Method and systems for scan conversion with higher resolution

1. A method, comprising:

acquiring a first ultrasound dataset;

generating a second ultrasound dataset from the first ultrasound dataset with a neural network, the second ultrasound dataset larger than the first ultrasound dataset;

scan converting the second ultrasound dataset to generate an image; and

training the neural network with a plurality of acquired ultrasound datasets of a first size and a plurality of size-reduced ultrasound datasets of a second size generated from the acquired ultrasound datasets of the first size.

2. The method of claim 1 , wherein the second ultrasound dataset is two to four times larger than the first ultrasound dataset.

3. The method of claim 1 , wherein an image generated by scan converting the first ultrasound dataset is a same size as the image generated from the second ultrasound dataset.

4. The method of claim 1 , wherein training the neural network with the plurality of acquired ultrasound datasets of the first size and the plurality of size-reduced ultrasound datasets of the second size comprises:

inputting the plurality of size-reduced ultrasound datasets to the neural network to obtain a plurality of augmented ultrasound datasets of the first size;

scan converting the acquired ultrasound datasets to generate a plurality of ground truth images;

scan converting the plurality of augmented ultrasound datasets to generate a plurality of predicted images; and

updating weights of the neural network according to a difference between the plurality of ground truth images and the plurality of predicted images.

5. The method of claim 1 , wherein the first ultrasound dataset is of the second size and wherein the second ultrasound dataset is of the first size.

6. The method of claim 1 , further comprising displaying the image in real-time as the ultrasound dataset is acquired.

7. The method of claim 1 , wherein acquiring the ultrasound dataset comprises scanning a subject with an ultrasound probe.

8. A method, comprising:

acquiring a first ultrasound dataset;

generating a second ultrasound dataset from the first ultrasound dataset with a neural network, the second ultrasound dataset larger than the first ultrasound dataset; and scan converting the second ultrasound dataset to generate an image, wherein scan converting the second ultrasound dataset comprises applying one of nearest neighbor interpolation, linear interpolation, bilinear interpolation, cubic interpolation, or bicubic interpolation to geometrically transform the second ultrasound dataset into the image.

9. The method of claim 8 , wherein the second ultrasound dataset is two to four times larger than the first ultrasound dataset.

10. The method of claim 8 , wherein an image generated by scan converting the first ultrasound dataset is a same size as the image generated from the second ultrasound dataset.

11. The method of claim 8 , further comprising training the neural network with a plurality of acquired ultrasound datasets of a first size and a plurality of size-reduced ultrasound datasets of a second size generated from the acquired ultrasound datasets of the first size.

12. The method of claim 11 , wherein training the neural network with the plurality of acquired ultrasound datasets of the first size and the plurality of size-reduced ultrasound datasets of the second size comprises:

inputting the plurality of size-reduced ultrasound datasets to the neural network to obtain a plurality of augmented ultrasound datasets of the first size;

scan converting the acquired ultrasound datasets to generate a plurality of ground truth images;

scan converting the plurality of augmented ultrasound datasets to generate a plurality of predicted images; and

updating weights of the neural network according to a difference between the plurality of ground truth images and the plurality of predicted images.

13. The method of claim 11 , wherein the first ultrasound dataset is of the second size and wherein the second ultrasound dataset is of the first size.

14. An ultrasound system, comprising:

an ultrasound probe including an array of transducer elements; and

a processor coupled to the ultrasound probe and configured with executable instructions stored on non-transitory memory that when executed during operation of the ultrasound system, cause the processor to:

acquire, with the ultrasound probe, a first ultrasound dataset;

generate a second ultrasound dataset from the first ultrasound dataset with a neural network, the second ultrasound dataset larger than the first ultrasound dataset; and

scan convert the second ultrasound dataset to generate an image, wherein scan converting the second ultrasound dataset to generate the image comprises applying one of nearest neighbor interpolation, linear interpolation, bilinear interpolation, cubic interpolation, or bicubic interpolation to geometrically transform the second ultrasound dataset into the image.

15. The system of claim 14 , further comprising a display device communicatively coupled to the processor, wherein the processor is further configured with executable instructions stored on non-transitory memory that when executed cause the processor to output, via the display device, the image.

16. The system of claim 14 , wherein the second ultrasound dataset is at least two to four times larger than the first ultrasound dataset.

17. The system of claim 14 , wherein the neural network is trained on a plurality of acquired ultrasound datasets of a first size and a plurality of size-reduced ultrasound datasets of a second size generated from the acquired ultrasound datasets of the first size.

18. The system of claim 17 , wherein the neural network is trained with the plurality of acquired ultrasound datasets and the plurality of size-reduced ultrasound datasets by inputting the plurality of size-reduced ultrasound datasets into the neural network to obtain a plurality of augmented ultrasound datasets of the first size, scan converting the acquired ultrasound datasets to generate a plurality of ground truth images, scan converting the plurality of augmented ultrasound datasets to generate a plurality of predicted images, and updating weights of the neural network according to a difference between the plurality of ground truth images and the plurality of predicted images.

19. An ultrasound system, comprising:

an ultrasound probe including an array of transducer elements; and

a processor coupled to the ultrasound probe and configured with executable instructions stored on non-transitory memory that when executed during operation of the ultrasound system, cause the processor to:

acquire, with the ultrasound probe, a first ultrasound dataset;

generate a second ultrasound dataset from the first ultrasound dataset with a neural network, the second ultrasound dataset larger than the first ultrasound dataset; and

scan convert the second ultrasound dataset to generate an image, wherein the neural network is trained on a plurality of acquired ultrasound datasets of a first size and a plurality of size-reduced ultrasound datasets of a second size generated from the acquired ultrasound datasets of the first size.

20. The system of claim 19 , wherein the neural network is trained with the plurality of acquired ultrasound datasets and the plurality of size-reduced ultrasound datasets by inputting the plurality of size-reduced ultrasound datasets into the neural network to obtain a plurality of augmented ultrasound datasets of the first size, scan converting the acquired ultrasound datasets to generate a plurality of ground truth images, scan converting the plurality of augmented ultrasound datasets to generate a plurality of predicted images, and updating weights of the neural network according to a difference between the plurality of ground truth images and the plurality of predicted images.