Physiological Acoustics Lab

I worked in the physiological acoustics lab from June 2005 through May 2006 under the supervision of Dr. Monty Escabi. Observed and assisted in cat and rat surgery. Wrote MATLab application for sound experiment on human subjects. Carried out experiment and analyzed data to draw conclusions.

The study began in June of 2005 and it was first decided that I should observe the procedures of the laboratory for a couple of weeks to get some experience with the protocols and the house rules. For these weeks I observed in vivo experiments performed on cats and rats. The primary experiments were to observe the organization of spectrotemporal receptive fields in the auditory midbrain. The investigators were trying to study how inferior colliculus (IC) plays an important role in translating and consolidating auditory information in the transition to the auditory thalamus. It was suggested that the receptive field in the IC is systematic in a way that it reflects the patterns of the innervating afferents and the local collicular circuitry. I was able to observe the experimental procedure and later data acquisition and analysis through Matlab. These first few weeks gave me a good background on proper protocol, good practice, experimental techniques, and how to properly analyze the data.

First it had to be determined that the cat has a clean outer ear canal and a clean middle ear (fluid free, as determined by a otoscopic inspection) canal for it to qualify as a valid test subject. The cat was not allowed to eat the night prior to the experiment. Then the anesthetization took place, which proved to be a very important step due to the loss of one or two cats. A dose of Ketamine hydrochloride and acepromazine was inject intramuscularly. In order for an intravenous solution to be used with the cat, the fore-arm was shaved and wiped with betadine solution. Sometimes the vein was difficult to find, so an incision is made to the surrounding skin to expose the vein, which makes it easier to now insert the needle. The angiocatheter is then inserted into the vein and taped loosely to the arm. The cat is typically kept anesthetized with a mixture of sodium pentobarbital.

Next I was able to observe a cat tracheotomy that is used to reduce breathing noise that could possibly interfere with the auditory recordings. A small incision is made to the intercartilaginous area of the trachea and the muscles are detached to expose the trachea. Then an endotracheal tube is inserted and is sutured up to help contain it. The entire procedure is performed while the cat is under anesthetic. Now the cat is ready for the brain surgery.

The cat is put into a three-point orbito-spatial headholder that is made specifically for cats. The head-holder can be adjusted in the dorsal, anterior and lateral directions to a very specific amount. The scalp is then injected with a drug in order to prep the area to be cut. Various connective tissue and muscles are then removed in order to expose the bone. Holes are then put in to the skull and the appropriate parts are removed with bone cutters. To expose the midbrain and the brainstem, the cortical matter and dura that are dorsal were aspirated.

I also learned good experimental practice by watching people who had years of experience in this field. It is generally a good idea to keep the entire area clean, especially the tools that are used for the procedure. The dirty surgical tools can pose the risk of infection and ruin the entire experiment. However, some tools are electronic and too sensitive to be sterilized and since the procedure is run in a non-sterile environment it is not possible for everything to be completely sterile. This concludes the preparatory phase of the experiment.

The cat’s vital statistics were monitored and recorded to make sure that it is in good health. Statistics such as the heart rate, breathing rate, temperature and reflexes were recorded throughout the experiment. If any of the following were observed, the cat was removed from the experiment: excessive edema to the brain, severe infection, excessively low heart rate, damage to brain structures, anatomical abnormalities, or deafness. The anesthesia is adjusted accordingly. The animal is maintained in an anesthetized state through Ketamine and Diazepam. It was also injected with various drugs to prevent brain edema and excessive salivation. The experiments focused specifically on spectral processing, binaural processing, and temporal processing.

Another portion of the independent study involved the development of a program that could be used for running sound experiments on human test subjects. The program integrates Matlab and the TDT (Tucker-Davis Technologies) so that the experiment could be run in a sound-proof chamber. The objective of the program was to determine the threshold of the minimum detectable sampling rate of various sounds. To do so, it was necessary that the program take a directory of sound files, be able to play two sound files consecutively, and then wait for a button press but the subject. The program consists of a TDT file which is opened once and runs continuously and can take in an array which is a sound, play it, and also record and store a button press from the RP2. The Matlab portion consists of three separate functions: one to open and run the TDT file, another to send a file to the TDT to be played, and another to acquire the button press. The first method returns the RP variable which allows the other functions to take in this variable and use it to get parameter tags and set parameter tags. Since all these functions are separated, they can be called multiple times, or be put into data loops and invoked as needed.

Many problems were encountered along the way. The biggest was getting the TDT part to store the button press so that Matlab could just call GetTagVal and return the value easily. The TDT wanted to reset the value to zero. Another problem was getting the sound to play properly. The TDT example was designed to play the sound continuously, so it was necessary to devise a way to check at what point the TDT was in the buffer and detect whether it was on the second loop or not. Since it was possible to know at what point it is at in the buffer, it was possible to solve this problem. The way to get around this was to have Matlab have two checks of the position of the buffer. The first check should always be greater than the second, unless it is on the second loop. So when this is detected, in a while loop, the TDT would terminate playing the sound. Playing and stopping the sound was performed using soft triggers.

The program is at the point now where it can be used for many experiments that involve playing sounds through the TDT and/or require acquiring a button press. Since these methods are functionalized in Matlab, it is simple to invoke any function as desired with the appropriate variables to be passed in.