Deep learning for perfusion in medical imaging
For decision support based on perfusion in medical imaging, a machine-learned model, such as a model trained with deep learning, generates perfusion examination information from CT scans of the patient. Other information, such as patient-specific information, may be used with the CT scans to generate the perfusion examination information. Since a machine-learned model is used, the perfusion examination information may be estimated from a spatial and/or temporally sparse number of scan shots or amount of CT dose. The results of perfusion imaging may be provided with less than the current, standard, or typical radiation dose.
1. A method for estimating myocardial perfusion by a medical imager, the method comprising:
scanning a patient with a computed tomography (CT) system using one or more injection protocol parameters that provide an effective dose corresponding to no more than 100 CT dose index, the scanning providing sparse coronary CT data representing contrast agent in a heart of the patient over time;
acquiring information for the patient in addition to the sparse coronary CT data;
estimating the myocardial perfusion by a machine-learned model having been trained by deep learning using training data comprising sequential CT images, associated injection protocol parameters, and associated patient specific information, the machine-learned model outputting the myocardial perfusion in response to input of the sparse coronary CT data, the one or more injection protocols, and the information; and
displaying the myocardial perfusion.
2. The method of claim 1 wherein scanning comprises scanning with fewer than five shots, and wherein estimating comprises estimating in response to input of the sparse coronary CT data from no more than the five shots.
3. The method of claim 1 wherein acquiring the information comprises acquiring patient attributes for the patient.
4. The method of claim 1 wherein acquiring the information comprises acquiring computed tomography angiography image of the patient.
5. The method of claim 1 wherein acquiring the information comprises acquiring an arterial enhancement curve information for the patient.
6. The method of claim 1 wherein acquiring the information comprises acquiring static myocardial measure for the patient.
7. The method of claim 1 wherein estimating comprises estimating the myocardial perfusion as a parametric image.
8. The method of claim 1 wherein estimating further comprises estimating the myocardial perfusion as a peak, time-to-peak, blow flow, blood volume, mean transit time, or combination thereof.
9. The method of claim 1 further comprising determining, by an image processor, a diagnosis or therapy from the estimated myocardial perfusion, and wherein displaying comprises displaying the diagnosis or therapy.
10. The method of claim 1 wherein estimating comprises estimating with the machine-learned model comprising a generative network having been trained adversarially to generate synthetic images from a fewer number of real perfusion images.
11. A system for decision support based on perfusion in medical imaging, the system comprising:
a computed tomography (CT) scanner for scanning a patient while contrast agent is in the patient using one or more injection protocol parameters that provide an effective dose corresponding to no more than 100 CT dose index, the CT scanner configured to output sparse CT data for the patient;
a memory configured to store a machine-learned model trained using training data comprising CT images and associated patient specific information;
an image processor configured to estimate perfusion information by input of the sparse CT data, the one or more injection protocol parameters, and patient-specific information to the machine-learned model, the perfusion information provided with no more than the six or fewer scan shots of the sparse CT data to represent the contrast agent; and
a display configured to display the perfusion information.
12. The system of claim 11 wherein the patient-specific information comprises a patient attribute, an angiography image, an arterial enhancement curve, and/or quantification from the sparse CT data.
13. The system of claim 11 wherein the perfusion information comprises a perfusion image, a quantitative value of perfusion, a diagnosis, and/or a therapy.
14. A method for decision support based on perfusion measured by a medical imager, the method comprising:
scanning, using one or more injection protocol parameters that provide an effective dose corresponding to no more than 100 CT dose index, a patient with a computed tomography (CT) system, the scanning providing sparse CT data representing contrast agent in the patient over time;
generating perfusion examination information by a machine-learned model applied to the sparse CT data, the one or more injection protocols, and patient specific information; and
outputting the perfusion examination information.
15. The method of claim 14 wherein generating comprises generating a diagnosis and/or therapy as the perfusion examination information.
16. The method of claim 14 wherein generating comprises generating by the machine-learned model comprising a deep-learned neural network applied to the sparse CT data and personalized patient information.