Reconstructing image data from magnetic resonance k-space data

A method for generating magnetic resonance image data, the method comprising:

• acquiring k-space data with an irregular sampling pattern via a magnetic resonance imaging system;
• and
• reconstructing image data using the acquired k-space data, using a trained machine learning model, the reconstructing being computer-implemented and comprising:
• receiving the acquired k-space data, wherein the acquired k-space data has an irregular sampling pattern in k-space;
• and
• in order to generate image data, performing an iterative machine-learning-based k-space-to-image reconstruction on the acquired k-space data with multiple reconstruction steps, the multiple reconstruction steps comprising:
• iterating between reconstruction steps of:
• applying a conventional image reconstruction involving a parallel imaging model relating image data to k-space data;
• and
• applying the trained machine learning model to enhance image data,
• wherein at least one of the conventional image reconstruction steps, at least one additional reconstruction step, or the at least one conventional image reconstruction step and the at least one additional reconstruction step comprise adapting data at a current iteration state to be more suitable for the trained machine learning model, to correspond more closely to training data the trained machine learning model has been trained with, or a combination thereof.

In a non-transitory computer-readable storage medium that stores instructions executable by a computer to reconstruct image data from magnetic resonance k-space data using a trained machine learning model, the instructions comprising:

• receiving acquired k-space data, wherein the acquired k-space data has an irregular sampling pattern in k-space;
• and
• in order to generate image data, performing an iterative machine-learning-based k-space-to-image reconstruction on the acquired k-space data with multiple reconstruction steps, the multiple reconstruction steps comprising:
• iterating between reconstruction steps of:
• applying a conventional image reconstruction involving a parallel imaging model relating image data to k-space data;
• and
• applying the trained machine learning model to enhance image data,
• wherein at least one of the conventional image reconstruction steps, at least one additional reconstruction step, or the at least one conventional image reconstruction step and the at least one additional reconstruction step comprise adapting data at a current iteration state to be more suitable for the trained machine learning model, to correspond more closely to training data the trained machine learning model has been trained with, or a combination thereof.

A magnetic resonance imaging system comprising:

• a processing unit configured to reconstruct image data from magnetic resonance k-space data using a trained machine learning model, the processing unit being configured to reconstruct the image data comprising the processing unit being configured to:
• receive acquired k-space data, wherein the acquired k-space data has an irregular sampling pattern in k-space;
• and
• in order to generate image data, perform an iterative machine-learning-based k-space-to-image reconstruction on the acquired k-space data with multiple reconstruction steps, the multiple reconstruction steps comprising:
• iteration between reconstruction steps of:
• application of a conventional image reconstruction involving a parallel imaging model relating image data to k-space data;
• and
• application of the trained machine learning model to enhance image data,
• wherein at least one of the conventional image reconstruction steps, at least one additional reconstruction step, or the at least one conventional image reconstruction step and the at least one additional reconstruction step comprise adaptation of data at a current iteration state to be more suitable for the trained machine learning model, to correspond more closely to training data the trained machine learning model has been trained with, or a combination thereof.