Magnetomotive Displacement of the Tympanic Membrane Using Magnetic Nanoparticles: Toward Enhancement of Sound Perception

Abstract

Objective: A novel hearing-aid scheme using magnetomotive nanoparticles (MNPs) as transducers in the tympanic membrane (TM) is proposed, aiming to noninvasively and directly induce a modulated vibration on the TM.

Methods: In this feasibility study, iron oxide (Fe₃O₄) nanoparticles were applied on ex vivo rat TM tissues and allowed to diffuse over ∼2 h. Subsequently, magnetic force was exerted on the MNP-laden TM via a programmable electromagnetic solenoid to induce the magnetomotion. Optical coherence tomography (OCT), along with its phase-sensitive measurement capabilities, was utilized to visualize and quantify the nanometer-scale vibrations generated on the TM tissues.

Results: The magnetomotive displacements induced on the TM were significantly greater than the baseline vibration of the TM without MNPs. In addition to a pure frequency tone, a chirped excitation and the corresponding spectroscopic response were also successfully generated and obtained. Finally, visualization of volumetric TM dynamics was achieved.

Conclusion: This study demonstrates the effectiveness of magnetically inducing vibrations on TMs containing iron oxide nanoparticles, manipulating the amplitude and the frequency of the induced TM motions, and the capability of assessing the magnetomotive dynamics via OCT.

Significance: The results demonstrated here suggest the potential use of this noninvasive magnetomotive approach in future hearing aid applications. OCT can be utilized to investigate the magnetomotive dynamics of the TM, which may either enhance sound perception or magnetically induce the perception of sound without the need for acoustic speech signals.

Fig. 1

Ear anatomy and the hearing principle. The sound waves (S) travel through the ear canal (EC) to the middle ear where they vibrate the ear drum/tympanic membrane (TM) and the auditory ossicles (AO). These vibrations reach the inner ear, where the hair cells inside the cochlea (C) convert the vibration to neural signals so that the brain can interpret sound.

Fig. 2

Schematic of the magnetomotive OCT (MM-OCT) system. A superluminescent diode (center wavelength ~1310 nm) produces near-infrared light that is sent through a single-mode fiber, where a 2×2 fiber coupler (FC) splits the light into two beams. One beam travels to the reference arm, which is composed of a fixed mirror (M); the other beam passes through the sample arm to the TM tissue sample. The backscattered light from both the sample and reference arms interfere, and the resulting interference pattern is detected by the spectrometer. A magnetic solenoid (MS) is placed in the sample arm, where the light beam passes through the bore of the MS to the tissue sample. The alternating magnetic field (AMF) is generated by applying different driving voltage waveforms to the MS with a programmable power supply. Other optical components in the system include the collimator (C), the achromatic lens (L), and the galvanometer scanner (G).

Fig. 3

Validation of the correlation between the modulated frequency applied (f_d) and the dominant frequency detected (f_m) from the 10 mg/ml MNP-laden TM tissues (N = 12–16 for each frequency value). The dashed line denotes the linear fit of the median f_d. For both (a) and (b), the “+” symbols shown beyond the whisker regions indicate the outliers (defined with an outlier coefficient of 3).

Fig. 4

Representative data of the TM samples applied with (I) 10 mg/ml MNP, (II) 5 mg/ml MNP, and (III) PBS solution (control). (a) Microscopic images (100x) of the representative histologic slices of TM tissues with iron-oxide staining. Indicated by the arrows, clusters of iron oxide MNPs (blue) are observed on the TM tissues (pink). (b) Representative structural OCT images and (c) the corresponding MM-OCT images of the TM samples that were mechanically perturbed around their resonance frequencies. In (c), structural intensity (red) is overlaid with the MM-displacement (green). (d) Frequency analysis of the three representative groups shows the existence of dominant frequencies in the mechanical spectra of the MNP-laden samples. Since the resonance frequency of each TM tissue sample is not the same, the dominant frequencies detected differ as well. Abbreviations: Outer ear (OE), middle ear (ME), and malleus (M).

Fig. 5

Box plots of the MM-displacement amplitude of the rat TM samples applied with PBS (control), 5 mg/ml MNP, and 10 mg/ml MNP, where the total number of data points are N = 12, 12, 16, respectively. The median values of each group are indicated along the boxes. The symbols “*” and “**” denote p-values < 10⁻⁴ and 10⁻⁵, respectively.

Fig. 6

Representative datasets showing the spectroscopic response of the MNP-laden TM tissue. After applying a chirped excitation (10–500 Hz) to the TM tissue, both (a) the spectrogram and (b) the mechanical spectrum showed that the largest mechanical vibration appeared around ~350 Hz. The spectroscopic response agrees with the trend shown in (c) the B-mode MM-OCT data, which show the superimposition of the structural image (red) and the spatial mapping of the MM-displacement (green) induced with various modulated frequencies. The scale bars represent ~200 μm.

Fig. 7

Spatiotemporal visualization of the MM dynamics of an ex vivo rat TM tissue laden with MNPs. A sinusoidal magnetic force (460 Hz) was applied to the TM, where the representative frame shows the TM moving toward the middle ear directions. The corresponding vibration-amplified 4D-OCT dataset covering the TM motion of one entire sinusoidal cycle is provided in Media 1. The scale bars in each direction (x, y, and z) represent ~200 μm.