Multi-shot echo planar imaging through machine learning
Systems and methods are provided for improving MRI data acquisition efficiency while providing more detailed information with high resolution and isotropic resolution without gaps. Improved data acquisition efficiency may be achieved by implementing a machine learning algorithm with a hardware processor and a memory to estimate imperfections in fast imaging sequences, such as a multi-shot echo planar imaging (MS-EPI) sequence. These imperfections, such as patient motion, physiological noise, and phase variations, may be difficult to model or otherwise estimate using standard physics-based reconstructions.
1. A method for reducing image artifacts using a magnetic resonance imaging system, comprising:
a) acquiring image data of a subject using a magnetic resonance imaging (MRI) system implementing a rapid imaging sequence;
b) providing to a computer system a trained machine learning algorithm that has been trained on training data comprising corrupted image data and physiological artifacts in order to reconstruct images from image data such that the reconstructed images depict reduced physiological artifacts;
c) applying the acquired image data to the trained machine learning algorithm, and generating an output that comprises one or more images reconstructed from the image data, wherein the one or more images depict reduced physiological artifacts; and
wherein the trained machine learning algorithm is a residual machine learning algorithm that is trained on the training data to map between corrupted images and image artifacts.
2. The method of claim 1 wherein the rapid imaging sequence is a multi-shot echo planar imaging (MS-EPI) sequence.
3. The method of claim 2 wherein the physiological artifacts include shot-to-shot motion artifacts.
4. The method of claim 2 wherein the one or more reconstructed images depict contrast weighting associated with gradient-echo (GRE) contrast.
5. The method of claim 1 wherein the residual machine learning algorithm comprises a convolutional neural network.
6. The method of claim 1 wherein the physiological artifacts include at least one of patient motion, physiological noise, or phase variations.
7. A system for reducing image artifacts using a magnetic resonance imaging system, comprising:
a computer system configured to
i) acquire image data of a subject using a magnetic resonance imaging (MRI) system implementing a rapid imaging sequence;
ii) implement a trained machine learning algorithm that has been trained on training data comprising corrupted image data and physiological artifacts in order to reconstruct images from image data such that the reconstructed images depict reduced physiological artifacts;
iii) apply the acquired image data to the trained machine learning algorithm, and generate an output that comprises one or more images reconstructed from the image data, wherein the one or more images depict reduced physiological artifacts; and
wherein the trained machine learning algorithm is a residual machine learning algorithm that is trained on the training data to map between corrupted images and image artifacts.
8. The system of claim 7 wherein the rapid imaging sequence is a multi-shot echo planar imaging (MS-EPI) sequence.
9. The system of claim 8 wherein the physiological artifacts include shot-to-shot motion artifacts.
10. The system of claim 8 wherein the one or more reconstructed images depict contrast weighting associated with gradient-echo (GRE) contrast.
11. The system of claim 7 wherein the residual machine learning algorithm comprises a convolutional neural network.
12. The system of claim 7 wherein the physiological artifacts include at least one of patient motion, physiological noise, or phase variations.
13. A non-transitory computer-readable medium storing instructions, the instructions comprising: A method for reducing image artifacts using a magnetic resonance imaging system, comprising:
one or more instructions that, when executed by one or more processors, cause the one or more processors to:
i) acquire image data of a subject using a magnetic resonance imaging (MRI) system implementing a rapid imaging sequence;
ii) implement a trained machine learning algorithm that has been trained on training data comprising corrupted image data and physiological artifacts in order to reconstruct images from image data such that the reconstructed images depict reduced physiological artifacts;
iii) apply the acquired image data to the trained machine learning algorithm, and generate an output that comprises one or more images reconstructed from the image data, wherein the one or more images depict reduced physiological artifacts, and
wherein the trained machine learning algorithm is a residual machine learning algorithm that is trained on the training data to map between corrupted images and image artifacts.
14. The non-transitory computer-readable medium of claim 13 wherein the rapid imaging sequence is a multi-shot echo planar imaging (MS-EPI) sequence.
15. The non-transitory computer-readable medium of claim 14 wherein the physiological artifacts include shot-to-shot motion artifacts.
16. The non-transitory computer-readable medium of claim 14 wherein the one or more reconstructed images depict contrast weighting associated with gradient-echo (GRE) contrast.
17. The non-transitory computer-readable medium of claim 13 wherein the residual machine learning algorithm comprises a convolutional neural network.
18. The non-transitory computer-readable medium of claim 13 wherein the physiological artifacts include at least one of patient motion, physiological noise, or phase variations.