Systems and methods for image generation

1 . A method for image generation implemented on a computing device having at least one processor and at least one storage device, the method comprising:

obtaining a first medical image and a second medical image of a target subject, the first medical image being generated based on first scan data collected using a first imaging modality, the second medical image being generated based on second scan data collected using a second imaging modality;

generating a pseudo-second medical image by transforming the first medical image, the pseudo-second medical image being a simulated image corresponding to the second imaging modality;

determining registration information between the first medical image and the second medical image based on the second medical image and the pseudo-second medical image;

generating a target medical image of the target subject based on the first medical image, the second medical image, and the registration information;

segmenting a region of interest (ROI) from the target medical image or the second medical image;

obtaining a determination result by determining whether the ROI needs to be corrected using a judgment model, wherein the judgment model is a trained machine learning model and includes a first sub-model and a second sub-model, the first sub-model being configured to determine whether the ROI includes lesion, and the second sub-model being configured to determine whether the ROI includes image artifact; and

determining whether the target medical image needs to be corrected based on the determination result, wherein the determining whether the ROI needs to be corrected using the judgment model includes:

determining whether the ROI needs to be corrected by inputting the ROI into the first sub-model and the second sub-model successively; or

determining whether the ROI needs to be corrected by inputting the ROI into the second sub-model and the first sub-model successively; or

determining whether the ROI needs to be corrected by inputting the ROI into the first sub-model and the second sub-model simultaneously.

2 . The method of claim 1 , further comprising:

generating a warped second medical image by warping the second medical image based on the registration information; and

generating a fusion image by fusing the first medical image and the warped second medical image.

3 . The method of claim 1 , wherein the generating a pseudo-second medical image by transforming the first medical image includes:

generating the pseudo-second medical image by inputting the first medical image into a modality transformation model, wherein

the modality transformation model is generated by training an initial model using a plurality of training samples, each of the plurality of training samples including a sample first medical image of a sample subject acquired using the first imaging modality and a sample second medical image of the sample subject acquired using the second imaging modality.

4 . The method of claim 1 , wherein the method further includes:

in response to determining that the target medical image needs to be corrected, correcting the target medical image.

5 . The method of claim 1 , wherein the first medical image of the target subject is generated based on the first scan data using a first reconstruction algorithm, and the method further includes:

generating a warped second medical image by warping the second medical image based on the registration information;

storing the warped second medical image into a storage device; and

retrieving the wrapped second medical image from the storage device for reconstructing a third medical image of the target subject corresponding to the first imaging modality based on the first scan data using a second reconstruction algorithm, wherein the first reconstruction algorithm is different from the second reconstruction algorithm.

6 . The method of claim 5 , wherein the method further includes:

in response to detecting that the second medical image and the first medical image are generated, triggering the determination of the registration information and the generation of the warped second medical image.

7 . The method of claim 5 , wherein the method further includes:

in response to detecting that the warped second medical image is generated, triggering the reconstruction of the third medical image.

8 . A method for image generation implemented on a computing device having at least one processor and at least one storage device, the method comprising:

obtaining a target medical image of a target subject;

segmenting a region of interest (ROI) from the target medical image;

obtaining a determination result by determining whether the ROI needs to be corrected using a judgment model, the judgment model being a trained machine learning model and including a first sub-model and a second sub-model, the first sub-model being configured to determine whether the ROI includes lesion, and the second sub-model being configured to determine whether the ROI includes image artifact;

determining whether the target medical image needs to be corrected based on the determination result; and

in response to determining that the target medical image needs to be corrected, correcting the target medical image, wherein the determining whether the ROI needs to be corrected using a judgment model includes:

determining whether the ROI needs to be corrected by inputting the ROI into the first sub-model and the second sub-model successively; or

determining whether the ROI needs to be corrected by inputting the ROI into the second sub-model and the first sub-model successively; or

determining whether the ROI needs to be corrected by inputting the ROI into the first sub-model and the second sub-model simultaneously.

9 . The method of claim 8 , wherein the obtaining a target medical image of a target subject includes:

obtaining a first medical image and a second medical image of the target subject, the first medical image being generated based on first scan data collected using a first imaging modality, the second medical image being generated based on second scan data collected using a second imaging modality;

generating a pseudo-second medical image by transforming the first medical image, the pseudo-second medical image being a simulated image corresponding to the second imaging modality;

determining registration information between the first medical image and the second medical image based on the second medical image and the pseudo-second medical image; and

generating the target medical image of the target subject based on the first medical image, the second medical image, and the registration information.

10 . The method of claim 9 , wherein the first medical image of the target subject is generated based on the first scan data using a first reconstruction algorithm, and the method further includes:

generating a warped second medical image by warping the second medical image based on the registration information;

storing the warped second medical image into a storage device; and

retrieving the wrapped second medical image from the storage device for reconstructing a third medical image of the target subject corresponding to the first imaging modality based on the first scan data using a second reconstruction algorithm.

11 . A method for image generation implemented on a computing device having at least one processor and at least one storage device, the method comprising:

obtaining a first medical image of a target subject, wherein the first medical image is generated based on first scan data using a first reconstruction algorithm, the first scan data being collected using a first imaging modality;

generating a warped second medical image by warping a second medical image based on registration information between the first medical image and the second medical image, the second medical image being generated based on second scan data collected using a second imaging modality; and

reconstructing one or more third medical images of the target subject corresponding to the first imaging modality based on the first scan data and the warped second medical image, wherein

each third medical image is reconstructed using a second reconstruction algorithm different from the first reconstruction algorithm, and

in response to detecting that the warped second medical image is generated, the reconstruction of the one or more third medical images is triggered.

12 . The method of claim 11 , wherein the reconstructing one or more third medical images of the target subject corresponding to the first imaging modality based on the first scan data and the warped second medical image includes:

storing the warped second medical image into a storage device; and

retrieving the wrapped second medical image from the storage device for reconstructing the one or more third medical images of the target subject corresponding to the first imaging modality based on the first scan data and the warped second medical image.

13 . The method of claim 11 , wherein the registration information is determined by:

generating a pseudo-second medical image by transforming the first medical image, the pseudo-second medical image being a simulated image corresponding to the second imaging modality; and

determining the registration information between the first medical image and the second medical image based on the second medical image and the pseudo-second medical image.

14 . The method of claim 11 , wherein the method further includes:

generating a target medical image of the target subject based on the first medical image, the second medical image, and the registration information;

segmenting a region of interest (ROI) from the target medical image;

obtaining a determination result by determining whether the ROI needs to be corrected using a judgment model, the judgment model being a trained machine learning model and including a first sub-model and a second sub-model, the first sub-model being configured to determine whether the ROI includes lesion, and the second sub-model being configured to determine whether the ROI includes image artifact;

determining whether the target medical image needs to be corrected based on the determination result.

15 . The method of claim 11 , wherein the first scan data is positron emission tomography (PET) scan data, the first reconstruction algorithm is a filtered back projection (FBP) algorithm, and the second reconstruction algorithm is an iterative reconstruction algorithm.

16 . The method of claim 11 , wherein the one or more third medical images include multiple third medical images reconstructed using different second reconstruction algorithms.

17 . The method of claim 11 , wherein the first scan data is collected in a first scan of the first imaging modality, and the method further comprises:

obtaining third scan data of the target subject, the third scan data being collected in a second scan of the first imaging modality performed after the first scan;

retrieving the wrapped second medical image from the storage device; and

reconstructing one or more fourth medical images of the target subject corresponding to the first imaging modality based on the third scan data and the warped second medical image.

18 . The method of claim 11 , wherein the method further includes:

in response to detecting that the second medical image and the first medical image are generated, triggering the determination of the registration information and the generation of the warped second medical image before the one or more third medical images are reconstructed.