Medical image processing method, apparatus, and system

1 . A medical image processing method, characterized by comprising:

acquiring a first scout image obtained after a scout scan is performed on a subject to be examined;

determining a correspondence between a scout image and a section image on the basis of a deep learning algorithm or a machine learning algorithm;

acquiring training data, the training data comprising training input data and training output data, the training input data comprising a scout image obtained by performing a scout scan in advance on a subject under examination, and the training output data comprising a section image obtained by performing an axial scan or a spiral scan on a region of interest of the subject under examination or an index identifier of the section image;

using the training data to train a neural network model so as to obtain the correspondence;

determining, according to a preset correspondence between a scout image and a section image, a predicted section image corresponding to a first region of interest in the first scout image; and

displaying the predicted section image and scanning parameters corresponding to the predicted section image.

2 . The method according to claim 1 , further comprising:

updating the predicted section image according to adjusted scanning parameters and/or according to an adjusted first region of interest, and displaying in real time the updated predicted section image and scanning parameters corresponding to the updated predicted section image.

3 . The method according to claim 2 , further comprising:

determining scanning parameters for clinical use according to the updated predicted section image and the scanning parameters corresponding to the updated predicted section image, and using the scanning parameters for clinical use to perform an axial scan or a spiral scan on the subject to be examined so as to obtain a diagnostic section image.

4 . The method according to claim 1 , wherein the first scout image comprises a normal scout image or a lateral scout image.

5 . The method according to claim 1 , wherein the scanning parameters comprise at least one of a noise index, a scan tube current and a scanning voltage.

6 . The method according to claim 1 , wherein the step of determining, according to the preset correspondence between the scout image and the section image, a predicted section image corresponding to the first region of interest in the first scout image comprises:

inputting an image of the first region of interest of the first scout image into the trained neural network model to obtain the predicted section image.

7 . The method according to claim 1 , wherein subject-under-examination features and/or regions of interest which correspond to different training input data are different; and subject-under-examination features and/or regions of interest and/or scanning parameters which correspond to different training output data are different.

8 . The method according to claim 1 , wherein the step of determining, according to the preset correspondence between the scout image and the section image, a predicted section image corresponding to the first region of interest in the first scout image comprises:

finding, in the correspondence, candidate scout images corresponding to subjects under examination matching a subject-under-examination feature of the subject to be examined;

comparing the first scout image to the candidate scout images, and selecting a second scout image matching the first scout image from among the candidate scout images; and

using a section image in the correspondence that corresponds to a first region of interest of the second scout image as the predicted section image.

9 . The method according to claim 8 , wherein the subject-under-examination features comprise at least one of: body size, sex and age.

10 . A medical image processing apparatus, characterized by comprising:

an acquisition unit, which acquires a first scout image obtained after a scout scan is performed on a subject to be examined;

a first determination unit, which determines, according to a preset correspondence between a scout image and a section image, a predicted section image corresponding to a first region of interest in the first scout image; and

a display unit, which displays the predicted section image and scanning parameters corresponding to the predicted section image

a third determination unit, which determines the correspondence on the basis of a deep learning algorithm or a machine learning algorithm

an acquisition module, which acquires training data, the training data comprising training input data and training output data, the training input data comprising a scout image obtained by performing a scout scan in advance on a subject under examination, and the training output data comprising a section image obtained by performing an axial scan or a spiral scan on a region of interest of the subject under examination or an index identifier of the section image; and

a training module, which uses the training data to train a neural network model so as to obtain the correspondence.

11 . The apparatus according to claim 10 , further comprising:

an update unit, which updates the predicted section image according to adjusted scanning parameters and/or according to an adjusted first region of interest,

wherein the display unit displays in real time the updated predicted section image and scanning parameters corresponding to the updated predicted section image.

12 . The apparatus according to claim 11 , further comprising:

a second determination unit, which determines scanning parameters for clinical use according to the updated predicted section image and the scanning parameters corresponding to the updated predicted section image, and uses the scanning parameters for clinical use to perform an axial scan or a spiral scan on the subject to be examined so as to obtain a diagnostic section image.

13 . The apparatus according to claim 12 , wherein the first determination unit inputs an image of the first region of interest of the first scout image into the trained neural network model to obtain the predicted section image.

14 . The apparatus according to claim 10 , wherein the first determination unit comprises:

a lookup module, which finds, in the correspondence, candidate scout images corresponding to subjects under examination matching a subject-under-examination feature of the subject to be examined;

a selection module, which compares the first scout image to the candidate scout images, and selects a second scout image matching the first scout image from among the candidate scout images; and

a determination module, which uses a section image in the correspondence that corresponds to a first region of interest of the second scout image as the predicted section image.

15 . A medical image processing system, characterized by comprising:

a scan device, configured to perform a scout scan on a subject to be examined so as to obtain a first scout image;

a processor, which determines, according to a preset correspondence between a scout image and a section image, a predicted section image corresponding to a first region of interest in the first scout image, wherein the processor is configured to determine, on the basis of a deep learning algorithm or a machine learning algorithm, a correspondence between a scout image and a section image by acquiring training data comprising training input data and training output data, the training input data comprising a scout image obtained by performing a scout scan in advance on a subject under examination, and the training output data comprising a section image obtained by performing an axial scan or a spiral scan on a region of interest of the subject under examination or an index identifier of the section image, and using the training data to train a neural network model so as to obtain the correspondence, and further configured to determine, according to the correspondence, a predicted section image corresponding to a first region of interest in the first scout image; and

a display, which displays the predicted section image and scanning parameters corresponding to the predicted section image.