I was thinking about how to accurately negotiate an electrode with respect to a human brain, to minimise damage when accessing internal structures, then I started thinking about ultrasonic imaging and a way to generate a 3D model of the brain, or the internal structure of any solid object, by means of an array of piezoelectric transducers attached to the exterior of the object, where the transducers take turns as transmitters and receivers and map out the density variations within the object through acoustic transmission, reflection and refraction at interfaces between materials with different acoustic impedances.
I also thought about ways to train an artificial neural net as a signal processor to make sense of the data from the transducers and construct a 3D model from it, which got me onto thinking about how to construct a simulation of sound propagation in heterogeneous solids to use as a training set generator, and also about constructing a physical apparatus for the same purpose.
I thought about electrical impedance too, and of using it to build a 3D model of the electrical conductivity variations in the brain through similar techniques to using acoustic impedance for a 3D density model.
