Automatic computation of BPE level

1 . A method, comprising:

training a deep learning (DL) model with a plurality of sets of training data from training subjects, each set of training data of the plurality of sets of training data including two input images of a breast of a respective one of the training subjects, ground truth versions of the two input images with background parenchymal enhancement (BPE) segmented, and a ground truth BPE score for both of the two input images, the DL model further trained with clinical information including contrast injection timing and acquisition timing for each set of training data, the two input images comprising a training low energy image and a training recombination image created from the training low energy image and a training high energy image;

obtaining a low energy image and a high energy image of a breast of a patient acquired with an x-ray imaging system configured to perform dual-energy imaging;

entering the low energy image and a recombination image generated from the low and high energy images of the breast of the patient acquired via the x-ray imaging system as input to the DL model;

receiving, as output from the DL model, a BPE assessment based on the low energy image and the recombination image of the breast of the patient; and

displaying the one or more images and the BPE assessment at a display device.

2 . The method of claim 1 , wherein an output of the DL model includes one or more images showing a segmentation of BPE within the breast, and the one or more images showing the segmentation of the BPE and the BPE assessment are displayed on the display device.

3 . The method of claim 1 , further comprising inputting additional clinical information into the DL model, and receiving the BPE assessment as an output of the DL model, the additional clinical information including at least one of:

metadata of the one or more images;

a timing of an injection of contrast into the patient;

acquisition times of the one or more images;

menstrual cycle information of the patient;

an age of the patient;

demographic data of the patient;

a presence or absence of one or more conditions of the patient; and

a definition of a specified portion of the one or more images, the specified portion to be excluded from the BPE assessment.

4 . The method of claim 1 , wherein:

the BPE assessment is a score indicating a percentage of tissues of the breast of the patient showing BPE with respect to a surface area of the breast; or wherein the BPE assessment is a score indicating a percentage of tissues of the breast of the patient showing BPE, with respect to a total volume of the breast.

5 . The method of claim 1 , wherein the BPE assessment is a classification of the breast of the patient into one of a plurality of categories.

6 . The method of claim 1 , wherein the breast is a first breast of the patient and the BPE assessment includes a first level of BPE of the first breast, and further comprising determining whether an asymmetry is detected between the first BPE level and a second BPE level of a second breast of the patient.

7 . The method of claim 6 , wherein determining whether the asymmetry is detected between the first BPE level and the second BPE level further comprises:

inputting a low energy image and a recombination image of the second breast into the DL model, and receiving a second BPE assessment of the second breast as a second output of the DL model, the second BPE assessment including the second BPE level; and

in response to a difference between the first BPE level and the second BPE level being greater than a threshold difference, displaying an indication that the asymmetry was detected on the display device.

8 . The method of claim 6 , wherein the DL model is trained to predict the asymmetry between the first BPE level and the second BPE level, and determining whether the asymmetry exists between the first BPE level and the second BPE level further comprises:

inputting the low energy image and the recombination image of the first breast and a low energy image and a recombination image of the second breast into the DL model, and receiving a predicted BPE asymmetry between the first BPE level and the second BPE level as an output of the DL model; and

displaying an indication of the predicted BPE asymmetry was detected on the display device.

9 . The method of claim 6 , further comprising determining whether the asymmetry exists between the first BPE level and the second BPE level based on a first time at which the low energy image of the first breast was acquired and a second time at which the low energy image of the second breast was acquired.

10 . The method of claim 1 , further comprising acquiring a craniocaudal view (CC) and a mediolateral oblique (MLO) view of the breast of the patient during a CEM exam, and generating the BPE assessment of the breast of the patient using the DL model based on both of the CC and MLO views.

11 . An image processing system, comprising:

a display device;

an x-ray imaging system configured to perform dual-energy imaging;

non-transitory memory storing instructions executable by one or more processors to:

train a deep learning (DL) model with a plurality of sets of training data from training subjects, each set of training data of the plurality of sets of training data including two input images of a breast of a respective one of the training subjects, ground truth versions of the two input images with background parenchymal enhancement (BPE) segmented, and a ground truth BPE score for both of the two input images, the two input images comprising a training low energy image and a training recombination image created from the low energy image and a training high energy image;

receive a first set of medical images of a first breast of a patient acquired via the x-ray imaging system, the first set of medical images of the first breast including a low energy image and a first high energy image;

input the the first low energy image and a first recombination image generated from the low energy image and the first high energy image into the DL model, the DL model trained to perform an assessment of a first level of BPE of the first breast based on the first low energy image and the first recombination image;

receive a first BPE assessment of the first breast as an output of the DL model, the first BPE assessment including at least one of a score and a classification indicating a percentage of glandular tissues of the first breast showing BPE with respect to a surface or volume of the first breast; and

display the first BPE assessment on the display device.

12 . The image processing system of claim 11 , wherein further instructions are stored in the non-transitory memory that when executed, cause the processor to input additional information into the DL model, and generate the first BPE assessment of the first breast based on the first low energy image and the first recombination image and the additional information, the additional information including at least one of:

metadata of the one or more medical images;

a timing of an injection of contrast into the patient;

acquisition times of the one or more medical images;

menstrual cycle information of the patient;

an age of the patient;

demographic data of the patient;

a presence or absence of one or more conditions of the patient; and

a definition of a specified portion of the one or more medical images, the specified portion to be excluded from the first BPE assessment.

13 . The image processing system of claim 11 , wherein further instructions are stored in the non-transitory memory that when executed, cause the processor to:

receive a second set of medical images of a second breast of the patient acquired with the x-ray imaging system, the second set including a second low energy image of the second breast and a second high energy image of the second breast;

input the second low energy image and a second recombination image generated from the second low energy image and the second high energy image into the DL model to obtain a second BPE assessment of a second level of BPE of the second breast based on the second low energy image and the second recombination image;

determine whether a BPE asymmetry exists between the first level of BPE of the first breast and the second level of BPE of the second breast; and

in response to determining that the BPE asymmetry exists, display an indication that the BPE asymmetry was detected on the display device.

14 . The image processing system of claim 13 , wherein further instructions are stored in the non-transitory memory that when executed, cause the processor to determine whether the BPE asymmetry exists based on a first timing of a first acquisition of the first set of medical images and a second timing of a second acquisition of the second set of medical images.

15 . The image processing system of claim 11 , wherein the output of the DL model includes one or more images showing a segmentation of BPE within the first breast, and the one or more images showing the segmentation of the BPE and the BPE assessment are displayed on the display device.

16 . A method for an x-ray imaging system configured to perform dual-energy imaging, the method comprising:

training a deep learning (DL) model with a plurality of sets of training data from training subjects, each set of training data of the plurality of sets of training data including two input images of a breast of a respective one of the training subjects, ground truth versions of the two input images with background parenchymal enhancement (BPE) segmented, and a ground truth BPE score for both of the two input images, the two input images comprising a training low energy image and a training recombination image created from the training low energy image and a training high energy image;

receiving a plurality of medical images of breasts of a patient acquired via the x-ray imaging system during a breast exam of the patient, the x-ray imaging system configured to obtain different views of the breasts at different times of the breast exam, the plurality of medical images including a first set of medical images of a first breast of the patient acquired via the imaging system at a first time of the breast exam and a second set of medical images of a second breast of the patient acquired via the imaging system at a second time of the breast exam, wherein the first set of medical images includes a first low energy image and a first high energy image, and wherein the second set of medical images includes a second low energy image and a second high energy image;

inputting the first low energy image, the second low energy image, a first recombination image generated from the first low energy image and the first high energy image, a second recombination image generated from the second low energy image and the second high energy image, the first time, and the second time into the DL model to obtain an assessment of a BPE asymmetry between the first breast and the second breast;

displaying at least one image of the first set of medical images of the first breast on a display device of the imaging system;

displaying at least one image of the second set of medical images of the second breast on the display device; and

displaying an indication of the BPE asymmetry on the display device.