A novel method for the simultaneous acquisition of auditory transient and steady state responses

This dissertation describes a novel method, designed for the simultaneous acquisition of Auditory Steady State Responses (ASSRs) and Auditory Transient Responses (ATRs). The method is based on the recently introduced Continuous Loop Averaging Deconvolution (CLAD) method (Delgado and Ozdamar, 2004). Specially tailored CLAD sequences were generated while ensuring equal spacing between all stimuli, except one slightly jittered stimulus to generate near steady state conditions. The nonisochronous stimulus was used to eliminate the singularity problem of the perfect steady-state stimuli, allowing the recovery of the ATR. A new technique named Multi Sequence Synthesis (MSS) that uses a variety of sequences to eliminate the noise amplification problems caused by the deconvolution process was designed and tested. Selected sequences were used to evoke Quasi Steady-State Responses (QSSR), and ATRs were extracted using the MSS method. A new method was also developed to construct the ASSRs from the QSSRs. This method computes the effect of the nonisochronous stimulus and constructs the real steady state response using the ATR. The idea is to subtract the effect of the response generated by the jittered stimulus from the raw recording and to add the effect of a virtual response that occurs at the steady-state enforcing stimulus location. The above mentioned methods and various sequence types were evaluated with simulations and real recordings from six subjects using click stimuli. It was found that the MSS method extends the usage of the CLAD technique by eliminating the limitations on sequences caused by the deconvolution process. Results showed that the proposed methods are capable of constructing both the ATR and ASSR information using the QSSRs. The threshold detections yielded similar results for the QSSRs and the MSS constructed recordings, showing the possibility of using a single QSSR with a properly selected sequence. Such sequences may be generated by eliminating the current limits on the data acquisition length and sampling rate. The method developed in this study may be applied to clinical use, utilizing the steady state information for hearing screening while using the transient response information for detecting neurological problems