Adapting a media signal to an MRI scan using a generative AI system

Abstract

The invention relates to a method for performing an MRI examination of a patient using an MRI system, the MRI system, a computer-implemented method for providing a media signal using an AI system, the AI ​​system, and an associated computer program product. The inventive method for performing an MRI examination of a patient using an MRI system comprises the following steps: - Adjusting an examination parameter of the MRI examination to the patient using a control unit of the MRI system, - Providing the adapted investigation parameter via the control unit as an input parameter for a generative AI system, wherein the generative AI system is configured to generate a media signal as an output parameter in response to an input of the provided input parameter and to make the generated media signal available for retrieval, wherein the generated media signal depends on the adapted investigation parameter and includes an audio track and / or a video track, - Receiving the generated media signal by means of a playback unit, - Performing the MRI examination of the patient according to the adapted MRI examination parameters using the MRI system in such a way that the received media signal is played back to the patient via the playback unit during the procedure.

Description

[0001] The invention relates to a method for performing an MRI examination of a patient using an MRI system, the MRI system itself, and an associated computer program, and relates in particular to the application of these technologies to improve image quality and diagnostic accuracy. This application addresses the problem that MRI examinations are often unpleasant and / or stressful for patients, especially due to the loud noises and / or the long duration of the MRI examination.

[0002] MRI scans are generally very loud and therefore perceived as unpleasant by many patients. The main reason for the high noise levels is the gradient coils of the MRI system's magnet unit, which are typically subjected to rapidly changing currents during the scan. The resulting forces cause mechanical distortions, vibrations, and / or movements of the gradient coil. These forces are partially transmitted to the other components of the MRI system, causing the MRI system, especially its tube, to act like a loudspeaker. Noise levels of sometimes exceeding 100 dBA are regularly reached during MRI scans, making hearing protection mandatory for the patient. Therefore, media signals, particularly audio tracks, are conventionally played through loudspeakers or headphones for the patient to hear, in order to distract and / or entertain them.Due to the high volume of the gradient coils, the media signal typically cannot be perceived without interference.

[0003] Over the past few decades, various hardware-based approaches and sequence / software techniques have been developed to make MRI scans significantly quieter. There are also approaches to equipping headphones with better noise isolation. Nevertheless, MRI scans can still be perceived as very loud.

[0004] DE 10 2014 222 496 A1 relates to a method for magnetic resonance imaging with a music-based gradient profile, comprising the following process steps: providing a magnetic resonance sequence and a piece of music, modifying the piece of music taking into account at least one acoustic resonance parameter which characterizes at least one acoustic resonance frequency of the magnetic resonance device, thereby generating a modified piece of music, adapting a gradient profile of the magnetic resonance sequence using the modified piece of music, thereby generating an adapted magnetic resonance sequence, and acquiring magnetic resonance image data of the subject using the adapted magnetic resonance sequence.In other words, the aim is to adapt the MRI scan to a provided piece of music as the soundtrack of a video signal in such a way that the MRI scan is ideally perceived as less unpleasant at the same noise emission levels. Since the contrast of an MRI scan depends heavily on the chosen repetition times, echo times, and / or echo spacing, adapting the MRI scan to the provided music is typically only realistic in a few cases and / or only for specific MRI scans such as MR fingerprinting.

[0005] From WO 2010 / 018 534 A1, it is known to enhance the noise generated by the gradient coils in an MRI scanner by playing a suitable piece of music that matches the noise of the gradient coils in tempo and / or key when a patient undergoes an MRI examination. The prior art also includes designs in which music was generated to match the sound of the noise emissions of an MRI system (see https: / / www.youtube.com / watch?v=e9593SnYwyo), whereby the media signal is thus acoustically correlated with the noise emissions of the specified gradient profile, in particular the noise emissions are temporarily integrated into the audio track of the media signal.

[0006] According to EP 4 185 204 B1, during a medical examination, the patient is provided with an entertainment program while fully conscious. This program is dynamically created and adapted, and the duration of the entertainment is continuously adjusted to the actual length of the medical procedure, even if it lengthens or shortens during the procedure.

[0007] US 2011 / 0 142 250 A1 discloses that when a patient undergoes an MRI scan, the noise generated by the gradient coils in the MRI scanner is enhanced by playing a complementary piece of music that matches the gradient coil noise in tempo and / or key. Complementary music is pre-generated for specific gradient coil sequences. After selecting one or more sequences to be performed during the MRI scan, the corresponding complementary music is identified and played to the patient in the MRI scanner tunnel during the scan to reduce patient stress. The tempo and / or key of the complementary music is adjustable in advance or in real time, allowing the music to be rhythmically and harmonically synchronized with a specific gradient sequence.

[0008] Furthermore, many patients find MRI scans unpleasant, partly due to the unfamiliar environment and / or the narrow tube. Many patients are also nervous, as the scan results can be critical for their further treatment. During the scan, media signals with a video track are conventionally used to distract patients from these unpleasant thoughts and sensations. This is particularly effective for children, who ideally require less sedation as a result. The media signals used include, for example, cartoons, documentaries, or videos from a local video database and / or internet-based platforms. Patients can often select a media signal themselves, preferably to give them a sense of control over their surroundings.The problem with the media signals used so far may be that they are not tailored to the specific examination depending on the patient.

[0009] This applies, for example, to the total duration of the media signal. MRI scans can last between 5 and 45 minutes, depending on the body region and / or the patient. The exact total duration changes regularly during the MRI scan, especially if individual sequences need to be repeated. Therefore, it is virtually impossible to pre-select a media signal for the patient whose total duration essentially matches the total duration of the MRI scan. This typically results in the MRI scan ending before the media signal has finished, particularly before the video sequence can be viewed to completion, and / or the media signal ending before the MRI scan has finished. This can regularly irritate patients, and is especially frustrating for pediatric patients.

[0010] In particular, the media signals used so far are not tailored to the MRI scan. These signals typically do not contain any information about the MRI scan, such as the remaining scan time.

[0011] Currently, selecting suitable media signals requires the experience and / or intuition of the MRI system operator, particularly the radiographer who manages the patient and operates the system. Depending on the patient's preferences and / or the MRI examination, the operator manually selects media signals that may be suitable in terms of duration and / or content. To reduce the workload, the operator can regularly select a media signal that might be suitable for as many patients as possible. Alternatively, media signals can be played continuously for several hours across multiple MRI examinations and / or in a continuous loop. This conventionally avoids patient irritation caused by mismatches between the total duration of the MRI examination and the total duration of the media signal.

[0012] Another way to adapt media signals to MRI scans is to overlay the video track of the media signal with information about the MRI scan. For example, there is a known infotainment system where a progress bar can be displayed over the video track to visualize the remaining time of the MRI scan.

[0013] The invention is based on the objective of providing an improved method for performing an MRI examination of a patient using an MRI system, the improved MRI system and an associated computer program product.

[0014] The problem is solved by the features of the independent claims. Advantageous embodiments are described in the dependent claims.

[0015] The inventive method for performing an MRI examination of a patient using an MRI system comprises the following steps: - Adjusting an examination parameter of the MRI examination to the patient using a control unit of the MRI system, - Providing the adapted investigation parameter via the control unit as an input parameter for a generative AI system, wherein the generative AI system is configured to generate a media signal as an output parameter in response to an input of the provided input parameter and to make the generated media signal available for retrieval, wherein the generated media signal depends on the adapted investigation parameter and includes an audio track and / or a video track, - Receiving the generated media signal by means of a playback unit, - Performing the MRI examination of the patient according to the adapted MRI examination parameters using the MRI system in such a way that the received media signal is played back to the patient via the playback unit during the procedure.

[0016] The method according to the invention offers, in particular, improvements to make the MRI examination experience more pleasant and comfortable for the patient. By using the generative AI system with examination parameters adapted to the patient, a patient-dependent media signal is generated and played back during the MRI examination. This media signal can advantageously include an audio track and / or a video track based on the individual characteristics of the patient. This preferably provides the patient with a pleasant distraction and / or reduces feelings of anxiety. Overall, the invention contributes to improving the patient's comfort and satisfaction during the MRI examination.

[0017] Adapting the examination parameters to the individual patient enables, in particular, a personalized MRI scan, which can increase patient well-being and improve the quality of the examination results. Using a generative AI system to generate the patient-specific media signal can preferably increase patient compliance during the MRI scan by creating a more pleasant environment that is less distracting from the stress of the procedure. Playing the media signal during the MRI scan can especially help reduce motion artifacts, as the media signal entertains and calms the patient, potentially leading to higher image quality.

[0018] The media signal, which is part of or constitutes the generated output parameter, advantageously depends on both the adapted examination parameter and the patient. This ideally creates a close link between the individual needs of the patient and the execution of the MRI examination.

[0019] The MRI system according to the invention features - a magnetic unit comprising a main magnet and at least one gradient coil, - the control unit - and a playback unit, wherein the MRI system is configured to carry out the method according to the invention.

[0020] The design of the MRI system for performing the method according to the invention enables seamless integration of imaging and patient-oriented functions, which can advantageously increase system efficiency and user-friendliness. The control unit, which is specifically configured for the method, can allow the imaging parameters to be adapted to the patient, preferably leading to a reduction in examination time and an optimization of the workflow.

[0021] A computer-implemented method for providing a media signal using an AI system includes the following steps: - Receiving an examination parameter from a patient's MRI examination as an input parameter, - Generating a media signal as an output parameter in response to an input parameter, - Providing the generated media signal for retrieval, wherein the generated media signal depends on the adapted investigation parameter and includes an audio track and / or a video track.

[0022] A suitable AI system for this purpose points to - an input interface for receiving an input parameter, - a computing unit for generating an output parameter depending on the received input parameter by applying a generative AI model to the input parameter, - an output interface for providing the generated output parameter, wherein the AI ​​system is configured, in particular trained, to provide the media signal. The AI ​​system can be further developed, in particular, as described in the following embodiments.

[0023] The input interface typically allows for flexible adaptation of the AI ​​system to various input parameters, increasing its applicability to a wide range of clinical scenarios. The computing unit, specifically trained to generate media signals as output parameters, can improve diagnostic efficiency, particularly if the patient feels more comfortable during the MRI scan. The output interface facilitates the integration of data generated by the AI ​​system into clinical workflows, thus primarily supporting decision-making and patient care.

[0024] The term "MRI system" here refers to the medical device suitable for and / or used to perform MRI scans, and specifically includes a magnet unit, a control unit, and possibly an AI system. The MRI system may also be referred to as an MRI scanner.

[0025] The term "gradient coil" here refers to a coil in the MRI system that is used to generate magnetic gradients. These gradients enable the spatial coding of signals during the MRI examination.

[0026] MRI, also known as magnetic resonance imaging, is a medical imaging technique that uses superimposed magnetic fields and radio waves, generated by the magnet unit and / or at least one gradient coil, to produce detailed images of the patient. MRI scans are particularly useful for examining the patient's soft tissues. While MRI can theoretically be used on objects rather than patients, it is not the only option. Unlike X-rays or CT scans, the MRI system does not use ionizing radiation and is therefore a safe and non-invasive method for patient examination.

[0027] The term "patient" here refers to the person undergoing the MRI scan and who perceives the generated signal during the examination. During an MRI scan, the patient is moved into a tube surrounded by a magnetic field. This magnetic field aligns the hydrogen atoms in the patient's body. Radio waves are then transmitted through the body, disrupting the alignment of the hydrogen atoms. When the radio waves are switched off, the atoms return to their original alignment according to the material-specific T1 and T2 times, emitting signals that are detected by a receiver unit of the MRI system. These signals are then processed by a computer unit to generate material-differentiated images of the body.

[0028] MRI scans can be performed in different planes to examine various parts of the body. The images can also be created in 3D to provide an even more detailed representation of the patient's body. MRI scans are particularly useful for diagnosing diseases of the brain, spinal cord, joints, liver, and other organs. They can also be used to detect tumors, inflammation, infections, and other conditions.

[0029] MRI scans are painless and / or typically last between 5 and 45 minutes. During the scan, the patient must lie still to avoid motion artifacts. In some cases, a contrast agent may be injected to make certain areas of the body more visible. After the scan, the images can be evaluated, particularly by a radiologist, to make a diagnosis and / or monitor the progression of a disease.

[0030] The term "MRI scan" here refers to the process of performing an MRI examination, in which at least one image of at least one part of a patient's body is generated, primarily to obtain diagnostic information about that part of the body. The MRI scan is commonly called a scan.

[0031] The term "MRI sequence" here refers to the process of acquiring specific signals during an MRI scan. These specific signals are used to generate appropriately configured images of the patient's body and / or to obtain diagnostic information about them.

[0032] MRI examinations are typically performed according to an MRI protocol. This protocol includes, in particular, at least one MRI sequence. The protocol may also include several MRI sequences, which are typically performed sequentially in as uninterrupted a sequence as possible.

[0033] Each MRI sequence is typically performed according to its specific examination parameter(s). The MRI examination is usually performed according to a variety of examination parameters within a single MRI sequence or multiple MRI sequences. The MRI sequences within the MRI examination protocol may differ in at least one examination parameter, particularly in the scan volume, contrast, resolution, orientation, and / or the timing of contrast agent administration. If two MRI sequences differ only in the timing of contrast agent administration, the examination parameters of these two sequences are not identical. If two MRI sequences differ only in the absolute starting time of the first sequence, the examination parameters of these two sequences are typically identical.The absolute starting time is, in particular, a time of day.

[0034] The total duration of the MRI scan typically corresponds to the total duration of the MRI scan protocol. The total duration of the MRI scan is typically defined as the period between the start time of the first MRI sequence in the MRI scan protocol and the end time of the last MRI sequence in the MRI scan protocol. In particular, preparation or follow-up time, especially for positioning the patient on the examination table and / or moving the table with the patient into or out of the MRI scanner, is typically not included in the total duration of the MRI scan.

[0035] Typically, examination parameters can be divided into patient-specific and patient-non-customizable parameters. Adjusting a patient-specific parameter can, for example, trigger an automatic adjustment of a patient-non-customizable parameter by the control unit, and vice versa. It is conceivable that the control unit automatically adjusts another patient-specific parameter and / or another patient-non-customizable parameter after the examination parameter has been adjusted to the patient. Typically, one patient-specific parameter is adjusted to the patient.

[0036] The examination parameters can typically be divided into sequence parameters and workflow parameters. Sequence parameters relate in particular to how the magnetic fields and / or radio waves are generated, set, and / or varied using the MRI system. Such an examination parameter might include, for example, the repetition time, echo time, and / or echospacing of the MRI scan. Workflow parameters relate in particular to behavioral parameters that the patient is expected to follow for at least part of the total duration of the MRI scan in order to improve the quality of the generated images. Such an examination parameter might include, in particular, one or more specific time points and / or a total or remaining duration for the patient to remain lying down, hold their breath, stay still, and / or relax.

[0037] The examination parameter can be temporally variable, particularly with temporal resolution, and / or vary over time, for example, in its temporal sequence and / or intensity. The temporal sequence can include a multitude of different or identical repetition times, echo times, and / or echo spacings. Alternatively or additionally, the temporal sequence can include one or more points in time and / or the total or remaining duration of lying down, breath-holding, stillness, and / or relaxation. It is conceivable that the intensity of the temporally variable examination parameter changes, for example, differentiating between shallow and normal breathing. Alternatively or additionally, the intensities of the multiple radio waves can vary.

[0038] The term "control unit" here refers to a component of the MRI system that serves, in particular, to adapt the examination parameters to the patient and provide them as input parameters for the AI ​​system. The control unit may, in particular, include a processing unit and / or a logic module to execute program code that performs the adaptation of the examination parameters, the provision of the adapted examination parameters, the reception of the generated media signal, the execution of the MRI examination, and / or the playback of the received media signal. The control unit may also include a storage unit to save, buffer, and / or retrieve the adapted examination parameters and / or the generated media signal.

[0039] The control unit can be a local component of the MRI system, physically located within the MRI examination room or in a control room overlooking the MRI examination room. Alternatively, the control unit can be a remote component of the MRI system, which may be located, for example, in a server room belonging to the operator of the MRI system and / or in a cloud.

[0040] The control unit may include input devices for adjusting the examination parameter, in particular a keyboard, a computer mouse, and / or a screen with or without a touch-sensitive surface (touchscreen), and / or a sensor for patient detection. Adjusting the examination parameter can be done by a user of the MRI system and / or semi-automatically or fully automatically. It is conceivable that some examination parameters are adjusted automatically and others by the user. Adjusting the examination parameter means, in particular, changing and / or setting the examination parameter.

[0041] The control unit may, in particular, include an interface for providing the adapted test parameter. The adapted test parameter may, for example, be transferred from the control unit's interface to the AI ​​system's input interface. It is conceivable that the AI ​​system's input interface retrieves the adapted test parameter from the control unit, specifically from the control unit's interface. The interface may, in particular, be a network interface, especially an Ethernet, Wi-Fi, or mobile internet interface, and / or a data interface accessible to other units, especially the AI ​​system and / or a computing unit.

[0042] The control unit can typically provide the customized examination parameter in a file format and / or a specific signal waveform. The control unit and / or the MRI system typically includes a conversion unit for converting the customized examination parameter into the file format and / or the specific signal waveform.

[0043] Providing the adapted examination parameter as an input parameter for the generative AI system means, in particular, that an input parameter is provided for the generative AI system, where the input parameter includes at least the adapted examination parameter. The adapted examination parameter can be provided immediately after the examination parameter has been adapted, especially before any other examination parameter of the MRI examination has been adapted. Alternatively, it is conceivable that the adapted examination parameter is only provided once all necessary examination parameters of the MRI examination have been adapted, i.e., once the MRI examination can be performed.

[0044] The provision of the adjusted examination parameter typically occurs automatically. Alternatively or additionally, it is conceivable that the user of the control unit triggers the provision of the adjusted examination parameter, in particular by activating a key or button.

[0045] Providing the adapted investigation parameter via the control unit as an input parameter for the generative AI system can additionally include triggering the application of the AI ​​system's generative AI model to the adapted investigation parameter as input parameter. Triggering the application corresponds, in particular, to the commencement of the generation of the media signal by the AI ​​system.

[0046] The term "AI system" here refers to a generative system that generates an output parameter based on the provided input parameter. It is designed to generate the media signal depending on the input parameters. The AI ​​system is a system that uses artificial intelligence (AI) to generate output parameters based on input parameters. It specifically includes an input interface, a processing unit, and an output interface. Typically, the AI ​​system is network-based.

[0047] The AI ​​system can be internet-based, for example, cloud-based. The AI ​​system can be provided by an operator other than the operator and / or manufacturer of the MRI system. The AI ​​system can be subject to a fee per application, based on the adapted examination parameters. Alternatively, the AI ​​system can be provided as part of a data center operated by the MRI system operator and / or manufacturer, and / or as part of the control unit.

[0048] The input interface is designed to receive at least one input parameter, which is provided as input for the AI ​​system. This input parameter(s) can contain various pieces of information relevant for generating the output parameter. In particular, the adapted investigation parameter is one or more of the input parameters. Receiving the input parameter can be analogous to inputting the input parameter in that the input parameter is automatically entered after it is received.

[0049] The term "processing unit" generally refers to a component of a system, particularly an AI system, an MRI system, or a control unit, that is capable of performing calculations and processing. In this context, the processing unit refers to the unit responsible for generating the output parameter based on the input parameters.

[0050] The processing unit is designed to generate the output parameter based on the received input parameter. This is achieved by applying a generative AI model to the input parameter(s). This new data resembles the training data but is not identical to it. The generative AI model is an artificial intelligence model specifically capable of generating new data that depends on the input data. Here, "intelligence" refers to the AI ​​system's ability to generate an output parameter based on the input parameter, thus adjusting the media signal accordingly. "Reaction" refers to the fact that the generative AI system generates an output parameter in response to the input parameter.

[0051] Examples of generative AI models include Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), and / or autoregressive models. Generative AI models learn from a large amount of training data to understand the underlying probability distributions of the data. Once trained, the model can generate new data by drawing random samples from the learned distribution.

[0052] To generate the output parameter, the generative AI model can be trained to generate a media signal from scratch and / or to retrieve and adapt an existing media signal. The existing media signal can, for example, be retrieved from a database by the AI ​​model. In principle, the generated media signal can then be used as the basis for further generation. The database can be network-, internet-, or cloud-based.

[0053] The AI ​​model is configured such that the generated output parameter exhibits a dependency on the input parameter. This dependency can mean a similarity and / or a fit. In other words, the generated output parameter resembles the input parameter and / or is adapted to the input parameter. The dependency can affect only a part of the output parameter, specifically only one or a few characteristics. The term "dependency" here refers to the fact that the generated media signal depends on various factors, such as the adapted examination parameter and / or the patient.

[0054] The dependency can be set, in particular, according to an instruction parameter. The instruction parameter can be provided, for example, in addition to the adapted examination parameter. Alternatively or additionally, the AI ​​system can be preconfigured according to the instruction parameter. The instruction parameter specifically defines the dependency between the output parameter and the input parameter. Specifically, the instruction parameter can specify the degree of dependency and / or its characteristic component. Typically, the degree of dependency is greater than 0% and / or less than 100%. The instruction parameter can be set, for example, by the user and / or the patient.It is sometimes conceivable that the dependency is set according to several instruction parameters, where one instruction parameter is specified and / or set by the user and another instruction parameter is specified and / or set by the patient.

[0055] The generated media signal includes, in particular, an audio track and / or a video track. If the generated media signal includes both an audio track and a video track, the audio and video tracks are typically synchronized in time and / or content. The term "audio track" here refers to the acoustic component of the media signal. If the generated media signal does not include a video track, it is, in particular, a music signal. The term "video track" here refers to the visual component of the media signal. If the generated media signal does not include an audio track, it is, in particular, a silent film signal.

[0056] The output interface is designed to provide the generated output parameter. This output parameter specifically contains the results obtained by applying the generative AI model to the input parameter(s) by the AI ​​system.

[0057] Providing the adapted examination parameter via the control unit as an input parameter for the generative AI system can additionally include keeping the control unit and / or the MRI system ready to receive the generated media signal. This readiness to receive the signal can, in particular, include starting a shutdown period (timeout), during which the generated media signal is typically received by the MRI system. If the generated media signal is not received within the shutdown period, the control unit can, for example, provide the adapted examination parameter again as an input parameter for the generative AI system and / or output an error signal to the user.

[0058] The playback unit typically receives the generated media signal before the MRI scan and its playback. It is conceivable that the MRI scan has already begun when the playback unit receives the generated media signal. Playback of the received media signal can begin immediately after receiving the generated media signal, and / or at the start of the MRI scan, and / or after a predefined delay.

[0059] The control unit and / or the playback unit and / or the MRI system may have an interface for receiving the generated media signal. In particular, a media unit may have the interface for receiving the generated media signal. Specifically, the media unit and / or the control unit and / or the MRI system and / or the playback unit may receive the generated media signal. Receiving the generated media signal by the playback unit may involve receiving the generated media signal by the control unit and / or the MRI system and / or the media unit before receiving it by the playback unit. For example, the generated media signal may be transmitted from the output interface of the AI ​​system to the interface of the control unit and / or the playback unit and / or the MRI system and / or the media unit.It is conceivable that the interface of the control unit and / or the playback unit and / or the MRI system and / or the media unit retrieves the generated media signal from the output interface of the AI ​​system.

[0060] The interface can be, in particular, a network interface, especially an Ethernet, Wi-Fi, or mobile internet interface, and / or a data interface to which other units, especially the AI ​​system and / or a computing unit, have access. The interface can, in particular, be the same interface used for providing the adapted test parameter or a different interface.

[0061] The media unit can be part of the control unit and / or the playback unit and / or the MRI system. Alternatively, it is conceivable that the media unit is not part of the control unit and / or the playback unit and / or the MRI system, but rather an independent component or part of the AI ​​system, or is cloud-based or network-based. The media unit and / or the control unit and / or the MRI system can be configured to store and / or buffer and / or retrieve the generated media signal. In particular, the media unit can be configured to transmit and / or stream the generated media signal to the playback unit and / or play it back via the playback unit. The media unit can, in particular, include a media server which is configured to store and / or buffer and / or retrieve the generated media signal and / or stream and / or transmit it to the playback unit.

[0062] Streaming the generated media signal means, in particular, that the complete media signal is not transmitted to the media unit and / or the playback unit, but only a portion of the media signal sufficient for the current input, especially the current playback. Streaming enables faster and more efficient transmission and playback of the media signal without requiring its complete transmission. Streaming can occur in real time or with a slight delay. In this case, the streaming unit forms at least part of the playback unit.

[0063] Alternatively or additionally, the media unit can transmit the generated media signal completely to the playback unit. This means that the media signal is only played back once it has been completely transmitted from the media unit to the playback unit. Transmitting the media signal completely requires more time and / or storage space than streaming, but advantageously enables uninterrupted playback.

[0064] The playback unit and / or the media unit and / or the control unit and / or the MRI system can be configured to pause and / or start the playback of the received media signal. It is also conceivable that the playback unit and / or the media unit could be configured to play the received media signal in a continuous loop.

[0065] After the MRI system receives the generated media signal, the control unit can, in particular, issue a readiness signal to acknowledge receipt. If the MRI scan is feasible, the receipt of the generated media signal can, for example, determine an absolute time at which the MRI scan and / or the playback of the generated media signal begins. Specifically, the receipt can trigger the execution of the MRI scan and / or the playback of the received media signal.

[0066] The term "performing the MRI examination" here refers to the practical implementation of the MRI examination using the described steps and parameters according to the adjusted examination parameters. Performing the MRI examination of the patient specifically includes activating the magnet unit and / or at least one gradient coil and / or emitting the radio waves using the MRI system. Performing the MRI examination typically does not include moving the patient table.

[0067] Ideally, the MRI scan and the playback of the received media signal occur simultaneously and / or with maximum temporal overlap. In principle, it is conceivable that the MRI scan begins before the playback of the received media signal. Alternatively, the playback of the received media signal can begin before the MRI scan.

[0068] Preferably, the MRI scan ends simultaneously with the playback of the received media signal. In principle, it is conceivable that the MRI scan ends before the playback of the received media signal is complete. Alternatively, the playback of the received media signal can end before the MRI scan is complete.

[0069] The input of the received media signal by the playback unit specifically includes reproducing the received media signal using the playback unit. The input of the received media signal preferably starts at the beginning of the received media signal and / or ends at the end of the received media signal. In principle, it is conceivable that the input of the received media signal starts several seconds or, for example, minutes after the beginning of the received media signal and / or ends several seconds or, for example, minutes before the end of the received media signal. The "several seconds" are, in particular, a maximum of 10 seconds, preferably 2 seconds.

[0070] The playback process can include fading in the video track. It can also include playing back the audio track. Before or after playback, the video track is typically hidden and / or the audio track is muted.

[0071] The playback unit can include a projection and / or a display unit for displaying the video track. The projection unit can, for example, include a projector and / or a projection surface. The projection surface can be a screen and / or the inner wall of the MRI system bore.

[0072] The playback unit may include an audio unit for outputting the audio track. The audio unit may include a speaker and / or headphones.

[0073] The playback unit is specifically MRI-compatible. The playback unit is designed in such a way that the patient can perceive the display of the video track and / or the output of the audio track. Perception by the patient specifically means hearing the audio track and / or seeing the video track. The term "perception" here refers to the patient's acquisition and processing of the media signal.

[0074] The phrase "so that" is used in particular to emphasize the dependency of the specific generation of the output parameter on the respective input parameter. The AI ​​system is specifically designed or trained to take this dependency into account.

[0075] In one embodiment, the method further comprises the steps of: receiving patient-specific preference information from the patient via the control unit and providing the received patient-specific preference information as an additional input parameter, in addition to the adapted examination parameter, for the generative AI system, so that the generated media signal additionally depends on the patient-specific preference information. Providing the received preference information as an additional input parameter for the generative AI system means, in particular, that at least one input parameter is provided for the generative AI system, wherein the at least one input parameter comprises the adapted examination parameter and the patient-specific preference information.Incorporating patient-specific preference information allows for even greater personalization of the media signal, which can further enhance patient satisfaction and relaxation. By including the patient's preferences, the generative AI system can generate content specifically tailored to the patient's interests and needs, improving the effectiveness of distraction during the examination. Adapting the media signal to the patient's preferences can promote patient cooperation and thus reduce the likelihood of re-scans due to patient movement. The patient-specific preference information can be provided as an instruction parameter and / or entered into the AI ​​system to establish and / or modify the relationship between the adapted examination parameter and the generated media signal.The patient-specific preference information typically relates, in particular, only to the content of the media signal and / or not to the total duration of the media signal.

[0076] In one embodiment, the patient is provided with several media signal content types for selection, with the patient-specific preference information comprising the media signal content type selected by the patient. The selected media signal content type is typically limited to an instruction parameter that pertains solely to the content of the media signal. Media signal content types can be, for example, various music genres and / or film genres and / or pre-generated media signals. It is conceivable that the patient selects a pre-generated media signal, such as a video from an internet-based platform, as the media signal content type, with the pre-generated media signal being used as an instruction parameter in such a way that the generated media signal is based on the pre-generated media signal.In this case, the context of the generated media signal and the context of the pre-configured media signal are typically the same and / or similar, at least for a period of time. In other words, the generated media signal is essentially an adapted pre-configured media signal, with the adaptation depending on the adjusted examination parameter. For patient selection, the control unit can have an input device and / or be connected to an input device that the patient uses to select one of several media signal content types. Determining the patient's own preference information increases their engagement and control over the examination process, which can lead to greater satisfaction.Directly entering patient preferences can improve the accuracy of the captured information and thus increase the relevance of the generated media signal for the patient. Using an input method to capture preferences can make the media signal personalization process more efficient and reduce preparation time for the MRI scan.

[0077] In one embodiment, the adapted examination parameter is time-variable, so that the audio and / or video track of the generated media signal is rhythmically and / or harmonically and / or image-wise synchronized with the time-variable examination parameter by means of the AI ​​system. Synchronizing the audio and / or video track of the media signal with the examination parameter of the MRI scan can advantageously help the patient prepare for the examination and better follow the instructions. Synchronizing the media signal with the examination parameter can improve the coordination between the MRI system and the patient, leading to a more efficient examination.Synchronizing the media signal with the time-varying examination parameter can help reduce the number of necessary interactions between the user and the patient, thus decreasing the user's workload. Rhythmic synchronization means, in particular, that changes in acoustic level and / or frequency and / or image content are synchronized with the sound emissions of the MRI scan, specifically those encompassed by the examination parameter. Harmonic synchronization means, in particular, that the acoustic frequencies of the audio track are synchronized with the frequencies of the sound emissions of the MRI scan. Image content synchronization means, in particular, that the image content of the video track is synchronized with the intensity of the sound emissions and / or the remaining duration of the MRI scan. For example, the video track could show a person climbing a mountain.The generated media signal is created by the AI ​​system in such a way that the mountaineer reaches the summit when the MRI scan is finished. This synchronization can, in particular, mean that the remaining time of the MRI scan is translated into a narrative arc of suspense.

[0078] In one embodiment, the generated media signal comprises a video track, wherein the video track includes a graphic representation of a person interacting with the patient in such a way that, during the MRI examination, information is conveyed to the patient by the depicted person, depending on the examination parameters. The graphic representation of the person in the video track interacting with the patient can provide clear and understandable instructions, which can improve patient compliance during the examination. The provision of information by the depicted person, depending on the sequence parameters, can reduce the need for verbal communication and thus minimize disruption for the patient during the MRI examination.Visual guidance via video can improve the comprehension and adherence to instructions, particularly for patients who have difficulty with auditory instructions, thus contributing to higher image quality and a more efficient examination. The graphic representation of the person can be, in particular, an avatar. The interaction between the graphic representation and the patient can be one-sided, in that the graphic representation provides information that the patient can perceive, but not vice versa. Alternatively, the interaction can be structured so that the information is also played as part of the audio track, while the person is graphically represented as speaking, and / or the information is graphically represented through gestures and / or behaviors of the person, such as remaining still, holding their breath, staying still, and / or relaxing.

[0079] In one embodiment, the adapted examination parameter includes a repetition time, an echo time, and / or an echo spacing of the MRI scan, such that the generated media signal depends on the repetition time, the echo time, and / or the echo spacing. Specifying examination parameters such as repetition time, echo time, and / or echo spacing allows for precise control of image quality and / or contrast ratios, leading to improved diagnostic accuracy of the MRI images.

[0080] In one embodiment, the examination parameter includes one or more time points and / or a total or remaining duration for the patient to remain lying down, hold their breath, stay still, and / or relax, such that the generated media signal depends on the one or more time points and / or the total or remaining duration of the lying down, breath-holding, staying still, and / or relaxing. Defining one or more time points for lying down, breath-holding, staying still, and / or relaxing helps reduce motion artifacts, which improves image quality and diagnostic accuracy of the MRI examination. Coordinating patient actions during the MRI examination improves patient comfort and can reduce the need for repeat scans, saving time and costs.The term "remaining still" refers to the condition in which the patient should not get up from the examination table during the MRI scan, both before and after the scan. The term "breath-holding" refers to the condition in which the patient should hold their breath during the MRI scan to minimize motion artifacts. The term "remaining still" refers to the condition in which the patient should remain still during the MRI scan, particularly regardless of breath-holding, to minimize motion artifacts. The term "relaxing" refers to the condition in which the patient should typically remain relaxed and / or move freely during breaks between MRI sequences, as such movements do not affect image quality or the MRI scan itself.

[0081] In one embodiment, the examination parameter includes the total duration of the MRI scan, such that the total duration of the generated media signal essentially corresponds to the total duration of the MRI scan. In this case, the input of the received media signal preferably ends essentially at the same time as the end of the received media signal and the end of the MRI scan. "Essentially" in this context means, in particular, a deviation of up to 30 s, preferably 10 s, and most advantageously 2 s, between the absolute end times of the input media signal and the execution of the MRI scan. Advantageously, the absolute end time of the input media signal corresponds to the absolute end time of the execution of the MRI scan.It is particularly advantageous if the playback of the received media signal starts at the beginning of the received media signal and the start of the MRI scan, and ends at the end of the received media signal and the end of the MRI scan. Specifying a total MRI scan duration allows for precise planning and scheduling of examinations, leading to more efficient utilization of the MRI system. Coordinating the duration of the media signal with the scan duration can improve the patient experience by reducing the perceived length of the scan and by providing patients with better information and entertainment. This, in turn, reduces patient irritation. In particular, aligning the total durations leads to less dissatisfaction, especially for pediatric patients.

[0082] In one embodiment, the generated media signal comprises an audio track, wherein the examination parameter includes a gradient profile of at least one gradient coil of a magnet unit of the MRI system, such that the audio track of the generated media signal is rhythmically and / or harmonically synchronized with noise emissions of the gradient profile in such a way that the noise emissions are superimposed on or integrated into the played-in media signal. The noise emissions, in particular the frequency responses, can correspond to the Fourier transforms of the gradient profiles and are thus advantageously predictable to a certain extent or calculable in advance. The synchronization of the generated media signal with the noise emissions of the gradient profile can reduce the noise and / or stress burden for the patient by superimposing disturbing noises or integrating them into a harmonious soundscape.Improved patient acceptance through a more comfortable acoustic environment can reduce motion artifacts and increase image quality, contributing to more efficient diagnosis. The gradient profile is primarily determined by the repetition time, echo time, and / or echospacing of the MRI scan.

[0083] In one embodiment, after the MRI examination has begun, an examination parameter is adjusted again, the adjusted examination parameter being provided as an input parameter for the generative AI system, the generative AI system being configured to adapt the generated media signal to the adjusted examination parameter in response to input of the provided adjusted examination parameter, in order to generate and make available an adapted media signal, the adapted media signal being dependent on the adjusted examination parameter and based on the generated media signal, the playback of the adapted media signal as a replacement for the generated media signal continuing for the remainder of the MRI examination.The readjustment of the examination parameter after the start of the MRI scan and the corresponding adjustment of the media signal by the generative AI system enable a flexible response to changing examination conditions, thus improving image quality and examination efficiency. Providing an adapted media signal that depends on the readjusted examination parameter and therefore on the patient can personalize the patient experience and increase comfort during the examination, promoting patient cooperation and improving the quality of the results. The transition is particularly seamless, so the patient is ideally unaware that the adapted media signal is being played instead of the generated one. In other words, the switch from the generated to the adapted media signal is preferably as seamless as possible.Re-adjusting an examination parameter can involve a previously adjusted parameter or a different parameter. In particular, an examination parameter can be adjusted for an MRI sequence that has already begun but is not yet complete, or that is already complete, or that has not yet begun. Re-adjusting an examination parameter specifically includes repeating a completed MRI sequence. The fact that the adjusted media signal is based on the generated media signal means, in particular, that the adjusted media signal is generated from the content of the generated media signal, such that the context of the generated media signal persists in the adjusted media signal for at least a certain period of time, preferably several seconds, and preferably minutes.The context of the media signal relates in particular to a content section characterized by harmonies and / or rhythms and / or visual content. In contrast to the generation of the generated media signal, the content type of the media signal is typically not changed in the case of an adapted media signal.

[0084] According to the invention, at least one gradient coil of the MRI system is energized as a playback unit for playing back the audio track of the media signal during the MRI examination for at least a portion of the total duration of the generated media signal. In other words, the playback unit of the MRI system specifically comprises the at least one gradient coil. Using the gradient coil to play back the media signal can reduce the need for additional hardware, resulting in cost savings and a simplified system configuration. This embodiment is particularly advantageous when the generated media signal is rhythmically and / or harmonically synchronized with the noise emissions of the gradient. The "at least part" of the total duration can, in particular, refer to those time periods during which no signals from the MRI images need to be acquired.Alternatively or additionally, it is conceivable that at least one gradient coil is used as the playback unit over the entire duration of the media signal.

[0085] In one embodiment, the MRI system further incorporates the AI ​​system. Adding the AI ​​system to the MRI system can improve diagnostic accuracy through advanced image analysis and pattern recognition, particularly without requiring communication via an internet connection. The AI ​​system can contribute to the real-time optimization of imaging parameters, potentially leading to personalized patient examinations and improved image quality.

[0086] A computer program product according to the invention comprises program code means for executing the method according to the invention when the computer program product is executed in the computing unit.

[0087] The computer program enables simple and rapid implementation of the procedure into existing AI and / or MRI systems, facilitating updates and expansions of system functionality. Direct loading into the processing unit's memory ensures high compatibility and ease of use, promoting acceptance in clinical practice. Execution of the program code can lead to standardized application of the procedure, improving the reproducibility of results and the comparability of data across different sites.

[0088] The computer program product can be a computer program or comprise a computer program. The computer program product particularly includes the program code means that implement the process steps according to the invention. This allows the process according to the invention to be defined and executed repeatably, and enables control over the transfer of the process according to the invention. The computer program product is preferably configured such that the computing unit can execute the process steps according to the invention by means of the computer program product. The program code means can, in particular, be loaded into a memory of the computing unit and typically executed by means of a processor of the computing unit with access to the memory.When the computer program product, in particular the program code, is executed in the processing unit, all embodiments of the described method according to the invention can typically be carried out. The computer program product is, for example, stored on a physical, computer-readable medium and / or digitally stored as a data packet in a computer network. The computer program product can represent the physical, computer-readable medium and / or the data packet in the computer network. Thus, the invention can also start from the physical, computer-readable medium and / or the data packet in the computer network. The physical, computer-readable medium is usually directly connectable to the processing unit, for example, by inserting the physical, computer-readable medium into a DVD drive or plugging it into a USB port, thereby allowing the processing unit to access the physical, computer-readable medium, particularly for reading.The data packet can preferably be retrieved from the computer network. The computer network can contain the computing unit itself or be indirectly connected to the computing unit via a wide-area network (WAN) or a (wireless) local area network (WLAN or LAN) connection. For example, the computer program product can be stored digitally on a cloud server at a storage location within the computer network and transferred to the computing unit via the WAN over the internet and / or via WLAN or LAN, particularly by accessing a download link that points to the storage location of the computer program product.

[0089] Features, advantages, or alternative embodiments mentioned in the description of the device are also transferable to the method, and vice versa. In other words, claims relating to the method can be further developed with features of the device, and vice versa. In particular, the device according to the invention can be used in the method.

[0090] Regardless of the grammatical gender of a particular term, persons with male, female or other gender identities are included. Exemplary embodiments of the drawing

[0091] The invention will now be described and explained in more detail with reference to the exemplary embodiments shown in the figures. In principle, structures and units that remain essentially the same in the following figure descriptions will be named with the same reference numeral as when the respective structure or unit first appeared.

[0092] They show: Fig. 1 a method according to the invention for performing an MRI examination of a patient using an MRI system, Fig. 2 a first embodiment of the method, Fig. 3 a second embodiment of the method, Fig. 4 a third embodiment of the method, Fig. 5 a fourth embodiment of the method, Fig. 6 a fifth embodiment of the method, Fig. 7 a sixth embodiment of the method, Fig. 8 a seventh embodiment of the method, Fig. 9 a method according to the invention for providing a media signal by means of an AI system, Fig. 10 an AI system according to the invention and Fig. 11 an MRI system according to the invention.

[0093] Fig. Figure 1 shows a method according to the invention for performing an MRI examination of a patient using an MRI system in a flowchart comprising steps S100 to S104: Procedure step S100 describes the adjustment of an examination parameter of the MRI examination to the patient using a control unit of the MRI system.

[0094] The adapted examination parameter can specify a repetition time, an echo time, and / or an echospacing for the MRI examination, such that the generated media signal depends on the repetition time, the echo time, and / or the echospacing. Alternatively or additionally, the examination parameter can specify one or more time points and / or a total or remaining duration for the patient to remain lying down, hold their breath, remain still, and / or relax, so that the generated media signal depends on one or more time points and / or a total or remaining duration for remaining lying down, holding their breath, remaining still, and / or relaxing. Procedure step S101 identifies the provision of the adapted examination parameter by the control unit as an input parameter for a generative AI system, wherein the generative AI system is configured toIn response to input of the provided input parameter, to generate a media signal as an output parameter and to make the generated media signal available for retrieval, wherein the generated media signal depends on the adapted investigation parameter and includes an audio track and / or a video track.

[0095] Process step S102 characterizes the reception of the generated media signal by means of a playback unit.

[0096] Procedure step S103 characterizes the performance of the MRI examination of the patient according to the adapted MRI examination parameter using the MRI system in such a way that, during this time, according to procedure step S104, the received media signal is played back for perception by the patient using the playback unit.

[0097] Fig. Figure 2 shows a flowchart of a first embodiment of the method.

[0098] Procedure step S105 indicates that the patient is provided with several media signal content types for selection by the patient and that patient-specific preference information includes the media signal content type selected by the patient.

[0099] Procedure step S106 marks the receipt of the patient's individual preference information by the control unit.

[0100] Procedure step S107 identifies the provision of the received patient-specific preference information as a further input parameter in addition to the adapted examination parameter for the generative AI system, so that the generated media signal additionally depends on the patient-specific preference information.

[0101] Fig. Figure 3 shows a flowchart of a second embodiment of the method.

[0102] Procedure step S108 indicates that the adapted examination parameter is time-variable, so that the audio track and / or the video track of the generated media signal are rhythmically and / or harmonically and / or image-content synchronized with the time-variable examination parameter by means of the AI ​​system.

[0103] Fig. Figure 4 shows a flowchart of a third embodiment of the method.

[0104] Procedure step S109 indicates that the generated media signal includes a video track, wherein the video track includes a graphic representation of a person for such interaction with the patient that information is fed to the patient during the MRI examination by means of the depicted person depending on the examination parameter.

[0105] Fig. Figure 5 shows a flowchart of a fourth embodiment of the method.

[0106] Procedure step S110 indicates that the adapted examination parameter encompasses a total duration of the MRI examination, so that the total duration of the media signal essentially corresponds to the total duration of the MRI examination.

[0107] Fig. Figure 6 shows a flowchart of a fifth embodiment of the method.

[0108] Procedure step S111 indicates that the generated media signal includes an audio track, wherein the adapted examination parameter includes a gradient profile of at least one gradient coil of a magnet unit of the MRI system, such that the audio track of the generated media signal is rhythmically and / or harmonically synchronized with noise emissions of the gradient profile in such a way that the noise emissions are superimposed or integrated with the played-in media signal.

[0109] Fig. Figure 7 shows a flowchart of a sixth embodiment of the method.

[0110] Procedure step S112 indicates that the generated media signal includes an audio track, wherein at least one gradient coil of the MRI system is energized as a playback unit for playing the audio track of the media signal during the MRI examination for at least part of the total duration of the generated media signal.

[0111] Fig. Figure 8 shows a flowchart of a seventh embodiment of the method.

[0112] Procedure step S113 marks the start of the MRI examination.

[0113] Procedure step S114 marks the beginning of the playback of the received media signal during this process.

[0114] Procedure step S115 indicates that an examination parameter is adjusted again after the MRI examination has begun.

[0115] Procedure step S116 indicates that the re-adjusted investigation parameter is provided as an input parameter for the generative AI system, wherein the generative AI system is configured to adapt the generated media signal to the re-adjusted investigation parameter in response to an input of the provided re-adjusted investigation parameter in order to generate and provide an adapted media signal for retrieval, wherein the adapted media signal depends on the re-adjusted investigation parameter and is based on the generated media signal.

[0116] Procedure step S117 indicates that the playback with the adapted media signal as a replacement for the generated media signal continues during the remaining duration of the MRI examination.

[0117] Procedure step S118 marks the end of the playback of the received media signal and / or the performance of the MRI examination.

[0118] Fig. Figure 9 shows a flowchart of a procedure for providing a media signal using an AI system.

[0119] Procedure step S200 identifies the receipt of an examination parameter from a patient's MRI examination as an input parameter.

[0120] Procedure step S201 characterizes the generation of a media signal as an output parameter in response to an input parameter.

[0121] Procedure step S202 characterizes the provision of the generated media signal for retrieval, wherein the generated media signal depends on the adapted examination parameter and includes an audio track and / or a video track.

[0122] Fig. Figure 10 shows a block diagram of an AI system according to the invention.

[0123] The AI ​​system 10 has an input interface 11 for receiving an input parameter. The AI ​​system 10 also has a computing unit 12 for generating an output parameter based on the received input parameter by applying a generative AI model to the input parameter. The AI ​​system 10 additionally includes an output interface 13 for providing the generated output parameter. The AI ​​system 10 is particularly suitable for executing a procedure according to Fig. 9 trained, especially coached.

[0124] Fig. Figure 11 shows a schematic representation of an MRI system 100 according to the invention.

[0125] The MRI system 100 comprises a magnet unit 101. The magnet unit 101 includes a main magnet 102 and at least one gradient coil 103. The MRI system 100 further comprises a control unit 105 and, in this embodiment, optionally a display unit 104.

[0126] The playback unit 104 includes headphones for perceiving an audio track of the generated media signal. The headphones are worn by patient P. The playback unit 104 also includes a display unit for perceiving a video track of the generated media signal.

[0127] Although the invention has been illustrated and described in detail by the preferred embodiments, the invention is nevertheless not limited by the disclosed examples and other variations can be derived by the person skilled in the art without leaving the scope of protection of the invention.

Claims

[1] Procedure for performing an MRI examination of a patient using an MRI system, comprising the steps: - Adjusting an examination parameter of the MRI examination to the patient using a control unit of the MRI system, - Providing the adapted investigation parameter via the control unit as an input parameter for a generative AI system, wherein the generative AI system is configured to generate a media signal as an output parameter in response to an input of the provided input parameter and to make the generated media signal available for retrieval, wherein the generated media signal depends on the adapted investigation parameter and includes an audio track and / or a video track, - Receiving the generated media signal by means of a playback unit, - Performing the MRI examination of the patient according to the adapted MRI examination parameters using the MRI system in such a way that the received media signal is played back to the patient via the playback unit during the examination, characterized by , - that the generated media signal includes an audio track, wherein at least one gradient coil of the MRI system is powered as a playback unit for playing the audio track of the media signal during the performance of the MRI examination for at least part of the total duration of the generated media signal. [2] The method of claim 1, further comprising: - Receiving patient-specific preference information via the control unit, - Providing the received patient-specific preference information as an additional input parameter in addition to the adapted examination parameter for the generative AI system, so that the generated media signal additionally depends on the patient-specific preference information. [3] Method according to claim 2, wherein several media signal content types are provided to the patient for selection by the patient and the patient-specific preference information includes the media signal content type selected by the patient. [4] Method according to one of the preceding claims, wherein the adapted investigation parameter is time-variable, so that the audio track and / or the video track of the generated media signal is rhythmically and / or harmonically and / or image-content synchronized with the time-variable investigation parameter by means of the AI ​​system. [5] Method according to one of the preceding claims, wherein the generated media signal comprises a video track, wherein the video track comprises a graphic representation of a person for such interaction with the patient that information is fed to the patient during the MRI examination by means of the depicted person depending on the examination parameter. [6] Method according to one of the preceding claims, wherein the adapted examination parameter comprises a repetition time, an echo time and / or an echospacing of the MRI examination, such that the generated media signal depends on the repetition time and / or the echo time and / or the echospacing. [7] Method according to any of the preceding claims, wherein the examination parameter comprises one or more time points and / or a total or remaining time for the patient to remain lying down, hold their breath, remain still and / or relax, such that the generated media signal depends on the one or more time points and / or the total or remaining time for the patient to remain lying down, hold their breath, remain still and / or relax. [8] Method according to one of the preceding claims, wherein the examination parameter comprises a total duration of the MRI examination, such that the total duration of the media signal substantially corresponds to the total duration of the MRI examination. [9] Method according to one of the preceding claims, wherein the generated media signal comprises an audio track, wherein the adapted examination parameter comprises a gradient profile of at least one gradient coil of a magnet unit of the MRI system, such that the audio track of the generated media signal is rhythmically and / or harmonically synchronized with noise emissions of the gradient profile in such a way that the noise emissions are superimposed or integrated with the played-in media signal. [10] A method according to one of the preceding claims, wherein after the start of the MRI examination, an examination parameter is adjusted again, wherein the adjusted examination parameter is provided as an input parameter for the generative AI system, wherein the generative AI system is configured to adapt the generated media signal to the adjusted examination parameter in response to an input of the provided adjusted examination parameter in order to generate an adapted media signal and make it available for retrieval, wherein the adapted media signal depends on the adjusted examination parameter and is based on the generated media signal, wherein the playback with the adapted media signal as a replacement for the generated media signal continues during the remaining duration of the MRI examination. [11] MRI system (100), having - a magnetic unit (101) comprising a principal magnet (102) and at least one gradient coil (103), - a control unit (105) - and a playback unit (104), wherein the MRI system (100) is configured to perform a method according to any one of claims 1 to 10. [12] MRI system (100) according to claim 11, further comprising an AI system (10), wherein the AI ​​system (10) has an input interface (11) for receiving an input parameter, a computing unit (12) for generating an output parameter depending on the received input parameter by applying a generative AI model to the input parameter, has an output interface (13) for providing the generated output parameter, wherein the AI ​​system (10) is trained, in particular trained, to perform the following steps: - Receiving an examination parameter from a patient's MRI examination as an input parameter, - Generating a media signal as an output parameter in response to an input parameter, - Providing the generated media signal for retrieval, wherein the generated media signal depends on the adapted investigation parameter and includes an audio track and / or a video track. [13] Computer program product, comprising program code means for executing a method according to any one of claims 1 to 10 when the computer program product is executed in a computing unit.

Images