Communication system for magnetic resonance imaging

A computer program comprising machine executable instructions stored on a non-transitory computer readable medium, which when executed by a processor controlling a computational system, wherein the computational system is configured to control a medical system that comprises a magnetic resonance imaging communication system, wherein the magnetic resonance imaging communication system comprises:

• a subject microphone configured to acquire measured audio data from a subject during operation of a magnetic resonance imaging system, wherein the magnetic resonance imaging communication system is configured to acquire the measured audio data as a continuous data stream during acquisition of k-space data by the magnetic resonance imaging system;
• at least one neural network processing module, wherein the at least one neural network processing module is configured to output at least one audio signal in response to receiving the measured audio data as input, wherein the at least one audio signal comprises a subject speech audio signal, wherein the at least one neural network processing module is configured to repeatedly provide the at least one audio signal in response to receiving the continuous data stream;
• wherein the at least one neural network if formed as a a-neuromorphic processor or the at least one neural network comprises a first neural network processing module and a second neural network processing module, wherein the first neural network processing module is configured to output the subject speech audio signal in response to receiving the measured audio data, wherein the second neural network is configured to output the gradient noise signal in response to receiving the measured audio data, and wherein the first neural network processing module and the second neural network processing module are configured to operate in parallel and wherein execution of the machine executable instructions causes the computational system to:
• control the subject microphone to acquire the measured audio data;
• and
• receive the at least one audio signal in response to inputting the measured audio data into the at least one neural network processing module.

A method of operating a medical system, wherein the medical system comprising a magnetic resonance imaging communication system, wherein the magnetic resonance imaging communication system comprises:

• a subject microphone configured to acquire measured audio data from a subject during operation of a magnetic resonance imaging system, wherein the magnetic resonance imaging communication system is configured to acquire the measured audio data as a continuous data stream during acquisition of k-space data by a magnetic resonance imaging system;
• at least one neural network processing module, wherein the at least one neural network processing module is configured to output at least one audio signal in response to receiving the measured audio data as input, wherein the at least one audio signal comprises a subject speech audio signal, wherein the at least one neural network processing module is configured to repeatedly provide the at least one audio signal in response to receiving the continuous data stream;
• wherein the at least one neural network if formed as a neuromorphic processor or the at least one neural network comprises a first neural network processing module and a second neural network processing module, wherein the first neural network processing module is configured to output the subject speech audio signal in response to receiving the measured audio data, wherein the second neural network is configured to output the gradient noise signal in response to receiving the measured audio data, and wherein the first neural network processing module and the second neural network processing module are configured to operate in parallel wherein the method comprises:
• controlling the subject microphone to acquire the measured audio data;
• and
• receiving the at least one audio signal in response to inputting the measured audio data into the at least one neural network processing module.