Synopsis

Control of imaging gradient |dB/dt| rms is desired to limit potential gradient induced heating of medical implants. Current controls for |dB/dt| rms have been implemented on a per sequence basis. In this abstract we show that controlling gradient |dB/dt| rms on a per sequence basis may be overly conservative for certain examinations containing sequences with transient durations of high |dB/dt| rms. As such there is the potential to unnecessarily restrict clinical sequences with acceptable device heating. Alternative means for controlling |dB/dt| rms, such as calculating |dB/dt| rms using a rolling average across sequences, should therefore be considered.

Introduction

For the assessment of MR conditional safety of medical device implants, there are many hazards to consider. One hazard, identified in ISO/TS 10974, is gradient induced device heating. The imaging gradient dB/dt field induces eddy currents on conductive surfaces of an Active Implantable Medical Device (AIMD) resulting in device heating. The device heating caused by gradient field induced eddy currents is proportional to the square of |dB/dt| rms¹.

MR manufactures have therefore worked to develop means to control the imaging gradient |dB/dt| rms. Per IEC 60601-2-33, control of |dB/dt| rms can be implemented based on numerical evaluation of the gradient output for each individual sequence²:

$$\mid\frac{dB}{dt}\mid_{rms}=\mid\frac{dB}{dt}\mid_{peak}\sqrt{\text{slew percentage}}$$

While calculating |dB/dt| rms on a per sequence basis is an effective means of controlling gradient output, it may be overly restrictive as many scan sequences with high |dB/dt| rms are relatively short in comparison to the thermal time constants of AMID’s containing conductive surfaces large enough to generate gradient induced device heating.

In this abstract we therefore explore the use of a rolling average as an alternative method for controlling |dB/dt| rms. This is a similar approach to the one used by MR manufactures for controlling RF induced heating of patients, where SAR limits are defined for a 6 minute averaging time as opposed to per sequence.

Methods

A patient examination containing a series of scan sequences, including high |dB/dt| rms sequences, was simulated with the maximum |dB/dt| rms for each sequence, calculated per IEC 60601-2-33, shown.

A 3 minute rolling average |dB/dt| rms was then calculated across the patient examination where |dB/dt| rms was averaged over the preceding 3 minute time period using the |dB/dt| rms for each sequence.

The gradient induced device heating for a representative AIMD, an implantable cardioverter defibrillator (ICD), was then calculated for the patient examination where the temperature response ($$$\triangle T$$$) of the ICD was modeled using a first order system:

$$\triangle T=\triangle T_{max}\ \left(1-e^{-t/\tau}\right)$$

Where $$$t$$$ is time, $$$\tau$$$ is the thermal time constant of the ICD (equal to 8 minutes), and $$$\triangle T_{max}$$$ is the maximum temperature rise of the ICD for the applied |dB/dt| rms. The $$$\triangle T_{max}$$$ used for calculating device heating for any given sequence was scaled according to the following equation¹:

$$\triangle T_{max}=\triangle T_{\text{overall max}}\left[\frac{\mid\frac{dB}{dt}\mid_{rms}}{\mid\frac{dB}{dt}\mid_{\text{rms overall max}}}\right]^{2}$$

Where $$$\mid\frac{dB}{dt}\mid_{\text{rms overall max}}$$$ is the maximum |dB/dt| rms of the clinical sequences within the patient examination, and $$$\triangle T_{\text{overall max}}$$$ is the maximum possible device heating when $$$\mid\frac{dB}{dt}\mid_{\text{rms overall max}}$$$ is applied. $$$\triangle T_{\text{overall max}}$$$ was set equal to 1 in order to normalize the temperature rise calculations.

For comparison purposes, the temperature rise of the ICD, using the model described above, was calculated for continuous exposure to the maximum |dB/dt| rms of the clinical sequences (which represents the clinical sequence control limit) and for continuous exposure to the maximum 3 minute rolling average |dB/dt| rms (which represent the rolling average control limit).

Results

The 3 minute rolling average |dB/dt| rms was calculate for a simulated patient examination made up of clinical sequences. The results are shown in Figure 1. As can be seen, the maximum 3 minute rolling average |dB/dt| rms (42T/s rms) is significantly less than the maximum |dB/dt| rms for the clinical sequences (61T/s rms).

The temperature rise of an ICD as a result of this same clinical examination was calculated and is shown in Figure 2.

These heating results were then compared to the temperature rise of the ICD from continuous exposure to the maximum |dB/dt| rms of the clinical sequences (i.e. the sequence control limit) and for continuous exposure to the maximum 3 minute rolling average |dB/dt| rms (i.e. the rolling average control limit), as shown in Figure 3.

As can be seen in Figure 3, the heating of the device from the clinical examination was less than half of the sequence control limit, demonstrating that the sequence control limit can be overly conservative. Alternatively, the maximum device heating from the clinical examination matched closer to the rolling average control limit while still remaining below the rolling average control limit.

Discussion/Conclusions

Acknowledgements

References

1. ISO/TS 10974, 2018, "Assessment of the safety of magnetic resonance imaging for patients with an active implantable medical device" ISO, Geneva, Switzerland, www.astm.org.

2. IEC 60601-2-33, 2015, “Particular requirements for the basic safety and essential performance of magnetic resonance equipment for medical diagnosis” International Electrotechnical Commission (IEC), Geneva, Switzerland, www.iec.ch.