Introduction

The use of magnetic resonance imaging (MRI) has grown steadily over the past years¹, making it the imaging modality of choice for the diagnosis of a large number of pathologies². In parallel, the use of implantable hearing systems has been adopted as a treatment modality for several different forms of disabling hearing loss. As with other active implantable devices, these auditory implants contain conductive and/or magnetic materials that could lead to mutual interference with electromagnetic fields present during scanning. Depending on the magnitude of these mutual interferences, the patient can be exposed to a number of risks.
Manufacturers of implantable devices are required to perform several tests to demonstrate their device’s behavior during an MRI examination³. Based on the results of these tests, the manufacturer can indicate if and under which conditions their device can be used in MRI in the device labelling⁴. Industry standards used in these tests describe general risks that apply to all (active) implantable devices³, but device specific hazards are not described in these documents. The presented work describes the design of a miniature vibrometer for use in MRI, that will be used to quantify the risk of unintended stimulation for patients with acoustically stimulating auditory implants. Unintended stimulation during MRI is a known phenomenon^5,6 for these patients which needs to be quantified to prevent harm.

Materials and methods

A prototype MRI-safe vibrometer has been designed which uses light with a wavelength of 1300 nm and is a next iteration of another vibrometer developed by our group working with light of 650 nm wavelength. This setup has been characterized on the bench and is being upgraded to increase the available optical power and with it the signal to noise ratio. The updated vibrometer is coupled to single-mode optical fibers which guide the light from the control room into the MRI suite, keeping all sensitive measurement equipment outside of the MR environment. Inside the MR suite, the distal end of the fiber, a miniature collimator is used. It is aiming the light onto the tip of a middle ear actuator (Cochlear Ltd., Sydney, AU) positioned inside the scanner bore. Vibrations induced by the switching gradient magnetic field or electromagnetic RF field can be detected by measuring the changes in light intensity reflected on the actuator tip. By changing the actuator position and orientation inside the scanner and altering the scanner pulse sequences, typical and worst-case scenarios regarding unintentional output can be determined.
The functionality of this prototype was assessed by inducing vibrations with a known amplitude and frequency while simultaneously measuring the vibrometer response. Vibrations were induced using a shaker (LDS V201, Bruel&Kjaer, Nærum, DK) and the magnitude of the vibrations was measured using a commercial laser Doppler vibrometer system (OFV-534, Polytec GmbH, Waldbronn, DE). The setup was mounted on a vibration isolation station (M-VIS3048, Newport Spectra-Physics, Utrecht, NL) to reduce the influence of environmental noise on the measurements. Data was recorded using an external soundcard (Fireface UC, RME Audio AG, Haimhausen, DE) at a sampling frequency of 48 kHz and analysed in Matlab (The Mathworks, Natick, MA, USA). Vibrations were induced at 22 logarithmically spaced frequencies between 0.1 and 10 kHz.
The obtained transfer function is intended to be used to calibrate actuator tip vibration signals induced by interactions with the electromagnetic fields present during MRI scanning (Ingenia 1.5T, Philips Healthcare, Best, NL). In parallel, a surgical feasibility study is ongoing to implant this novel miniature collimator in a human temporal bone. This will allow measuring the displacements of the actuator when mounted in a realistic environment.

Results and discussion

The previous version vibrometer allows measuring vibrations above a noise floor of 0.2-9.2 µm/s in function of the signal frequency. Assuming that these vibrations are transferred directly to the stapes footplate without any amplification, these values can be converted to estimated sound pressure levels of 48-86 dB SPL⁷. Induced sounds that could provide harm to the patient can therefore be measured using this setup. It is expected that the prototype under development will outperform these results.
Surgical feasibility work indicates that the miniature collimator can be implanted in a human temporal bone using the fixation system of a different middle ear implant (Codacs, Cochlear Ltd, Sydney, AU). The light emitted from the collimator can easily be aimed at the ossicles and the implanted actuator in contact with the ossicles.

Conclusion

Preliminary measurements using a previous generation vibrometer indicate that the system can measure unintended acoustic output. It is expected that the prototype under development will outperform this device both on the bench and in-situ in a temporal bone.

Acknowledgements

No acknowledgement found.