| 摘要 |
Disclosed is a method for interpreting image data that represents variations in density of matter inside a subject. A data processing system may receive image data from a scan of a subject. The data processing system automatically discerns one or more bones, one or more arteries, and one or more organs of the subject, based at least in part on the image data. The data processing system may also automatically identify one or more of the discerned bones, arteries, and organs. The data processing system may also automatically discern an outer or inner margin of a body wall of the subject. The data processing system may also produce output that depicts one or more of the identified bones, one or more of the identified arteries, one or more of the identified organs, and/or or the outer or inner margin of the body wall. Other embodiments are described and claimed. |
| 主权项 |
1. A method comprising:
receiving image data from a scan of a subject, the image data representing first voxels and second voxels; automatically discerning one or more bones of the subject, based at least in part on the image data; automatically identifying one or more of the discerned bones; automatically discerning one or more arteries of the subject, based at least in part on the image data; automatically identifying one or more of the discerned arteries; automatically discerning one or more organs of the subject, based at least in part on the image data; automatically identifying one or more of the discerned organs; automatically discerning an outer or inner margin of a body wall of the subject; and producing output that depicts one or more of the identified bones, one or more of the identified arteries, one or more of the identified organs, or the outer or inner margin of the body wall, wherein the outer or inner margin of the body wall of the subject is automatically discerned by automatic image recognition software executed by a processor, or at least one of the bones, arteries and organs of the subject is automatically discerned or automatically identified by the automatic image recognition software executed by the processor, wherein the automatically discerning the one or more bones, arteries and organs of the subject, the automatically identifying the one or more discerned bones, arteries and organs, and the automatically determining a branch point of one of the one or more arteries of the subject comprise:
for each of at least some of the first voxels, comparing the respective first voxel with a threshold value of radiodensity and with a radiodensity value of a respective one of the second voxels to produce an array, wherein the respective one of the second voxels is a neighboring voxel of the respective first voxel, andusing the array, converting the image data to an output image wherein the identified one or more discerned bones, arteries and organs, and the determined branch point of one of the one or more arteries of the subject are identified,wherein the array is produced by operations comprising:
for each of the compared first voxels, based at least in part on the comparison, assigning a respective identifier to the respective compared first voxel, the respective identifier indicating a respective particular one of the identified bones, arteries, or organs, or the determined branch point; andfor each of the compared first voxels, storing in the array a radiodensity value of the respective compared first voxel, a radiodensity gradient value of the respective compared first voxel, and the respective identifier assigned to the respective compared first voxel, wherein the radiodensity gradient value of the respective compared first voxel indicates a value of radiodensity of the respective compared first voxel relative to radiodensity values of voxels neighboring the respective compared first voxel. |
