Abstract
This thesis describes efforts to establish a biophysical model which corresponds to the real physiological structure of the auditory system. By using this model, a lot of phenomena in true human hearing can be simulated. Based on the models of the middle ear to the cochlea and outer hair cells (Liu and Neely, 2009, 2010), we integrate several auditory models (Meddis, 1986; Sumner et al., 2002; Hewitt et al., 1992) and a new model－a tuberculoventral(TUB) cell model with delayed, frequency-specific inhibition－to construct the auditory pathway from middle ear to the brainstem. In the cochlear nucleus, TUB cells and T-multipolar cells are distributed tonotopically. In other words, every TUB cell and T-multipolar cell in different place has its own best resonance frequency. The T-multipolar cell can be inhibited by TUB cell which has the same best resonance frequency. If the acoustic stimulation is a pure tone of low frequency, the corresponding auditory nerve fibers will fire regularly and generate a special effect called phase locking. Because TUB cells and T-multipolar cells receive input from the same group of auditory nerve fibers, the delayed, frequency-specific inhibition of TUB cells can suppress the stable triggering in the phase locking cycle of T-multipolar cells. Since the input of medial olivocochlear(MOC) interneurons are from T-multipolars, this inhibition can lower the firing rate of MOC interneurons and cause the masking effect in the cochlea to reduce indirectly. If the acoustic stimulation is irregular background noise, there is no phase locking effect, so the TUB cells can not inhibit the T-multipolars effectively. Therefore the masking effect in the cochlea from MOC interneurons will be unaffected. The discrepancy of the inhibition from TUB cells can cause different masking intensity between background noise and pure tone, so the unmasking effect in the cochlea from MOC interneurons can be simulated.