Systems and methods for facilitating opportunistic screening for cardiomegaly
A computer-implemented method for facilitating opportunistic screening for cardiomegaly includes obtaining a set of computed tomography (CT) images. The set of CT images captures at least a portion of a heart of a patient, and the set of CT images is captured for a purpose independent of assessing cardiomegaly. The method further includes using the set of CT images as an input to an artificial intelligence (AI) module configured to determine a heart measurement based on CT image set input. The method also includes obtaining heart measurement output generated by the AI module and, based on the heart measurement output, classifying the patient into one of a plurality of risk levels for cardiomegaly. The classification is operable to trigger additional action based on the corresponding risk level for the patient.
1. A computer-implemented method for facilitating opportunistic screening for cardiomegaly, comprising:
obtaining a set of computed tomography (CT) images, the set of CT images capturing at least a portion of a heart of a patient, the set of CT images being captured for a purpose independent of assessing cardiomegaly, the set of CT images comprising CT images captured in a non-gated manner;
using the set of CT images as an input to an artificial intelligence (AI) module configured to determine a heart measurement based on CT image set input by:
identifying a subset of CT images from the CT image set input, the subset of CT images providing one or more largest measurements associated with a heart represented in the CT image set input, and
providing heart measurement output based on the one or more largest measurements associated with the heart represented in the CT image set input, wherein the AI module is trained using training data comprising (i) a plurality of training sets of CT images and (ii) for each training set of CT images of the plurality of training sets of CT images, an identification of a respective subset of CT images and a respective heart measurement based on the respective subset of CT images;
obtaining heart measurement output based on output of the AI module; and
based on the heart measurement output, classifying the patient into one of a plurality of risk levels for cardiomegaly, the classification being operable to trigger additional action based on the corresponding risk level for cardiomegaly.
2. The computer-implemented method of claim 1 , wherein the set of CT images comprises a set of abdominal CT images.
3. The computer-implemented method of claim 1 , wherein the set of CT images comprises a set of chest CT images.
4. The computer-implemented method of claim 1 , wherein the AI module comprises a machine learning module.
5. The computer-implemented method of claim 4 , wherein the one or more largest measurements associated with the heart represented in the CT image set input comprises a ventricular wall thickness and/or an axial length and/or area of one or more of: the heart represented in the CT image set input, a right ventricle thereof, a left ventricle thereof, a right atrium thereof, and/or a left atrium thereof.
6. The computer-implemented method of claim 1 , wherein the AI module comprises a machine learning module configured to:
identify a subset of CT images from the CT image set input, each CT image of the subset of CT images providing a representation of a heart represented in the CT image set input; and
provide heart measurement output comprising a volume measure of the heart represented in the CT image set input, the volume measure being determined using the subset of CT images.
7. The computer-implemented method of claim 4 , wherein:
the machine learning module is further configured to determine whether the CT image set input comprises a set of abdomen CT images or a set of chest CT images, and
the heart measurement output is at least partially based on whether the CT image set input comprises a set of abdomen CT images or a set of chest CT images.
8. The computer-implemented method of claim 1 , wherein classifying the patient into one of the plurality of risk levels for cardiomegaly is further based on one or more patient attributes of the patient.
9. The computer-implemented method of claim 8 , wherein the one or more patient attributes comprise one or more of: body mass index (BMI), body surface area, inner chest size, outer chest size, height, weight, girth, age, gender, race, and/or field of view of the axial CT image.
10. The computer-implemented method of claim 1 , wherein, for at least one of the plurality of risk levels for cardiomegaly, the additional action comprises providing a notification to one or more relevant entities.
11. The computer-implemented method of claim 10 , wherein the one or more relevant entities comprises the patient or a physician for the patient.
12. The computer-implemented method of claim 10 , wherein providing the notification to the one or more relevant entities comprises generating a report and providing the report to the one or more relevant entities.
13. The computer-implemented method of claim 12 , wherein the report comprises the heart measurement output and/or one or more subsets of CT images of the set of CT images.
14. The computer-implemented method of claim 13 , wherein the report comprises a percentile ranking associated with the heart measurement output relative to heart measurements of other patients.
15. The computer-implemented method of claim 1 , wherein the heart measurement output comprises a major axis of at least part of the heart of the patient.
16. A system for facilitating opportunistic screening for cardiomegaly, the system comprising:
one or more processors; and
one or more hardware storage devices that store instructions that are executable by the one or more processors to configure the system to:
access a set of computed tomography (CT) images, the set of CT images capturing at least a portion of a heart of a patient, the set of CT images being captured for a purpose independent of assessing cardiomegaly, the set of CT images comprising CT images captured in a non-gated manner;
process the set of CT images using an artificial intelligence (AI) module configured to determine a heart measurement based on CT image set input by causing the AI module to:
identify a subset of CT images from the CT image set input, the subset of CT images providing one or more largest measurements associated with a heart represented in the CT image set input, and
provide heart measurement output based on the one or more largest measurements associated with the heart represented in the CT image set input, wherein the AI module is trained using training data comprising (i) a plurality of training sets of CT images and (ii) for each training set of CT images of the plurality of training sets of CT images, an identification of a respective subset of CT images and a respective heart measurement based on the respective subset of CT images; and
based on the heart measurement output, classify the patient into one of a plurality of risk levels for cardiomegaly, the classification being operable to trigger additional action based on the corresponding risk level for cardiomegaly.