Systems and methods for pulmonary perfusion analysis using dynamic radiography
Described herein are systems, methods, and computer-readable medium for detecting a perfusion abnormality of a subject. In one embodiment, a method includes the following: obtaining, by dynamic radiography, imaging data for a dynamic series of a plurality of x-ray images that include areas of a subject corresponding to pulmonary vasculature; identifying, based on the imaging data, a dynamic signal corresponding to changing blood volume during the cardiac cycle of the subject; decomposing the dynamic signal into periodic components in frequency space; identifying, from the periodic components in frequency space, signals oscillating at the heart rate of the subject; generating, based on the identified signals oscillating at the heart rate of the subject, a perfusion map representation corresponding to pulmonary tissue perfusion in the subject; and detecting, based at least in part on the generated perfusion map representation, a perfusion abnormality of the subject.
1 . A method for detecting a perfusion abnormality of a subject, comprising:
obtaining, by dynamic radiography, imaging data for a dynamic series of a plurality of x-ray images that include areas of a subject corresponding to pulmonary vasculature;
identifying, based on the imaging data, a dynamic signal corresponding to changing blood volume during the cardiac cycle of the subject;
decomposing the dynamic signal into periodic components in frequency space;
identifying, from the periodic components in frequency space, signals oscillating at the heart rate of the subject;
generating, based on the identified signals oscillating at the heart rate of the subject, a perfusion map representation corresponding to pulmonary tissue perfusion in the subject; and
detecting, based at least in part on the generated perfusion map representation, a perfusion abnormality of the subject.
2 . The method of claim 1 , wherein the dynamic series of images is obtained, at least in part, using fluoroscopy and does not require the use of a contrast agent in the subject.
3 . The method of claim 1 , wherein decomposing the dynamic signal into periodic components comprises using a Fourier transform.
4 . The method of claim 1 , wherein generating the perfusion map representation comprises generating a colormap that represents an amount of signal oscillating at the heart rate of the subject, based on filtering signals in frequency space.
5 . The method of claim 1 , wherein generating the perfusion map representation comprises identifying and combining multiple frequency bands.
6 . The method of claim 1 , wherein detecting the perfusion abnormality comprises determining, from a product of the decomposed dynamic signal, pulmonary resistance associated with a time delay between cardiac contraction and blood flow to the lungs of the subject.
7 . The method of claim 1 , wherein obtaining the x-ray imaging data further comprises applying a dual energy technique to obtain two datasets associated with the area of the subject that corresponds to the pulmonary vasculature, for two different respective x-ray energies, and reconstructing images at a desired energy level such as to enhance contrast of a blood pool for optimizing signal measurements.
8 . A method for detecting a perfusion abnormality of a subject, comprising:
obtaining, by dynamic radiography using a CT scanner, imaging data of three-dimensional dynamic x-ray images that include areas of the subject which correspond to pulmonary vasculature, and wherein the imaging data includes a plurality of voxels;
identifying, from the obtained imaging data, a dynamic signal corresponding to each of the plurality of voxels;
decomposing, on a voxel-by-voxel basis, the respective dynamic signal into periodic components that correspond to periodic signals in frequency space;
identifying, from the periodic components, the magnitude of signal oscillating at the heart rate;
generating, based on the identified magnitude of signal, a three-dimensional perfusion map representation; and
detecting, based at least in part on the generated perfusion map representation, a perfusion abnormality of the subject.
9 . A system for detecting a perfusion abnormality of a subject, comprising:
a dynamic radiography device configured to obtain imaging data for a dynamic series of a plurality of x-ray images that include areas of a subject corresponding to pulmonary vasculature;
one or more processors; and
a memory coupled to at least the one or more processors that stores instructions which, when executed by a computer, cause the system to perform functions that include:
identifying, based on the imaging data, a dynamic signal corresponding to changing blood volume during the cardiac cycle of the subject;
decomposing the dynamic signal into periodic components in frequency space;
identifying, from the periodic components in frequency space, signals oscillating at the heart rate of the subject;
generating, based on the identified signals oscillating at the heart rate of the subject, a perfusion map representation corresponding to pulmonary tissue perfusion in the subject; and
detecting, based at least in part on the generated perfusion map representation, a perfusion abnormality of the subject.
10 . A system for detecting a perfusion abnormality of a subject, comprising:
a computerized tomography (CT) scanner configured to obtain imaging data of three-dimensional dynamic x-ray images that include areas of the subject which correspond to pulmonary vasculature, and wherein the imaging data includes a plurality of voxels;
one or more processors; and
a memory coupled to at least the one or more processors that stores instructions which, when executed by a computer, cause the system to perform functions that include:
identifying, from the obtained imaging data, a dynamic signal corresponding to each of the plurality of voxels;
decomposing, on a voxel-by-voxel basis, the respective dynamic signal into periodic components that correspond to periodic signals in frequency space;
identifying, from the periodic components, the magnitude of signal oscillating at the heart rate;
generating, based on the identified magnitude of signal, a three-dimensional perfusion map representation; and
detecting, based at least in part on the generated perfusion map representation, a perfusion abnormality of the subject.
11 . A non-transitory computer-readable medium storing instructions which, when executed by a computer, cause a dynamic radiography system to perform functions that include:
obtaining imaging data for a dynamic series of a plurality of x-ray images that include areas of a subject corresponding to pulmonary vasculature;
identifying, based on the imaging data, a dynamic signal corresponding to changing blood volume during the cardiac cycle of the subject;
decomposing the dynamic signal into periodic components in frequency space;
identifying, from the periodic components in frequency space, signals oscillating at the heart rate of the subject;
generating, based on the identified signals oscillating at the heart rate of the subject, a perfusion map representation corresponding to pulmonary tissue perfusion in the subject; and
detecting, based at least in part on the generated perfusion map representation, a perfusion abnormality of the subject.
12 . A non-transitory computer-readable medium storing instructions which, when executed by a computer, cause a dynamic radiography system comprising a computerized tomography (CT) scanner to perform functions that include:
obtaining imaging data of three-dimensional dynamic x-ray images that include areas of the subject which correspond to pulmonary vasculature, and wherein the imaging data includes a plurality of voxels;
identifying, from the obtained imaging data, a dynamic signal corresponding to each of the plurality of voxels;
decomposing, on a voxel-by-voxel basis, the respective dynamic signal into periodic components that correspond to periodic signals in frequency space;
identifying, from the periodic components, the magnitude of signal oscillating at the heart rate;
generating, based on the identified magnitude of signal, a three-dimensional perfusion map representation; and
detecting, based at least in part on the generated perfusion map representation, a perfusion abnormality of the subject.