Image processing device and image processing method
AssigneeHamamatsu Photonics K.K.
InventorsKibo Ote, Fumio Hashimoto, Yuya Onishi
Life of this patent
Estimating the term…
Abstract
An image processing method, starting from a certain initial state, creates a tomographic image of a subject by repeatedly performing a reconstruction step, a CNN processing step, and an update step a plurality of times. In the reconstruction step, a first image is created by performing processing by using a list mode iterative reconstruction method. In the CNN processing step, input information is input to a CNN and a second image is created by the CNN by using a DIP technique, and the CNN is trained. In the update step, a third image is updated based on the first image and the second image.
Claims22
An image processing apparatus for creating a tomographic image based on list data collected by a radiation tomography apparatus, the image processing apparatus comprising:
- a reconstruction unit configured to newly create a first image by repeatedly performing processing of bringing an image obtained by updating the first image based on the list data by using a list mode iterative reconstruction method close to a difference between a second image and a third image;
- a CNN processing unit configured to input input information to a convolutional neural network, create the second image by the convolutional neural network, and train the convolutional neural network so as to bring the created second image close to a sum of the first image and the third image;
- and
- an update unit configured to update the third image based on the first image and the second image, wherein
- starting from an initial state of each of a training state of the convolutional neural network, the first image, the second image, and the third image, creation of the first image by the reconstruction unit, creation of the second image and training of the convolutional neural network by the CNN processing unit, and update of the third image by the update unit are repeatedly performed, and any one of the first image and the second image obtained by repeated processing is set as the tomographic image.
An image processing apparatus for creating a tomographic image based on list data collected by a radiation tomography apparatus, the image processing apparatus comprising:
- a reconstruction unit configured to create a first image by updating a third image based on the list data by using a list mode iterative reconstruction method;
- a CNN processing unit configured to input input information to a convolutional neural network, create a second image by the convolutional neural network, and train the convolutional neural network so as to bring the created second image close to the third image;
- and
- an update unit configured to update the third image based on the first image and the second image, wherein
- starting from an initial state of each of a training state of the convolutional neural network and the third image, creation of the first image by the reconstruction unit, creation of the second image and training of the convolutional neural network by the CNN processing unit, and update of the third image by the update unit are repeatedly performed, and any one of the first image, the second image, and the third image obtained by repeated processing is set as the tomographic image.
An image processing method for creating a tomographic image based on list data collected by a radiation tomography apparatus, the image processing method comprising:
- a reconstruction step of newly creating a first image by repeatedly performing processing of bringing an image obtained by updating the first image based on the list data by using a list mode iterative reconstruction method close to a difference between a second image and a third image;
- a CNN processing step of inputting input information to a convolutional neural network, creating the second image by the convolutional neural network, and training the convolutional neural network so as to bring the created second image close to a sum of the first image and the third image;
- and
- an update step of updating the third image based on the first image and the second image, wherein
- starting from an initial state of each of a training state of the convolutional neural network, the first image, the second image, and the third image, creation of the first image in the reconstruction step, creation of the second image and training of the convolutional neural network in the CNN processing step, and update of the third image in the update step are repeatedly performed, and any one of the first image and the second image obtained by repeated processing is set as the tomographic image.
An image processing method for creating a tomographic image based on list data collected by a radiation tomography apparatus, the image processing method comprising:
- a reconstruction step of creating a first image by updating a third image based on the list data by using a list mode iterative reconstruction method;
- a CNN processing step of inputting input information to a convolutional neural network, creating a second image by the convolutional neural network, and training the convolutional neural network so as to bring the created second image close to the third image;
- and
- an update step of updating the third image based on the first image and the second image, wherein
- starting from an initial state of each of a training state of the convolutional neural network and the third image, creation of the first image in the reconstruction step, creation of the second image and training of the convolutional neural network in the CNN processing step, and update of the third image in the update step are repeatedly performed, and any one of the first image, the second image, and the third image obtained by repeated processing is set as the tomographic image.
Dependencies are inferred from the claim text.