modeling

I've been hard at work at the questions I addressed in my last blog. So far I've figured out a few things...

some assumptions of the CRF model:
The model is linear, and therefore assumes validity only in the low-level linear area near the threshold of hearing. This is interesting because I am actually testing the model using a broad range of intensity levels, meaning much of what I measure will not be fully incorporated into the current model. I'm going to have to try to add this into the flowchart of the model somehow.

The model assumes that there is spatial dependence- in other words the cochlea has a frequency/space mapping. This is a pretty standard component of any cochlear model, but it means that I have to figure out the wave equation. I'm planning to do some major thinking and researching this weekend. Rob gave us a model he is working on, and I plan to use it as a starting point. Hopefully by next Friday I will have made significant strides.

The model assumes a long-wave approximation. Shera tested this assumption in 2005, and I'm in the process of working through his conclusions. The basic conclusion, however, is that this is not a terribly off assumption.

The model assumes the mechanical properties of the cochlear partition vary irregularly with position. This is consistent with previous arguments.

I've also been working through the equations dealing with R. This is somewhat related to the other project we have been working on in middle ear (which is another blog topic altogether). I've been looking at the SFOAE profile v. frequency and looking at the regular oscillations present. It's pretty obvious to me now that these are standing waves, and they are coming from the cochlea/stapes area. Yes, it all comes back to impedance.

I've found a few more papers to read and will update on my progress again soon.

I also have a far better understanding of the opposing research. Siegel used a "suppressor" tone far basal from the stimulus in order to eliminate the base of the cochlea as a contribution source. He used a tone far enough away from the area of measurement that the suppressor isn't truly suppressing, but it's doing something. I have my suspicions about what it's doing, and I've added a new test to my protocol. I really can't wait to get back into the booth to put all of this to the test.

My goal for this weekend is to put all of this together, and do a little Matlab coding to test out my ideas.