Search

EP-4551122-B1 - SYSTEM AND METHOD FOR PROCESSING ULTRASOUND IMAGING DATA

EP4551122B1EP 4551122 B1EP4551122 B1EP 4551122B1EP-4551122-B1

Inventors

  • WILD, Sebastian
  • EWALD, Arne
  • GOOSSEN, Andre
  • LOSSAU, Tanja
  • PETERS, JOCHEN
  • GROTH, ALEXANDRA
  • WAECHTER-STEHLE, IRINA
  • WEBER, Frank Michael

Dates

Publication Date
20260513
Application Date
20230628

Claims (13)

  1. A processing system (120) for processing multiplane ultrasound imaging data, the processing system being configured to: obtain a first set of two-dimensional ultrasound images (115) of an anatomical structure acquired, at a first time frame (f 1 ), by a multiplane ultrasound scanning device (110), wherein at least two of the two-dimensional images have been acquired at a differently-oriented plane; obtain a three-dimensional ultrasound image of the anatomical structure acquired, at a third time frame (f 3 ), by a three-dimensional ultrasound scanning device, wherein the third time frame is different to the first time frame (f 1 ); and generate an estimated three-dimensional ultrasound image (125) of the anatomical structure at the first time frame by processing at least the first set of two-dimensional ultrasound images and the three-dimensional ultrasound image of the anatomical structure using a machine-learning algorithm, wherein the machine-learning algorithm has been trained using a training algorithm configured to receive training input data entries and corresponding training output data entries, wherein each training input data entry comprises a set of two-dimensional ultrasound images of an anatomical structure for a subject and each corresponding training output data entry comprises a three-dimensional ultrasound image of the same anatomical structure, and wherein each training input data entry further comprises a three-dimensional ultrasound image of the same anatomical structure as the corresponding set of two-dimensional ultrasound images, said three-dimensional ultrasound image having been acquired at a different time to the set of two-dimensional ultrasound images of said training input data entry or to the three-dimensional image of the output data entry corresponding to said training input data entry.
  2. The processing system of claim 1, wherein each two-dimensional image in the first set of two dimensional images has been acquired at a differently-oriented plane.
  3. The processing system (120) of claim 1 or 2, wherein: the training input data entries comprise a group of first training inputs and the training output data entries comprise first known outputs, each first training input corresponding to a respective first known output; and the set of two-dimensional ultrasound images of each first training input comprises a set of simulated two-dimensional multiplane ultrasound images, wherein the set of simulated two-dimensional multiplane ultrasound images is generated from the three-dimensional ultrasound image of the respective first known output.
  4. The processing system (120) of claim 1 or 2, wherein: the training input data entries comprise second training inputs and the training output data entries comprise second known outputs, each second training input corresponding to a respective second known output; the set of two-dimensional ultrasound images of each second training input comprises a set of two-dimensional multiplane ultrasound images of the anatomical structure; and the three-dimensional ultrasound image of each second known output is represented by the set of two-dimensional multiplane ultrasound images of the corresponding second training input.
  5. The processing system (120) of any of claims 1 to 4, wherein the machine-learning algorithm is a generative adversarial network.
  6. The processing system (120) of any of claims 1 to 5, wherein the processing system is further configured to provide the estimated three-dimensional ultrasound image (125) to an image analysis algorithm developed for processing and/or analyzing three-dimensional imaging data.
  7. The processing system (120) of claim 6, wherein the image analysis algorithm is a model-based segmentation algorithm.
  8. The processing system (120) of any of claims 1 to 7, wherein the processing system is further configured to provide, at a display device (130), a visual representation of the estimated three-dimensional ultrasound image (125) of the anatomical structure.
  9. The processing system (120) of any of claims 1 to 8, wherein the processing system is configured to: obtain a second set of two-dimensional ultrasound images of the anatomical structure acquired, at a second time frame (f 2 ), by the multiplane ultrasound scanning device (110), wherein, in the second set, at least two of the two-dimensional ultrasound images have been acquired at a differently-oriented plane, and wherein the second time frame is different to the first time frame (f 1 ); and generate the estimated three-dimensional ultrasound image (125) of the anatomical structure at the first time frame by processing at least the first set of two-dimensional ultrasound images (115) and the second set of two-dimensional ultrasound images using the machine-learning algorithm.
  10. An ultrasound system, comprising: a multiplane ultrasound scanning device; and the processing system of any of claims 1 to 9.
  11. A computer-implemented method (200) for processing multiplane ultrasound imaging data, the computer-implemented method comprising: obtaining a first set of two-dimensional ultrasound images (115) of an anatomical structure acquired, at a first time frame (f 1 ), by a multiplane ultrasound scanning device (110), wherein at least two of the two-dimensional ultrasound images have been acquired at a differently-oriented plane; obtaining a three-dimensional ultrasound image of the anatomical structure acquired, at a third time frame (f 3 ), by a three-dimensional ultrasound scanning device, wherein the third time frame is different to the first time frame (f 1 ); and generating an estimated three-dimensional ultrasound image (125) of the anatomical structure at the first time frame by processing at least the first set of two-dimensional ultrasound images and the three-dimensional ultrasound image of the anatomical structure using a machine-learning algorithm, wherein the machine-learning algorithm has been trained using a training algorithm configured to receive training input data entries and corresponding training output data entries, wherein each training input data entry comprises a set of two-dimensional ultrasound images of an anatomical structure for a subject and each corresponding training output data entry comprises a three-dimensional ultrasound image of the same anatomical structure, and wherein each training input data entry further comprises a three-dimensional ultrasound image of the same anatomical structure as the corresponding set of two-dimensional ultrasound images, said three-dimensional ultrasound image having been acquired at a different time to the set of two-dimensional ultrasound images of said training input data entry or to the three-dimensional image of the output data entry corresponding to said training input data entry.
  12. The computer-implemented method (200) of claim 11, wherein the computer-implemented method further comprises: obtaining a second set of two-dimensional ultrasound images of the anatomical structure acquired, at a second time frame (f 2 ), by the multiplane ultrasound scanning device (110), wherein, in the second set, at least two of the two-dimensional ultrasound images has been acquired at a differently-oriented plane, and wherein the second time frame is different to the first time frame (f 1 ); and wherein the step of generating the estimated three-dimensional ultrasound image (125) of the anatomical structure at the first time frame comprises processing at least the first set of two-dimensional ultrasound images (115) and the second set of two-dimensional ultrasound images using the machine-learning algorithm.
  13. A computer program product comprising computer program code means which, when executed on a computer device having a processing system, cause the processing system to perform all of the steps of the method (200) according to claim 11 or 12.

Description

FIELD OF THE INVENTION The invention relates to the field of ultrasound imaging. BACKGROUND OF THE INVENTION When referring to ultrasound images, these can be acquired in either two or three dimensions. Two-dimensional (2D) ultrasound imaging remains a popular choice, as it allows a higher frame rate than three-dimensional imaging. In the case of contrast-enhanced ultrasound, two-dimensional ultrasound imaging results in less destruction of an injected contrast agent (also known as an ultrasound enhancing agent, UEA) compared with three-dimensional (3D) ultrasound imaging. However, many automated image analysis algorithms are designed to operate on three-dimensional ultrasound images, and cannot be used to analyze two-dimensional ultrasound images. There is therefore a need for an improved method for generating three-dimensional information from ultrasound imaging data. WO 2021/209348 A1 discloses an ultrasound roadmap image generation system. SUMMARY OF THE INVENTION The invention is defined by the claims. SUMMARY OF THE DISCLOSURE According to examples in accordance with an aspect of the disclosure, there is provided a processing system for processing multiplane ultrasound imaging data. The processing system is configured to: obtain a first set of two-dimensional ultrasound images of an anatomical structure acquired, at a first time frame, by a multiplane ultrasound scanning device (e.g., an imaging probe), wherein at least two of the two-dimensional images have been acquired at a differently-oriented plane; obtain a three-dimensional ultrasound image of the anatomical structure acquired, at a third time frame, by a three-dimensional ultrasound scanning device, wherein the third time frame is different to the first time frame; and generate an estimated three-dimensional ultrasound image of the anatomical structure at the first time frame by processing at least the first set of two-dimensional ultrasound images and the three-dimensional ultrasound image of the anatomical structure using a machine-learning algorithm. The first set of two-dimensional ultrasound images are multiplane images, i.e. images acquired/captured by an imaging probe at a same probe position and orientation and at/in a same time frame. The multiplane images (and therefore at least two of the images in the first set of two-dimensional ultrasound images) are differently-oriented to one another. In some examples, each image in the first set of two-dimensional ultrasound images has been acquired at a different-oriented plane. Preferably, one of the imaging axes of each multiplane image is at a same location and orientation (with respect to an imaged subject/individual) as an imaging axis of at least one other multiplane image. In other words, the multiplane images may share an imaging axis. Preferably, each two-dimensional image in the first set (i.e., each multiplane image) comprises image data representing a same axis that extends from the probe that captured the mulitplane images into the subject/individual being imaged. The inventors have recognized that multiplane imaging provides some three-dimensional information, and that the remaining information in a three-dimensional volume may be constructed by the use of a machine-learning algorithm. The machine-learning algorithm is a machine-learning algorithm that has been trained to recognize three-dimensional features of an anatomical structure based on two-dimensional multiplane images. The generated estimated three-dimensional image is thus a "full" three-dimensional image, which may then be treated in the same way as a three-dimensional image obtained by three-dimensional imaging. Since two-dimensional multiplane ultrasound imaging has a higher frame rate than three-dimensional ultrasound imaging, estimated three-dimensional ultrasound images may be produced at a higher frame rate than "real" three-dimensional ultrasound images. In addition to this, where a contrast (or enhancing) agent is used, two-dimensional multiplane ultrasound imaging will result in less destruction of the contrast agent, allowing a higher number of contrast-enhanced estimated three-dimensional ultrasound images to be produced. The use of multiplane images, rather than two-dimensional images acquired sequentially by moving/rotating an imaging probe between acquisitions, means that the images are automatically spatially and temporally aligned. This avoids the introduction of errors when aligning images. The temporal alignment is particularly valuable in applications that require a three-dimensional image from a particular point in time, such as cardiac imaging. Even when using cardiac gating, images from corresponding points of successive cardiac cycles will not align exactly, as different breathing states at the same point of the cardiac cycle, as well as irregularities in the cardiac cycle, will cause different heart shapes. Preferably, the first set of two-dimensional ultrasound images comprises fewer than 10 i