Controlling an alert signal for spectral computed tomography imaging
AssigneeKoninklijke Philips N.V.
InventorLiran Goshen
Life of this patent
Estimating the term…
Abstract
A mechanism for identifying potential areas of clinical relevance that are not readily visible in non-spectral CT image data. The non-spectral CT image data is processed using a machine-learning method to produce predicted spectral CT image data. The predicted spectral CT image data is compared to (actual) spectral CT image data to identify any differences between the two. Any identified differences are used to control an alert signal, for instance, to draw a clinician or other user's attention to possible areas of clinical relevance that are not readily visible in the non-spectral CT image data.
Claims15
A computer-implemented method of controlling an alert signal in computed tomography (CT) imaging comprising:
- obtaining non-spectral CT image data and spectral CT image data, each representing a same region of interest of a subject;
- processing the non-spectral CT image data, using a machine-learning algorithm, to produce predicted spectral CT image data, the predicted spectral CT image data being a prediction of the spectral CT image data derived from the non-spectral CT image data;
- processing the spectral CT image data and the predicted spectral CT image data to identify any differences between the spectral CT image data and the predicted spectral CT image data;
- and
- controlling an alert signal responsive to any identified differences.
A system for controlling an alert signal in computed tomography (CT) image processing, the system comprising:
- a memory that stores a plurality of instructions;
- and
- a processor assembly coupled to the memory and configured to execute the plurality of instructions to:
- obtain non-spectral CT image data and spectral CT image data, each representing a same region of interest of a subject;
- process the non-spectral CT image data, using a machine-learning algorithm, to produce predicted spectral CT image data, the predicted spectral CT image data being a prediction of the spectral CT image data derived from the non-spectral CT image data;
- process the spectral CT image data and the predicted spectral CT image data to identify any differences between the spectral CT image data and the predicted spectral CT image data;
- and
- control an alert signal responsive to any identified differences.
A non-transitory computer-readable medium for storing executable instructions, which cause a method to be performed to control an alert signal in computed tomography (CT) imaging, the method comprising:
- obtaining non-spectral CT image data and spectral CT image data, each representing a same region of interest of a subject;
- processing the non-spectral CT image data, using a machine-learning algorithm, to produce predicted spectral CT image data, the predicted spectral CT image data being a prediction of the spectral CT image data derived from the non-spectral CT image data;
- processing the spectral CT image data and the predicted spectral CT image data to identify any differences between the spectral CT image data and the predicted spectral CT image data;
- and
- controlling an alert signal responsive to any identified differences.
Dependencies are inferred from the claim text.