AI-based rendered volume auto-correction for fixed and mobile x-ray imaging modalities and other imaging modalities
Abstract:
A method for correcting inaccuracies in a three-dimensional (3D) rendered volume of an object due to deviations between an actual scanner translation speed and an expected scanner translation speed, the method comprising: placing a pre-measured reference adjacent to the object which is being scanned so that the pre-measured reference and the object are in the same scan field; scanning the object and the pre-measured reference so that the object and the pre-measured reference are both incorporated in a 3D rendered volume produced through scanning; comparing the 3D rendered volume of the pre-measured reference against the 3D volume of the true pre-measured reference and generating a correction map indicative of how the rendered 3D volume of the pre-measured reference should be adjusted so as to produce a more accurate 3D rendering of the pre-measured reference; and using the correction map to adjust the rendered 3D volume of the object.
Information query
Patent Agency Ranking
0/0