. The NAC PET image is displayed in a second viewport 1182. A third view port 1183 can display an overlay of the attenuation map and the NAC PET image. For example, a user can be permitted to register landmarks in the PET and MR or CT attenuation map to corresponding structures or landmarks in the NAC PET image. For instance, the operator, through the interface 12, selects the NAC PET image landmarks (e.g., using a mouse, stylus, or other suitable user input device) that correspond to landmarks in the attenuation map image. Alternately, the NAC PET attenuation map can be aligned automatically by a program in the processor 116. The processor 18 (Fig. 1) in the user interface 12 then performs correction algorithms and infers an appropriate tissue type to employ when filling in truncated areas in the attenuation map.

    [0028] The corrected attenuation map can then be used to reconstruct an artifact-free attenuation-corrected PET image, which may be used in other applications. For instance, a therapy planning station 130 can use the attenuation-corrected PET image to plan a therapy session. Once planned to the satisfaction of the operator, the planned therapy can, where appropriate to an automated procedure, be transferred to a therapy device 132 that implements the planned session. Other stations may use the attenuation-corrected PET image in various other planning processes.

    [0029] In another embodiment, the overlay displayed in viewport 1183 is adjustable to weight the CT or MR image data relative to the PET image, or vice versa. For instance a slider bar or knob (not shown), which may be mechanical or presented on the display 20 and manipulated with an input device, may be adjusted to vary the weight of the image or the PET image. In one example, an operator can adjust the image in viewport 1183 from purely CT or MR image data (as is shown in viewport 1181), through multiple and/or continuous combinations of CT/MR and PET image data, to purely PET image data (as is shown in viewport 1182). For instance, a ratio of CT or MR image data to PET image data can be discretely or continuously adjusted from 0:1 to 1:0. As another option, the CT or MR image can be displayed in grayscale and the PET image can be colorized. Anatomical landmarks in the CT or MR image help relate the PET image to the subject.

    [0030] The innovation has been described with reference to several embodiments. Modifications and alterations may occur to others upon reading and understanding the preceding detailed description. It is intended that the innovation be construed as including all such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.


    Claims

    1. An anatomical image correction system (10), including:

    a memory (20) that stores truncated anatomical image data acquired during a scan of a subject and positron emission tomography (PET) data acquired during a PET scan of the subject; and

    a processor (18) that reconstructs a non-attenuation-corrected (NAC) PET image from the acquired PET data, generates an attenuation map (30) from the acquired truncated anatomical image data, and compensates for truncated image data in the attenuation map (30) using a contour in the NAC PET image.


     
    2. The system according to claim 1, wherein the processor (18) determines a likelihood of accuracy of data in the attenuation map (30).
     
    3. The system according to claim 2, wherein the processor (18) identifies a truncated region of the attenuation map (30) by determining whether truncated data in the region has a likelihood of accuracy below a predetermined threshold.
     
    4. The system according to any one of the preceding claims, wherein the processor (18):

    identifies a contour in the NAC PET image that corresponds to a truncated region of the attenuation map (30);

    infers a tissue type for use in correcting the truncated region of the attenuation map (30) as a function of the identified contour; and

    fills the truncated region with interpolated data consistent with the inferred tissue type to correct the truncated region.


     
    5. The system according to claim 4, wherein the processor (18) employs the corrected attenuation map to reconstruct an attenuation-corrected PET image from the PET data.
     
    6. The system according to any one of the preceding claims, further including an imaging device (16) that acquires the truncated anatomical image, and a PET scanner (14) that acquires the PET data.
     
    7. The system according to any one of the preceding claims, further including a display (22) that presents the NAC PET image, the attenuation map (30), and an overlay of the NAC PET image and the attenuation map (30), to a user.
     
    8. The system according to any one of the preceding claims, wherein the memory stores and the processor executes machine executable instructions, including:

    a routine (50) for generating the NAC PET image;

    a routine (52) for segmenting the NAC PET image to identify the contour;

    a routine (54) for identifying a truncated region in the attenuation map (30);

    a routine (56) for correcting the truncated region using the identified contour as a guide; and

    a routine (58) for reconstructing an attenuation-corrected PET image using the corrected attenuation map.


     
    9. A computer implemented method of correcting an attenuation map using a non-attenuation-corrected (NAC) positron emission tomography (PET) image, including:

    generating the NAC PET image;

    segmenting the NAC PET image to identify a body contour;

    identifying a truncated region in the attenuation map (30);

    correcting the truncated region of the attenuation map (30) using the identified contour to infer an appropriate tissue type to use for truncation compensation; and

    reconstructing an attenuation-corrected PET image using the corrected attenuation map.


     
    10. The method according to claim 9, further including:

    sequentially performing a scan to acquire the truncated anatomical image data used to generate the attenuation map (30) and a PET scan to acquire PET data used to generate the NAC PET image and the attenuation corrected PET image; and

    using the acquired truncated image data and the body contour derived from the acquired PET data to identify patient motion between the truncated scan and the PET scan.


     
    11. The method according to either one of claims 9 or 10, further including displaying the attenuation map (30), the NAC PET image, and an overlay of the NAC PET image and the attenuation map, to a user.
     
    12. The method according to any one of the claims 9-11, further including:

    determining a likelihood of accuracy of data in the attenuation map (30); and

    identifying the truncated region of the attenuation map (30) by determining whether truncated data in the region has a likelihood of accuracy below a predetermined threshold.


     
    13. The method according to any one of the claims 9-12, further including identifying a plurality of body contours and selecting a body contour corresponding to the truncated region for use in correcting the truncated data in the attenuation map (30) after identifying the truncated region.
     
    14. A computer readable medium having stored thereon software for controlling one or more computers to perform the method according to any one of the claims 9-13.
     
    15. The method according to any one of the claims 9-13, further including employing the body contour as an emission boundary, executing a single-scatter simulation procedure for scatter correction, and executing a tail-fitting procedure to define and correct pure-scatter tails when reconstructing the attenuation-corrected PET image.
     


    Ansprüche

    1. Anatomisches Bildkorrektursystem (10), umfassend:

    einen Speicher (20), in dem abgeschnittene anatomische Bilddaten, die während eines Scans eines Subjekts erfasst wurden, und Positronenemissionstomographie (PET)-Daten, die während eines PET-Scans des Subjekts erfasst wurden, gespeichert werden; und

    einen Prozessor (18), der ein nicht-dämpfungskorrigiertes (non-attenuation corrected, NAC) PET-Bild aus den erfassten PET-Daten rekonstruiert, anhand der erfassten abgeschnittenen anatomischen Bilddaten eine Dämpfungskarte (30) erzeugt und unter Verwendung einer Kontur in dem NAC-PET-Bild eine Korrektur bezüglich der abgeschnittenen Bilddaten in der Dämpfungskarte (30) vornimmt.


     
    2. System nach Anspruch 1, wobei der Prozessor (18) eine Genauigkeitswahrscheinlichkeit der Daten in der Dämpfungskarte (30) ermittelt.
     
    3. System nach Anspruch 2, wobei der Prozessor (18) eine abgeschnittene Region der Dämpfungskarte (30) identifiziert, indem er ermittelt, ob abgeschnittene Daten in der Region eine Genauigkeitswahrscheinlichkeit unter einem vorgegebenen Schwellenwert aufweisen.
     
    4. System nach einem der vorhergehenden Ansprüche, wobei der Prozessor (18):

    eine Kontur in dem NAC-PET-Bild identifiziert, die einer abgeschnittenen Region der Dämpfungskarte (30) entspricht;

    eine Folgerung auf einen Gewebetyp zur Verwendung bei der Korrektur der abgeschnittenen Region der Dämpfungskarte (30) als eine Funktion der identifizierten Kontur vornimmt; und

    die abgeschnittene Region mit interpolierten Daten entsprechend dem gefolgerten Gewebetyp füllt, um die abgeschnittene Region zu korrigieren.


     
    5. System nach Anspruch 4, wobei der Prozessor (18) die korrigierte Dämpfungskarte nutzt, um aus den PET-Daten ein dämpfungskorrigiertes PET-Bild zu rekonstruieren.
     
    6. System nach einem der vorhergehenden Ansprüche, ferner umfassend eine Bildgebungsvorrichtung (16), die das abgeschnittene anatomische Bild erfasst, und einen PET-Scanner (14), der die PET-Daten erfasst.
     
    7. System nach einem der vorhergehenden Ansprüche, ferner umfassend eine Anzeige (22), die das NAC-PET-Bild, die Dämpfungskarte (30) und eine Überlagerung des NAC-PET-Bilds und der Dämpfungskarte (30) für einen Benutzer darstellt.
     
    8. System nach einem der vorhergehenden Anspruche, wobei der Speicher maschinenausführbare Anweisungen speichert und der Prozessor maschinenausführbare Anweisungen ausführt, einschließlich:

    einer Routine (50) zum Erzeugen des NAC-PET-Bilds;

    einer Routine (52) zum Segmentieren des NAC-PET-Bilds zum Identifizieren der Kontur;

    einer Routine (54) zum Identifizieren einer abgeschnittenen Region in der Dämpfungskarte (30);

    einer Routine (56) zum Korrigieren der abgeschnittenen Region unter Verwendung der identifizierten Kontur als Leitfaden; und

    einer Routine (58) zum Rekonstruieren eines dämpfungskorrigierten PET-Bilds unter Verwendung der korrigierten Dämpfungskarte.


     
    9. Computerimplementiertes Verfahren zum Korrigieren einer Dämpfungskarte unter Verwendung eines nicht-dämpfungskorrigierten (non-attenuation corrected, NAC) Positronenemissionstomographie (PET)-Bilds, umfassend:

    Erzeugen des NAC-PET-Bilds;

    Segmentieren des NAC-PET-Bilds, um eine Körperkontur zu identifizieren;

    Identifizieren einer abgeschnittenen Region in der Dämpfungskarte (30);

    Korrigieren der abgeschnittenen Region der Dämpfungskarte (30) unter Verwendung der identifizierten Kontur, um eine Folgerung auf einen geeigneten Gewebetyp zur Verwendung für die Abschneidekompensation vorzunehmen; und

    Rekonstruieren eines dämpfungskorrigierten PET-Bilds unter Verwendung der korrigierten Dämpfungskarte.


     
    10. Verfahren nach Anspruch 9, ferner umfassend:

    sequentielles Durchführen eines Scans zum Erfassen der abgeschnittenen anatomischen Bilddaten, die zur Erzeugung der Dämpfungskarte (30) verwendet werden, und eines PET-Scans zum Erfassen der PET-Daten, die zur Erzeugung des NAC-PET-Bilds und des dämpfungskorrigierten PET-Bilds verwendet werden; und

    Verwenden der erfassten abgeschnittenen Bilddaten und der von den erfassten PET-Daten abgeleiteten Körperkontur, um Patientenbewegung zwischen dem abgeschnittenen Scan und dem PET-Scan zu erkennen.


     
    11. Verfahren nach einem der Ansprüche 9 oder 10, ferner umfassend das Anzeigen der Dämpfungskarte (30), des NAC-PET-Bilds und einer Überlagerung des NAC-PET-Bilds und der Dämpfungskarte für einen Benutzer.
     
    12. Verfahren nach einem der Ansprüche 9 bis 11, ferner umfassend:

    Ermitteln einer Genauigkeitswahrscheinlichkeit der Daten in der Dämpfungskarte (30); und

    Identifizieren der abgeschnittenen Region der Dämpfungskarte (30) durch Ermitteln, ob abgeschnittene Daten in der Region eine Genauigkeitswahrscheinlichkeit unter einem vorgegebenen Schwellenwert aufweisen.


     
    13. Verfahren nach einem der Ansprüche 9 bis 12, ferner umfassend:
    Identifizieren einer Vielzahl von Körperkonturen und Auswählen einer der abgeschnittenen Region entsprechenden Köperkontur zur Verwendung bei der Korrektur der abgeschnittenen Daten in der Dämpfungskarte (30) nach dem Identifizieren der abgeschnittenen Region.
     
    14. Computerlesbares Medium, auf dem Software zum Steuern von einem oder mehreren Computern zur Durchführung des Verfahrens nach einem der Ansprüche 9 bis 13 gespeichert ist.
     
    15. Verfahren nach einem der Ansprüche 9 bis 13, ferner umfassend das Nutzen der Körperkontur als eine Emissionsgrenze, das Ausführen einer Einzelstreuungs-Simulationsprozedur zur Streuungskorrektur, das Ausführen einer Tail-Fitting-Prozedur zum Definieren und Korrigieren von reinen Streuungs-Tails beim Rekonstruieren des dämpfungskorrigierten PET-Bilds.
     


    Revendications

    1. Système de correction d'images anatomiques (10) comprenant :

    une mémoire (20) qui stocke des données d'images anatomiques tronquées acquises au cours d'un balayage d'un sujet et des données de tomographie à émission de positons (PET) acquises au cours d'un balayage PET du sujet ; et

    un processeur (18) qui reconstruit une image PET non corrigée en atténuation (NAC) à partir des données PET acquises, génère une carte d'atténuation (30) à partir des données d'images anatomiques tronquées acquises et compense les données d'images tronquées de la carte d'atténuation (30) en utilisant un contour dans l'image PET NAC.


     
    2. Système selon la revendication 1, dans lequel le processeur (18) détermine une vraisemblance de précision de données dans la carte d'atténuation (30).
     
    3. Système selon la revendication 2, dans lequel le processeur (18) identifie une région tronquée de la carte d'atténuation (30) en déterminant si les données tronquées de la région ont ou non une vraisemblance de précision en dessous d'un seuil prédéterminé.
     
    4. Système selon l'une quelconque des revendications précédentes, dans lequel le processeur (18) :

    identifie un contour dans l'image PET NAC qui correspond à une région tronquée de la carte d'atténuation (30) ;

    déduit un type de tissu pour utilisation dans la correction de la région tronquée de la carte d'atténuation (30) en fonction du contour identifié ; et

    remplit la région tronquée par des données interpolées compatibles avec le type de tissu déduit pour corriger la région tronquée.


     
    5. Système selon la revendication 4, dans lequel le processeur (18) emploie la carte d'atténuation corrigée pour reconstruire une image PET corrigée en atténuation à partir des données PET.
     
    6. Système selon l'une quelconque des revendications précédentes, comprenant en outre un dispositif d'imagerie (16) qui acquiert l'image anatomique tronquée et un scanneur PET (14) qui acquiert les données PET.
     
    7. Système selon l'une quelconque des revendications précédentes, comprenant en outre un appareil d'affichage (22) qui présente l'image PET NAC, la carte d'atténuation (30) et une superposition de l'image PET NAC et de la carte d'atténuation (30) à un utilisateur.
     
    8. Système selon l'une quelconque des revendications précédentes, dans lequel la mémoire stocke et le processeur exécute des instructions exécutables par machine comprenant :

    un programme (50) pour générer l'image PET NAC ;

    un programme (52) pour segmenter l'image PET NAC afin d'identifier le contour ;

    un programme (54) pour identifier une région tronquée dans la carte d'atténuation (30) ;

    un programme (56) pour corriger la région tronquée en utilisant le contour identifié comme guide ; et

    un programme (58) pour reconstruire une image PET corrigée en atténuation en utilisant la carte d'atténuation corrigée.


     
    9. Procédé mis en oeuvre par ordinateur pour corriger une carte d'atténuation en utilisant une image de tomographie à émission de positons (PET) non corrigée en atténuation (NAC), comprenant :

    la génération de l'image PET NAC ;

    la segmentation de l'image PET NAC pour identifier un contour corporel ;

    l'identification d'une région tronquée dans la carte d'atténuation (30) ;

    la correction de la région tronquée de la carte d'atténuation (30) en utilisant le contour identifié pour déduire un type de tissu approprié pour utilisation pour la compensation du troncage ; et

    la reconstruction de l'image PET corrigée en atténuation en utilisant la carte d'atténuation corrigée.


     
    10. Procédé selon la revendication 9, comprenant en outre :

    en séquence la réalisation d'un balayage pour acquérir les données d'images anatomiques tronquées utilisées pour générer la carte d'atténuation (30) et d'un balayage PET pour acquérir des données PET utilisées pour générer une image PET NAC et l'image PET corrigée en atténuation ; et

    l'utilisation des données d'images tronquées acquises et le contour corporel dérivé des données PET acquises pour identifier un mouvement du patient entre le balayage tronqué et le balayage PET.


     
    11. Procédé selon l'une quelconque des revendications 9 ou 10, comprenant en outre l'affichage de la carte d'atténuation (30), de l'image PET NAC et d'une superposition de l'image PET NAC et de la carte d'atténuation à un utilisateur.
     
    12. Procédé selon l'une quelconque des revendications 9 à 11, comprenant en outre :

    la détermination d'une vraisemblance de précision de données dans la carte d'atténuation (30) ; et

    l'identification de la région tronquée de la carte d'atténuation (30) en déterminant si des données tronquées de la région ou non ont une vraisemblance de précision en dessous d'un seuil prédéterminé.


     
    13. Procédé selon l'une quelconque des revendications 9 à 12, comprenant en outre une identification d'une pluralité de contours corporels et la sélection d'un contour corporel correspondant à la région tronquée pour utilisation dans la correction des données tronquées de la carte d'atténuation (30) après identification de la région tronquée.
     
    14. Support lisible par ordinateur sur lequel est stocké un logiciel pour commander un ou plusieurs ordinateurs afin d'effectuer le procédé selon l'une quelconque des revendications 9 à 13.
     
    15. Procédé selon l'une quelconque des revendications 9 à 13, comprenant en outre l'emploi du contour corporel comme limite d'émission, l'exécution d'une procédure de simulation de dispersion individuelle pour corriger la dispersion et l'exécution d'une procédure d'ajustement de queue afin de définir et de corriger des queues de pure dispersion lors de la reconstruction de l'image PET corrigée en atténuation.
     




    Drawing














    Cited references

    REFERENCES CITED IN THE DESCRIPTION



    This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

    Patent documents cited in the description