The Experimental Investigation And Finite Element Analysis On Efficiency Of Driving The Cochlea In Forward And Reverse Way

The middle ear implantable hearing devices have been used to overcome conductive hearing loss induced by middle ear pathologies or disorders. The device with forward driving (FD) commonly converts sound into the mechanical vibrations of the ossicular chain, which is transmitted into the cochlea through the oval window (OW). However, coupling the transducer to ossicles is difficult in the ear with middle ear diseases, such as the cholesteatoma, congenital aural atresia, or ossicular malformations, or with unsuccessful middle ear surgery. The implantable device with reverse driving (RD) provides an alternative way of coupling sound to cochlea by driving the round window membrane (RWM). According to the reports in literature, the middle ear implantable device with RD is able to drive the cochlea and induce the cochlear potentials, auditory brainstem responses, auditory-nerve potentials and differential pressure across the cochlear partition, which were comparable to that with FD by sound in the ear canal. Clinically, there have been numerous reports on the study and application of using the implantable device with RD. According to their results, improved pure-tone threshold or speech recognition scores after implanting the RD devices have also been achieved for almost all the patients.However, stimulation from the round window (RW) is not a physiologically sound transmission pathway. In FD, the tympanic membrane and ossicular chain transfer the acoustic pressure into cochlea at the oval window. The vibration of the stapes footplate serves as the input to cochlea, induces the wave propagation through the perilymph in scala vestibuli and scala tympani, and stimulates the basilar membrane (BM) vibration. In RD, the transducer directly stimulates the vibration of RWM, which is transmitted into the scala tympani and BM. Because of the asymmetrical internal structures of the cochlea and the different boundaries at the OW and RW, the efficiency of RD-induced BM vibration should not be the same as that for FD with the implant located on the ossicular chain or OW. In this study, we compared the efficiency of BM vibration between FD and RD in two ways---The experimental investigation and finite element analysis.In animal experimental, eight Hartley guinea pigs were used and laser Doppler vibrometer (LDV) was employed to measure the vibration of the targets including BM, incus tip and magnet bead. After anesthesia of the animal we accessed to the bulla through a postauricular dissection, drilled a hole on the cochlea bony wall, and exposed the basal turn of the BM. The glass microbead (30um) was placed on the center of the BM as the laser-reflecting target. The holes made on the cochlea and bulla were sealed with glass sheets. In forward driving,80dB SPL pure tones over the frequency range of1-40kHz from a function generator were presented to the ear canal near the eardrum by an insert earphone. In reverse driving, a small magnet (0.625mm in height,0.75mm in diameter) was placed on the middle ear side of the RWM and driven by a coil. The distance between the magnet and coil was2.0-2.5mm. The vibrations at the BM’s basal turn and the incus tip were measured with LDV under two driving conditions. The vibration at the RWM was also measured in RD. The efficiency of RD was compared with that of FD based on cochlear gain, i.e., the ratio of BM vibration to cochlear input, consisting of RWM vibration for RD and incus vibration for FD.In finite element (FE) analysis, a published human ear model was used. The FD was simulated in the model by presenting90dB SPL in the ear canal at2mm away from the tympanic membrane (TM), while the RD was simulated by driving a transducer placed on the lateral surface of the RWM. Two types of transducer were used in FE analysis:Reverse-1with large transducer:7mg in weight,1.2mm in length, and1mm2in cross-section area; Reverse-2with small transducer:2.2mg in weight,1.2mm in length, and0.314mm2in cross-section area. The area ratio of the large transducer to the RWM was about0.5, which was close to that used in the guinea pig experiments. Driving force with amplitude of0.05mN was applied onto the transducers along the normal direction of RWM for both transducers. The harmonic analysis over the auditory frequency range of1kHz-10kHz was conducted in these three models using Ansys. In FD, the magnitudes of the displacements at the central point of TM, footplate, RW and BM were picked up, and the vertical parts (along the normal direction) were calculated respectively. In RD, the magnitudes of the displacements at the central point of TM, footplate, transducer and BM were picked up, and the vertical parts were calculated respectively. And then, the cochlear gain in FD and RD as a function of stimulus frequency was derived by normalizing the BM displacement with respect to cochlear input. In FD, the displacement of the footplate was used to represent the cochlear input. In RD, the transducer transmitted vibration into cochlea via RWM, and thus the cochlear input was represented by the displacement of the transducer.The results demonstrated that electromagnetic coupling on the RWM can generate an effective BM vibration. The characteristic frequencies of BM vibration at the basal turn ranged between13and15kHz in both driving directions. FD had higher cochlear gain than the RD. Moreover, in RD, the bigger transducer will generate lower cochlear input impedance and induce higher cochlear gain. To the authors’ knowledge, this is the first study to measure BM vibration in RD. Thus, the present findings provide useful information for understanding the RD mechanism and efficiency when RD of the RWM is used for designing middle ear implantable devices.