Low radiation dose computed tomography perfusion (CTP) with improved quantitative analysis
AssigneeKoninklijke Philips N.V.
InventorsShiyu Xu, Hao Dang, Chuanyong Bai
Life of this patent
Estimating the term…
Abstract
A computer tomography scanner (102) includes a radiation source (112) configured to emit x-ray radiation, a detector array (116) configured to detect x-ray radiation and generate a signal indicative thereof, and a reconstructor (118) configured to reconstruct the signal and generate sequential spares time line perfusion volumetric image data. The computer tomography scanner further includes a processor (132) configured to process the sequential spares time line perfusion volumetric image data using a trained neural network of a perfusion data enhancing module (136) to produce sequential dense time line perfusion volumetric image data.
Claims14
A computed tomography scanner, comprising:
- a radiation source configured to emit x-ray radiation;
- a detector array configured to detect the x-ray radiation and generate a signal indicative thereof;
- a reconstructor configured to reconstruct the signal and generate sequential sparse time line perfusion volumetric image data;
- and
- a processor configured to process the sequential sparse time line perfusion volumetric image data using a trained neural network to produce sequential dense time line perfusion volumetric image data, wherein the processor trains the neural network with 1) training sets of sequential sparse time line perfusion volumetric image data, 2) training sets of sequential dense time line perfusion volumetric image data, and 3) training sets of perfusion parameters generated from the training sets of sequential dense time line perfusion volumetric image data.
A computed tomography scanner, comprising:
- a radiation source configured to emit x-ray radiation;
- a detector array configured to detect the x-ray radiation and generate a signal indicative thereof;
- a reconstructor configured to reconstruct the signal and generate sequential sparse time line perfusion volumetric image data;
- and
- a processor configured to train a neural network to predict sequential dense time line perfusion volumetric image data from 1) training sequential sparse time line perfusion volumetric image data, 2) training sequential dense time line perfusion volumetric image data, and 3) training sets of perfusion parameters generated from the training sets of sequential dense time line perfusion volumetric image data.
A non-transitory computer readable storage medium encoded with computer readable instructions, which, when executed by a processor of a computing system, cause the processor to:
- process sequential sparse time line perfusion volumetric image data using a trained neural network to produce sequential dense time line perfusion volumetric image data, wherein the neural network is trained with 1) training sets of sequential sparse time line perfusion volumetric image data, 2) training sets of sequential dense time line perfusion volumetric image data, and 3) training sets of perfusion parameters generated from the training sets of sequential dense time line perfusion volumetric image data.
Dependencies are inferred from the claim text.