Provision of corrected medical image data

A computer-implemented method for provision of a trained function, the computer-implemented method comprising:

• receiving a first training data record;
• receiving at least one further training data record;
• applying an image quality metric to the first training data record and the at least one further training data record, such that an image quality parameter is determined in each case;
• identifying the first training data record or the at least one further training data record as a comparison data record, the identifying of the comparison data record comprising comparing the image quality parameters, the comparison data record having a highest image quality;
• applying the trained function to input data, wherein the input data is based on the first training data record and the at least one further training data record;
• determining whether the first training data record has a higher image quality compared to the at least one further training data record based on the applying of the trained function;
• adjusting at least one parameter of the trained function based on a comparison of the comparison data record with a result of the applying of the trained function to the input data;
• and
• providing the trained function.

A provision unit comprising:

• a processor configured to provide corrected medical image data, the provision of the corrected medical image data comprising:
• receipt of medical image data of an examination object;
• creation of a first temporary data record, the creation of the first temporary data record comprising application of a first correction to the medical image data;
• creation of at least one further temporary data record, the creation of the at least one further temporary data record comprising application of a further correction to the medical image data, wherein the further correction corresponds to the first correction with at least one interference term that is applied to at least one parameter of the first correction;
• application of a trained function to input data, wherein the input data is based on the first temporary data record and the at least one further temporary data record, wherein at least one parameter of the trained function is based on an image quality metric;
• determination of whether the first temporary data record has a higher image quality compared to the at least one further temporary data record based on the application of the trained function to the input data;
• provision of the first temporary data record as the corrected medical image data when a result of the determining is positive;
• and
• repetition of the creation of the first temporary data record, the creation of the at least one further temporary data record, and the application of the trained function to the input data using the at least one further temporary data record as the medical image data when the result of the determining is negative.

A medical imaging device comprising:

• a provision unit comprising:
• a processor configured to provide corrected medical image data, the provision of the corrected medical image data comprising:
• receipt of medical image data of an examination object;
• creation of a first temporary data record, the creation of the first temporary data record comprising application of a first correction to the medical image data;
• creation of at least one further temporary data record, the creation of the at least one further temporary data record comprising application of a further correction to the medical image data, wherein the further correction corresponds to the first correction with at least one interference term that is applied to at least one parameter of the first correction;
• application of a trained function to input data, wherein the input data is based on the first temporary data record and the at least one further temporary data record, wherein at least one parameter of the trained function is based on an image quality metric;
• determination of whether the first temporary data record has a higher image quality compared to the at least one further temporary data record based on the application of the trained function to the input data;
• provision of the first temporary data record as the corrected medical image data when a result of the determining is positive;
• and
• repetition of the creation of the first temporary data record, the creation of the at least one further temporary data record, and the application of the trained function to the input data using the at least one further temporary data record as the medical image data when the result of the determining is negative,
• wherein the medical imaging device is configured to record the medical image data, receive the medical image data, provide the medical image data, or any combination thereof.

In a non-transitory computer-readable storage medium that stores instructions executable by a provision unit of a medical imaging device to provide corrected medical image data, by a training unit to provide a trained function, or by the provision unit to provide the corrected medical image data and the training unit to provide the trained function, the instructions comprising:

• providing the corrected medical image data, the providing of the corrected medical image data comprising:
• receiving medical image data of an examination object;
• creating a first temporary data record, the creating of the first temporary data record comprising applying a first correction to the medical image data;
• creating at least one further temporary data record, the creating of the at least one further temporary data record comprising applying a further correction to the medical image data, wherein the further correction corresponds to the first correction with at least one interference term that is applied to at least one parameter of the first correction;
• applying a trained function to first input data, wherein the first input data is based on the first temporary data record and the at least one further temporary data record, wherein at least one parameter of the trained function is based on an image quality metric;
• determining whether the first temporary data record has a higher image quality compared to the at least one further temporary data record based on the applying of the trained function to the first input data;
• providing the first temporary data record as the corrected medical image data when a result of the determining is positive;
• and
• repeating the creating of the first temporary data record, the creating of the at least one further temporary data record, and the applying of the trained function to the first input data using the at least one further temporary data record as the medical image data when the result of the determining is negative;
• providing the trained function, the providing of the trained function comprising:
• receiving a first training data record;
• receiving at least one further training data record;
• applying an image quality metric to the first training data record and the at least one further training data record, such that an image quality parameter is determined in each case;
• identifying the first training data record or the at least one further training data record as a comparison data record, the identifying of the comparison data record comprising comparing the image quality parameters, the comparison data record having a highest image quality;
• applying the trained function to second input data, wherein the second input data is based on the first training data record and the at least one further training data record;
• determining whether the first training data record has a higher image quality compared to the at least one further training data record based on the applying of the trained function;
• adjusting at least one parameter of the trained function based on a comparison of the comparison data record with a result of the applying of the trained function to the second input data;
• and
• providing the trained function;
• or
• a combination thereof.