Estimating position of an organ with a biomechanical model

Abstract

There is presented a method 100 and apparatus 200 to measure the surface of the patient (thorax and abdominal regions), e.g., during therapy delivery and (if necessary) while imaging. Together with biomechanical considerations the position of internal structures of the patient, such as an organ, and optionally a tumor in an organ, is inferred from the measured patient surface. In case the patient breaths and thus the organ and/or tumor moves, the position may be determined, which may be advantageous during, e.g., radiation therapy, since it enables that whenever the tumor is at the right position according to the radiation therapy plan, the radiation is switched on. In a specific embodiment, a finite element model is employed.

Claims

1. A method for estimating organ coordinates of at least a part of an organ which may move and/or change shape due to breathing in an associated patient, the method comprising: using a processor, obtaining a biomechanical model of said associated patient, said biomechanical model being based on three-dimensional data indicative of only internal structures of a thorax of said associated patient at a first point in time; using a surface determining unit comprising a surface scanner, a fiber optic shape sensing (OSS) garment, or a plurality of cameras, obtaining surface coordinates indicative of a position and shape at a second point in time of at least a part of the surface of the thorax of said associated patient and at least a part of the surface of the abdomen of said associated patient so as to enable estimating an external abdominal volume; and using the processor, obtaining the organ coordinates; including at least organ coordinates of a diaphragm by inputting said surface coordinates in said biomechanical model and outputting from said biomechanical model the organ coordinates, wherein said biomechanical model models at least a part of an internal abdominal volume, and tissue of the chest wall of said patient, as incompressible.

2. A method according to claim 1, wherein said biomechanical model employs as constraints, at least a part of the internal abdominal volume, and the surface coordinates indicative of a position and shape of at least a part of the surface of the thorax of said associated patient.

3. A method according to claim 1, wherein said biomechanical model is based on three-dimensional data indicative of only internal structures of the thorax of said associated patient obtained by a single 3D scan.

4. A method according to claim 1, wherein the surface coordinates serve as the exclusive input to the biomechanical model in the step of obtaining organ coordinates at a second point in time.

5. A method according to claim 1, wherein said organ is a lung including the lower boundary comprising the diaphragm.

6. A method according to claim 1, wherein the method further comprising determining a tumor coordinate indicative of a position of a tumor in the organ, wherein the determination is based on the organ coordinates.

7. A method according to claim 6, wherein the method further comprising repeatedly determining a tumor coordinate indicative of a position of a tumor in the organ, wherein the determination is based on the organ coordinates.

8. A method according to claim 6, wherein the biomechanical model comprises a patient-specific mesh of chest, the organ and abdomen.

9. A method according to claim 6, wherein the method further comprising any one of: outputting the tumor coordinate, so as to enable an associated device to receive the tumor coordinate, outputting a signal indicative of whether or not the tumor coordinate is within a pre-determined volume.

10. An apparatus for estimating coordinates of at least a part of an organ which may move due to breathing in an associated patient, the apparatus comprising: a non-transitory computer-readable medium suitable for comprising a biomechanical model of said associated patient, said biomechanical model being based on three-dimensional data indicative of internal structures of a thorax of said associated patient and a portion of an abdomen of said patient at a first point in time, a surface determining unit for obtaining surface coordinates indicative of a position and shape at a second point in time of at least a part of the surface of the thorax of said associated patient, and at least a part of the surface of the abdomen of said associated patient so as to enable estimating an external abdominal volume, a processor operatively connected to the computer readable medium and the surface determining unit for obtaining surface coordinates indicative of a position and shape of the surface of said associated patient at the second point in time, the processor being arranged for obtaining the organ coordinates, by receiving said surface coordinates, inputting said surface coordinates in said biomechanical model, and outputting from said biomechanical model the organ coordinates, wherein said biomechanical model is configured to models at least a part of an internal abdominal volume, and tissue of the chest wall, as incompressible.

11. An apparatus according to claim 10, wherein the surface determining unit for obtaining surface coordinates indicative of a position and shape of the surface of said associated patient, comprises any one of: a surface scanner, a fiber optic shape sensing (OSS) garment, and/or a plurality of cameras enabling photogrammetry.

12. An apparatus according to claim 10, further comprising a source of radiation for external beam therapy which is operatively connected to the processor.

13. An apparatus according to claim 12, wherein the source of radiation for external beam therapy is arranged for functioning dependent on the position of tumor.

14. An apparatus according to claim 10, wherein the computer-readable medium comprises the biomechanical model of said associated patient.

15. A non-transitory computer readable medium having a computer program stored thereon, the computer program enabling a processor to carry out the method of claim 1.

16. The apparatus according to claim 10, wherein the processor is further programmed to: determine a tumor coordinate indicative of a position of a tumor in the organ, wherein the determination is based on the organ coordinates.

17. An apparatus, comprising: a database storing a biomechanical model of an associated patient, the biomechanical model being based on three-dimensional data indicative of internal structures of at least a thorax of said associated patient at a first point in time; a surface determining unit comprising a surface scanner, a fiber optic shape sensing (OSS) garment, or a plurality of cameras configured to obtain surface coordinates indicative of a position and shape at a second point in time of at least a part of the surface of the thorax of said associated patient and at least a part of the surface of the abdomen of said associated patient so as to enable estimating an external abdominal volume; a processor operatively connected to the database and imaging device, the processor being programmed to: retrieve said biomechanical model from the database; receive said surface coordinates from the imaging device; input said surface coordinates into said biomechanical model; and output from said biomechanical model organ coordinates including at least organ coordinates of a diaphragm, wherein said biomechanical model is configured to model at least a part of an internal abdominal volume, and tissue of the chest wall, as incompressible.

18. The apparatus according to claim 17, wherein the processor is further programmed to: determine a tumor coordinate indicative of a position of a tumor in the organ from the organ coordinates.

19. The apparatus according to claim 18, further comprising: a radiation therapy source configured to deliver therapy to the associated patient; wherein the processor is configured to control operation of the radiation therapy source to delivery therapy to the determined position of the tumor.

20. The apparatus according to claim 17, wherein the processor is programmed to input only said surface coordinates into said biomechanical model whereby the biomechanical model outputs the organ coordinates including at least organ coordinates of the diaphragm using only said surface coordinates.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) Embodiments of the invention will be described, by way of example only, with reference to the drawings, in which

(2) FIG. 1 is a flow chart illustrating a method according to an embodiment of the invention,

(3) FIG. 2 illustrates an apparatus according to an embodiment of the invention,

(4) FIG. 3 is a schematic illustrating an embodiment of the invention,

(5) FIGS. 4-5 show diaphragm motion during breathing,

(6) FIG. 6 is a graph showing diaphragm position vs. abdominal volume,

(7) FIG. 7 is a flow chart illustrating a method according to an embodiment of the invention,

(8) FIGS. 8-9 illustrate examples of a lung finite element mesh and applied displacement boundary conditions.

DESCRIPTION OF EMBODIMENTS

(9) FIG. 1 is a flow chart illustrating a method (100) according to an embodiment of the invention for estimating organ coordinates, such as organ coordinates indicative of a position and shape, of at least a part of an organ which may move and/or change shape due to breathing in an associated patient, the method comprising:

(10) obtaining s102 a biomechanical model of said associated patient, said biomechanical model being based on three-dimensional data indicative of only internal structures of the thorax of said associated patient at a first point in time,

(11) obtaining s104 surface coordinates indicative of a position and shape at a second point in time of at least a part of the surface of the thorax of said associated patient, and at least a part of the surface of the abdomen of said associated patient so as to enable estimating an external abdominal volume,

(12) obtaining s106 the organ coordinates, by inputting said surface coordinates in said biomechanical model, and outputting from said biomechanical model the organ coordinates, wherein said biomechanical model models

(13) internal abdominal volume, and

(14) tissue of the chest wall of said patient,

(15) as substantially incompressible,

(16) wherein the method is further comprising determining s108 a tumor coordinate indicative of a position of a tumor in the organ, wherein the determination is based on the organ coordinates,

(17) wherein the method further comprising:

(18) outputting s110 the tumor coordinate, so as to enable an associated device to receive the tumor coordinate.

(19) FIG. 2 illustrates an apparatus (220) according to an embodiment of the invention for estimating coordinates, such as coordinates indicative of a position and shape, of at least a part of an organ which may move due to breathing in an associated patient (330), the apparatus comprising:

(20) a computer-readable medium (222) suitable for comprising a biomechanical model (352) of said associated patient, said biomechanical model being based on three-dimensional data indicative of only internal structures of the thorax of said associated patient of said associated patient at a first point in time,

(21) a surface determining unit (224) for obtaining surface coordinates indicative of a position and shape at a second point in time of at least a part of the surface of the thorax (356) of said associated patient, and at least a part of the surface of the abdomen (354) of said associated patient,

(22) a processor (226) operatively connected to the computer readable medium and the surface determining unit (224) for obtaining surface coordinates indicative of a position and shape of the surface of said associated patient at the second point in time, the processor being arranged for obtaining the organ coordinates, by receiving said surface coordinates, inputting said surface coordinates in said biomechanical model, and outputting from said biomechanical model the organ coordinates, wherein said biomechanical model models

(23) abdominal volume, and

(24) tissue of the chest wall, as incompressible.

(25) FIG. 3 is a schematic illustrating an embodiment of the invention where the organ is a lung, the figure showing an associated patient 330 which patient may in a first imaging phase 332 (which may correspond to a first point in time) be imaged by a surface determining unit 224 for obtaining surface coordinates 336 corresponding to the configuration of the lung and abdomen during the first imaging phase, such as surface coordinates indicative of a position and shape of at least a part of the surface of the thorax 340 of said associated patient, and at least a part of the surface of the abdomen 338 of said associated patient. The patient may furthermore in the first imaging phase 332 be imaged by an imaging unit 325 for obtaining 3D data 342 indicative of positions of internal structures, which imaging unit may be a CT-scanner or an NMR scanner. The 3D data 342 may be received by an image segmentation unit 344, which may provide information regarding lung surface 346, diaphragm surface 348, and abdomen surface 350 during the first imaging phase. The three-dimensional data may be input to a biomechanical model, such as a biomechanical model 352 of the thorax and the abdomen. It may be understood that the three-dimensional data may include surface coordinates 336 and data from the image segmentation unit 344.

(26) The patient may in a second imaging phase 334 (which may correspond to a second point in time) be imaged by a surface determining unit 224 for obtaining surface coordinates corresponding to the configuration of the lung and abdomen during the second imaging phase, such as surface coordinates indicative of a position and shape of at least a part of the surface of the thorax 356 of said associated patient, and at least a part of the surface of the abdomen 354 of said associated patient. The surface coordinates obtained in the second imaging phase 334 may be input to the biomechanical model 352.

(27) Although the configuration of the lung and abdomen in the first and second imaging phase are not necessarily the same, such as different, such as corresponding to different temporal positions in the breathing cycle, the biomechanical model 352 may provide data 358 corresponding to lung surface and data 360 corresponding to diaphragm surface in the second imaging phase 334.

(28) The data 358 corresponding to lung surface and data 360 corresponding to diaphragm surface in the second imaging phase 334 may be input to a biomechanical model 362 of lung and tumor, which may in turn output a tumor position 364 in the second imaging phase.

(29) In a specific embodiment, the second imaging phase includes radiation therapy. In that embodiment, the 3D data 342 may also be input to a radiation therapy planning workstation 366, which may output a radiation therapy plan 368. The tumor position 364 as well as the radiation therapy plan 368 may be sent to an online control computer 370 for controlling a radiation source 374, such as a linear accelerator. The online control computer 370 may then control the radiation source accordingly, in dependence of the radiation therapy plan 368 and the tumor position 364, such as control whether the radiation source should be on or off 372

(30) In the following the dataflow and workflow an exemplary specific embodiment where the second imaging phase includes radiation therapy for lung cancer radiation therapy is described (see FIG. 3):

(31) Before therapy (cf., the first imaging phase):

(32) 1. The thorax of the patient is scanned during a breath hold and a 3D CT scan is created. Structures of interest, such as all structures of interest are covered; for example the lung, the diaphragm, and the tumor.

(33) 2. For the biomechanical calculation of the diaphragm position, the knowledge of the shape of the abdomen may be utilized. Due to the X-ray dose associated with CT imaging, it is in general not beneficial to image the abdomen together with the thorax. Therefore, the surface scanner measures the patient's surface during 3D CT imaging. Treatment planning is done based on the 3D CT scan. In an alternative embodiment, which may be advantageous if there is no surface scanner available during CT imaging, the abdomen is imaged together with the thorax using CT imaging.
3. Segmentations of lung, diaphragm, tumor, skin, and other relevant structures are available after treatment planning.
4. External beam radiation therapy is planned based on this 3D CT scan. During therapy (cf., the second imaging phase):
5. For radiation therapy, the patient is positioned in the radiation therapy room and the LINAC is aligned according to the treatment plan.
6. A surface scanner measures the surface of the patient (thorax and abdominal regions).
7. Based on the surface measurement (thorax and abdomen) and the segmentation results (lung and diaphragm), the shape of the thoracic wall, i.e. the outer surface of the lung, is calculated taking into account biomechanical considerations.
8. Based on the outer shape of the lung, the position of the tumor is estimated e.g. with the help of biomechanical models.
9. When the tumor is at the planned position (and the patient position and shape is sufficiently close to the planning), radiation delivery is triggered and the beam is switched on.

(34) FIGS. 4-5 show diaphragm motion during breathing. It may be seen as a basic insight of the present inventors that the internal abdominal volume, including abdominal tissue, as well as the tissue of the chest wall is almost incompressible. Therefore, any change of the abdominal volume is related to a corresponding change of the diaphragm position, any change in the outer surface of the chest is related to a corresponding change of the lung surface.

(35) FIG. 4 shows that diaphragm motion (as indicated by the left arrow) is highly correlated with external abdominal volume as measured from the outside (as indicated by the right arrow).

(36) FIG. 5 shows that motion of the outside of the chest (as indicated by the upper arrow) is highly correlated with inside motion i.e. lung surface motion (as indicated by the lower arrow).

(37) FIG. 6 is a graph showing correlation of diaphragm position and abdominal volume for one of the volunteers from the study. More specifically, the graph shows diaphragm position in head-feet direction vs. external abdominal volume. The horizontal axis shows external abdominal volume (EAV) in cubic millimeters. The vertical axis shows diaphragm position (DP) in millimeters. The data markers represent paradox breathing (676), abdominal breathing (677), thoracal breathing (678) and normal breathing (679). In principle, the gas, which is included in the abdomen e.g. in the stomach or the intestines, is compressible. However, in the one hand the abdominal gas volume is relatively small in relation to the abdominal tissue; on the other hand, pressure changes during breathing are small in the abdomen. Therefore, volume changes of abdominal gas can be neglected as well. Real-time 4D MRI data has been acquired during a volunteer study at Philips Research in Hamburg (Measurement time for a complete volume 0.8 sec). A high correlation of external abdominal volume with diaphragm position has been detected. It is important to note that this correlation is independent of the breathing pattern or depth of breathing. It has been observed for normal, thoracal, abdominal, as well as paradox breathing patterns with deep and flat amplitude (see FIG. 6 for an example from the volunteer study mentioned above).

(38) Diaphragm:

(39) Based on biomechanical considerations, the diaphragm position can be predicted surprisingly well from a measurement of the complete abdominal volume. This has been shown with the help of a volunteer study in which 4D real-time MRI of the thorax and the abdomen was used to assess the breathing motion. The volunteers were asked to breath with flat and deep amplitudes in 4 different breathing patterns: normal, abdominal, thoracal, and paradox breathing. The latter is a yoga exercise with unnatural movements. 3D volumes were recorded every 0.7-0.8 seconds depending on the height of the volunteer. The biomechanically motivated correlation between abdominal volume and diaphragm position was verified for all volunteers, all amplitudes, and all breathing patterns.

(40) Thorax:

(41) The shape of the skin surface of the chest can be measured with the help of the devices such as the surface determining unit for obtaining surface coordinates described elsewhere in the present text. In a first order approximation the thickness of the thorax measured from lung surface to skin is constant. Therefore, it is possible to estimate the lung surface starting from a 3D image of the internal structures of the thorax together with measurements of the outer shape of the patient's thorax.

(42) Together with biomechanical considerations the position of a tumor inside the organ of a patient is inferred from the measured patient surface. This may be relevant, e.g., during therapy delivery, where the patient breaths and thus the tumor moves. Whenever the tumor is at the right position according to the radiation therapy plan, the radiation is switched on.

(43) Lung Surface Estimation

(44) The surface of the organ, such as the lung, is estimated based on the surface coordinates, such as the measurements of the surface scanner. A first order approximation to estimate the lung surface is to assume, that the thickness of the chest wall does not change during breathing. Therefore, the distance of lung surface to outer body surface is constant in the area of the chest. More complex biomechanical models can be used as well. In order to estimate the lower boundary of the lung, the diaphragm, more complex calculations may be carried out. The diaphragm might not be seen directly in CT and MRI images. It consists of muscular and ligamentous parts. The muscular (elastic) parts of the diaphragm are fixed to anatomical structures, which can be segmented from CT and MRI images (ribs, vertebra, and pericardium). The ligamentous (inelastic) parts can be found at the bottom of the lung forming the border between lung and abdominal cavity. A first order approximation to the estimation of the lower long surface is thus to take into account the elastic parts of the diaphragm during breathing and to calculate its shape based on the surface coordinates, such as the measured outer shape and the estimated external abdominal volume.

(45) Tumor Position Estimation

(46) Once the organ, such as lung, surface is estimated based on the measurement of the outer shape of the patient's body, the boundary conditions for a biomechanical model of the lung including the tumor are known. In a first order approximation, the lung tissue can be modeled as a spongy tissue in which an incompressible (or even rigid) tumor is embedded. FEM methods can be used to predict the tumor position based on this model.

(47) In another embodiment there is provided a method wherein the biomechanical model comprises a patient-specific mesh of chest, the organ and abdomen. In a particular embodiment, there is provided:

(48) 1. A patient-specific mesh of chest, lung, and abdomen. This mesh can be created from preoperative CT scans that are routinely acquired.

(49) 2. Surface coordinates, such as displacement surrogate signals (a combination of):

(50) a. Motion of the chest

(51) b. Motion of the umbilicus

(52) c. Motion of several skin markers on the abdomen

(53) d. Volume of the abdomen

(54) These parameters can be measured using optical shape sensing or other tracking methods such as optical or electromagnetic trackers.

(55) 3. A model-based approach to calculate the motion of the diaphragm from the surrogate signals.

(56) 4. An FE-based algorithm to compute the motion of the lung from the diaphragm motion.

(57) The model could be divided into two parts:

(58) The first part calculates the motion of diaphragm from the surrogate signal, measured in the abdominal region.

(59) The second part calculates the lung motion from the calculated diaphragm motion in part 1 and chest motion surrogate signal.

(60) In order to calculate the diaphragm motion form the surrogate signal a learning approach may be employed. In this method the diaphragm motion pattern is learned from a set of clinical data. There are several methods available to learn the diaphragm motion pattern. In one embodiment, principal component analysis can be used. In this method, the diaphragm motion of a group of patient is transferred into its principal components. A few components with the highest corresponding singular value can be selected as representatives of the diaphragm motion pattern. Other embodiments can include model fitting approaches or neural networks. Then the patient specific model is solved in a way that the calculated diaphragm motion has a similar pattern to the learnt pattern.

(61) In one embodiment an FE-based model can be used to calculate the diaphragm motion from the surrogate signal. In this case, the displacement vector u is confined to have the same pattern as the learnt displacement vector (a weighted summation of principal components). Moreover, the known displacements from the surrogate signals can be exchanged with unknown forces in equation (1) so that equation is sufficiently constrained to be solved. Note that most of entries in f are zero as there is no external force. In this case, the diaphragm motion can be calculated without a need to measure forces or pressures or use an inaccurate predefined value.

(62) The second part of the algorithm uses the calculated diaphragm motion from the first part. We assume that the lower portion of the lung is attached to the diaphragm and moves with it. Therefore, in equation (1) some of the displacement will be known that can be exchanged for corresponding unknown forces. The same can be said about the chest wall motion if measured. Therefore, the lung motion can be calculated without incorporation of force or pressure values.

(63) FIGS. 8-9 illustrate examples of a lung finite element mesh and applied displacement boundary conditions. More specifically, FIGS. 8-9 show two examples of the lung finite element mesh and applied displacement boundary conditions.

(64) FIG. 8 shows top nodes 890 being fixed to the chest bones, bottom nodes 891 move with the diaphragm and the side nodes are free.

(65) FIG. 9 shows top nodes 990 and side nodes are fixed to the chest bones, bottom nodes 991 move with the diaphragm and the side nodes are free.

(66) FIG. 7 is a flow chart illustrating a method according to an embodiment of the invention where the organ is a lung, wherein

(67) three-dimensional data is acquired 780, such as acquired with CT scanning (similar to the patient being imaged by an imaging unit 325 for obtaining 3D data 342 indicative of positions of internal structures, which imaging unit may be a CT-scanner or an NMR scanner in FIG. 3),

(68) the three-dimensional data is segmented 781 into lung, chest and abdomen (similar to the 3D data 342 being received by an image segmentation unit 344, which may provide information regarding lung surface 346, diaphragm surface 348, and abdomen surface 350 during the first imaging phase in FIG. 3),

(69) a patient specific mesh is generated 782,

(70) a finite element model 783 of the abdomen is generated,

(71) learnt constraints, the finite element model 783 of the abdomen and surface coordinates 785 are employed to calculate 786 diaphragm motion,

(72) a finite element model 787 of the chest and lung is generated,

(73) the finite element model 787 of the chest and lung, surface coordinates 788 indicative of a position and shape of at least a part of the surface of the thorax (356) of said associated patient, and the diaphragm motion 786 are employed to calculate lung motion.

(74) In addition to or as an alternative to the CT datasets, other imaging datasets such as MRI, Ultrasound, X-ray, PET-SPECT, flow etc. may be used. During therapy, the breathing pattern and cycle can also be identified.

(75) To achieve a patient-specific model from a generic database of the population, specific features can be extracted. One example of the same would be the shape of the diaphragm, wherein a person having a flat diaphragm could match better with those in the database that have a similar shape. Another example is the size of the diaphragm which may also be correlated with the patient's size and weight.

(76) In order to accurately predict lung and tumor motion, the motion at certain points like the umbilicus or sternum may be computed and transformed to the abdominal volume. To accommodate for variations in breathing patterns, each pattern could be sorted into a multi-dimensional phase space.

(77) It is understood by one skilled in the art that any patient or volunteer database may comprise datasets that are normalized, in order to account for inter-trial variations.

(78) To sum up, the invention provides a method and apparatus to measure the surface of the patient (thorax and abdominal regions), e.g., during therapy delivery and (optionally, such as if necessary) while imaging. Together with biomechanical considerations the position of internal structures of the patient, such as an organ, and optionally a tumor in an organ, is inferred from the measured patient surface. In case the patient breaths and thus the organ and/or tumor moves, the position may be determined, which may be advantageous during, e.g., radiation therapy, since it enables that whenever the tumor is at the right position according to the radiation therapy plan, the radiation is switched on. In a specific embodiment, a finite element model is employed.

(79) 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. Other variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word comprising does not exclude other elements or steps, and the indefinite article a or an does not exclude a plurality. A single processor or other unit may fulfill the functions of several items recited in the claims. The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measured cannot be used to advantage. A computer program may be stored/distributed on a suitable medium, such as an optical storage medium or a solid-state medium supplied together with or as part of other hardware, but may also be distributed in other forms, such as via the Internet or other wired or wireless telecommunication systems. Any reference signs in the claims should not be construed as limiting the scope.