Method for use in x-ray CT image reconstruction
A method for use in image reconstruction of CT projection data, which aims at reducing motion artefacts in reconstructed images caused by movement of anatomical bodies. Embodiments are based on mitigating motion artefacts based on restricting the range of data that is used for reconstructing each slice. More particularly, a sub-range of the projection data corresponding to each slice is selected, this sub-range being chosen based on determining one or more sub-windows of visibility of a target anatomical object or event within the projection data sequence. The event may be a particular phase of a movement cycle of the anatomical body. The structure could be a particular portion of the anatomical body which is of interest. Either approach leads to reduction of motion artefacts within a single slice, by restricting the data range, and focusing upon the data which is most relevant clinically.
1 . A medical computed tomography (CT) device, comprising:
a memory that stores a plurality of instructions; and
a processor that couples to the memory and is configured to execute the plurality of instructions to:
obtain an input CT projection data set comprising a plurality of data subsets, each subset corresponding to a different respective slice through a scanned body, and wherein each data subset comprises data corresponding to a defined range of acquisition angles, and a defined acquisition time range;
identify at least one temporal or angular sub-window within the input projection data set associated with visibility in said data set, if reconstructed, of a desired anatomical feature or event;
select from the data subset for each slice a temporal or angular sub-range of the data, smaller than a complete range for each slice, wherein selection of the sub-range is based on at least one of the identified at least one temporal or angular sub-window in the projection data;
generate a data output indicative of the selected sub-range for each slice;
perform a first image reconstruction of each of the slices using a complete data range of the data subset for each slice to derive a first image for each slice;
apply low-pass filtering to each of the first images;
perform a second image reconstruction for each of the slices using the identified sub-range of the data subset for each slice to derive a second image for each slice;
apply high pass filtering to each of the second images;
generate a combined image for each slice comprising a combination of the filtered first and second images, wherein generating the combined image for each slice comprises adding a low-frequency part of the first image reconstructed from the complete data range of the data subset to a high-frequency part of the second image reconstructed from the sub-range of the data subset; and
generate an output indicative of the combined image for each slice.
2 . The device as claimed in claim 1 , wherein the processor is configured to:
receive a signal generated by an external sensor configured to sense one or more body parameters related to an anatomical event, the signal corresponding to an output of the sensor over the course of the acquisition of the input projection data, and
determine the at least one sub-window associated with visibility of the anatomical object or event based at least in part on the received signal.
3 . The device as claimed in claim 1 , wherein the determining the at least one sub-window in the input projection data set comprises performing a preliminary image reconstruction for the at least one data subset and determining, based on the preliminary reconstruction, the at least one sub-window within the input projection data set within which the anatomical feature or event is visible.
4 . The device as claimed in claim 2 , wherein the selected angular sub-range for each data subset has a pre-defined angular width, and the selecting the data sub-range for a given data subset comprises determining an alignment of the sub-range within the full width of the data subset.
5 . The device as claimed in claim 4 , wherein determining the alignment of the data sub-range is performed based on the signal generated by the external sensor.
6 . The device as claimed in claim 1 , wherein selecting the data sub-range for at least one data subset comprises determining an angular width of the data sub-range based on one of the at least one temporal or angular sub-window.
7 . The device as claimed in claim 1 , wherein the projection data is representative of an anatomical object of interest, wherein the object of interest undergoes cyclical movement, and wherein the anatomical event corresponds to a phase of the cyclical movement.
8 . The device as claimed in claim 1 , wherein selecting the data sub-range for each slice is performed based in part on the data sub-ranges selected for at least a portion of the other slices.
9 . The device as claimed in claim 1 , wherein selecting the data sub-range for each data subset is based in part on maximizing temporal overlap or temporal proximity of the angular sub-ranges selected for the different data subsets, and is based in part on maximizing temporal overlap or temporal proximity of the angular sub-range selected for the data subsets of neighboring slices.
10 . The device as claimed in claim 1 , wherein the processor is configured to extract the data sub-range selected for each respective slice from the data subset for the slice.
11 . The device as claimed in claim 10 , wherein the processor is configured to perform image reconstruction for each slice using only the extracted data sub-range of the respective data subset for each slice.
12 . A computer implemented method for use in medical computed tomography (CT) image reconstruction, the method comprising:
Receiving an input CT projection data set comprising a plurality of data subsets, each subset corresponding to a different respective slice through a scanned body, and wherein each data subset comprises data corresponding to a defined range of acquisition angles, and a defined acquisition time range;
identifying at least one temporal or angular sub-window within the input projection data set associated with visibility in said data set, if reconstructed, of a desired anatomical feature or event;
selecting from the data subset for each slice a temporal or angular sub-range of the data, smaller than a complete range for each slice, wherein selection of the sub-range is based on at least one of the identified at least one temporal or angular sub-window in the projection data;
generating a data output indicative of the selected sub-range for each slice;
performing a first image reconstruction of each of the slices using a complete data range of the data subset for each slice to derive a first image for each slice;
applying low-pass filtering to each of the first images;
performing a second image reconstruction for each of the slices using the identified sub-range of the data subset for each slice to derive a second image for each slice;
applying high pass filtering to each of the second images;
generating a combined image for each slice comprising a combination of the filtered first and second images, wherein generating the combined image for each slice comprises adding a low-frequency part of the first image reconstructed from the complete data range of the data subset to a high-frequency part of the second image reconstructed from the sub-range of the data subset; and
generating an output indicative of the combined image for each slice.
13 . A non-transitory computer-readable medium for storing executable instructions, which cause a method to be performed for use in medical computed tomography (CT) image reconstruction, the method comprising:
Receiving an input CT projection data set comprising a plurality of data subsets, each subset corresponding to a different respective slice through a scanned body, and wherein each data subset comprises data corresponding to a defined range of acquisition angles, and a defined acquisition time range;
identifying at least one temporal or angular sub-window within the input projection data set associated with visibility in said data set, if reconstructed, of a desired anatomical feature or event;
selecting from the data subset for each slice a temporal or angular sub-range of the data, smaller than a complete range for each slice, wherein selection of the sub-range is based on at least one of the identified at least one temporal or angular sub-window in the projection data;
generating a data output indicative of the selected sub-range for each slice;
performing a first image reconstruction of each of the slices using a complete data range of the data subset for each slice to derive a first image for each slice;
applying low-pass filtering to each of the first images;
performing a second image reconstruction for each of the slices using the identified sub-range of the data subset for each slice to derive a second image for each slice;
applying high pass filtering to each of the second images;
generating a combined image for each slice comprising a combination of the filtered first and second images, wherein generating the combined image for each slice comprises adding a low-frequency part of the first image reconstructed from the complete data range of the data subset to a high-frequency part of the second image reconstructed from the sub-range of the data subset; and
generating an output indicative of the combined image for each slice.