Action Potential Final Report
Team Action Potential: Modeling the Electrical Currents in a Neuron
There were several steps that we needed to complete for our project: the first was to decide how to construct our model and attach the LED lights. There were several options, but we decided to go for the most straightforward one, which entailed constructing the model in SolidWorks with holes the size of LED lights. Joseph modeled the neuron, using mostly the spline tool and the extrusion tool in SolidWorks, and Kathryn helped while also planning the lighting effects and creating a MATLAB code to program them. We included every main element of a neuron: the dendrites, the cell body, with a nucleus, two ribosomes and two endoplasmic reticulums, and the axon, which in turn carried the axon terminals. We placed the lights in a way that, through MATLAB coding, we could animate the action potential process by activating the LEDs in a particular region of the cell in the appropriate order: dendrites, nucleus, axon, axon terminals. We ran into some trouble with the proportions of the model in SolidWorks, mainly because of the multiple scalings that we used, so we could not attach the LED lights to the model itself. To deal with this issue we picked out a piece of wood to mount our model on, and the staff in the Wood shop kindly drilled holes for the lights at locations we had previously marked. After this Joseph strung the ground wires together, attached to the shorter wire on each LED light, while Kathryn soldered the E5 Shield to the live wires and, in turn, to the longer wire on each LED light. Kathryn tweaked the program for each set of lights after the physical model had been completed and mounted to the board, and after that step was finished, our project was complete.
-Kathryn is very interested in Biology, so she had the idea to model a brain cell in Solidworks. Joseph thought the model would be a fun challenge, so that’s what we agreed on. Professor Cheever suggested that we use LED lights to animate the action potential of the cell and he explained to us how we should go about it, a process that we outlined in the abstract. Our goals were to construct a three-dimensional model of a neuron on SolidWorks, print it, and attach LED lights in the holes we had included. As we explained in the abstract, however, we could not include the cuts that would make the holes for the LED lights large enough, so our new goal became to attach the model to a board and let the LEDs shine through.
-This project was interesting and engaging because we were taking something as incredibly complex as the neuron, simplifying it, and essentially mimicking the way that the body transmits thought and emotion. We were also pleased to learn some basic electrical engineering techniques as well when attaching the LEDs, and found it to be a very enriching experience. We were motivated by our clearly defined goals and our interest in both the idea behind the project and the work the project involved.
Neurons are the cells that direct thought, senses, and feeling in our bodies.The membrane barrier surrounding the cell keeps concentrations of certain ions (salts: potassium+ and sodium+)at very specific levels both outside, and inside the fluid cytoplasm of the cell. Since the ions have charges, this results in an electrical potential between the inside and outside of the membrane. When the nerve is stimulated (by pressure, chemicals, or other interactions) the cell responds by changing those ion concentrations in a very specific way; the concentrations change such that the normal negative (resting) potential is switched to a positive potential in the area that received the stimulus. This causes the surrounding areas to react by temporarily switching potential as well; overall a chain reaction of negative-to-positive potential changes travels from the dendrites (structures receiving input) to the axon terminals (structures that communicate to other cells). (Sadava 2006)In this project, we used LED lights to simulate this electrical signal traveling down the cell.
The LED lights we used are pictured below (source Prof. Cheever pers. comm.).
Each LED has a ground pin and a power pin that had to be correctly wired to the E5 Shield. The ground pins were connected together and led to the ground hole on the E5 Shield. All of the LEDs were assigned power pin numbers that corresponded to where they plugged in on the E5 Shield. Those numbers were used to sequence the LEDs using matlab code. The LEDs were sequenced to first show insufficient stimulation in the dendrites (3 LEDs turn on and off at half brightness), then full signaling/action potential, as all the lights on the cell light up in order.
The matlab code was fairly simple (Appendix 1); it mainly operates using for-loops to separate various sections of the neuron (e.g. pins 11,12, and 7 are the axon terminal lights). Vectors were constructed for the necessary pin numbers, then the for-loop ran the vector's values through either the a.analogWrite function or the a.digitalWrite function. The a.analogWrite function allows a pulse-count input, which varies the brightness on the specified LED. The a.digitalWrite funtion has simpler inputs, and simply turns the number-specified LED on or off.
Completed Project Design
-Our project, as previously stated, includes a roughly seven-inch-by-five-inch three-dimensional model of a neuron, with LED lights programmed to illustrate the passage of an electrical impulse. The model’s foundation is a sketch in SolidWorks, for which we employed the spline function to make the appearance more organic and true to our subject, and the extrusion function to create a three-dimensional, printable model out of our sketch. The sketch was initially too small, so we scaled it to five times its size, which led to an unfortunate string of scaling that left us confused as to how to create enough material for the 3.9-millimeter cuts that our LED lights required. The relation of the original sketch to the three-dimensional model was unintelligible, but the project was saved through the kind actions of the wood-shop staff. Our printed model served as our reference for the circles we drew on our block of wood, which would then serve as the holes for the insertion of the LED lights. The holes that the kind man in the wood shop drilled were 3.9 millimeters in diameter on the bottom side of the wood, and only 2 millimeters on the top so that the LED lights could easily fit without coming through to the top side. We used tacky material to attach the model to the block of wood, and inserted the LEDs into the bottom drill-holes. We used five red LEDs at the holes corresponding to the dendrite and nucleus sections of the cell to illustrate unprocessed electrical impulses. For the axon and axon terminals we used green LEDs so that viewers could see that these impulses are different, i.e. strong enough to be transmitted. We used 3-millimeter LEDs for our project, as we have referenced before. These lights have two main parts: the light itself and the two wires attached behind it. The light itself is 3 millimeters in diameter but its base is 3.85 millimeters. Out of the base, the longer wire to protrude is called the anode and the shorter is the cathode. These are the wires we previously referred to as the “longer” and “shorter” wire, respectively. After we had inserted the LEDs into the bottom side of the wood, the wires were still exposed. Joseph attached a ground wire, or a negatively-charged electron conductor, to the cathodes of each light, carefully winding the ground wire around the cathode using a special tool. He then attached each cathode to the next cathode in the string of LED lights so that the entire display was connected by a negatively-charged conductor. This completed about half of the electrical work needed for the project.
-Kathryn completed the positively-charged portion of the electrical work. She first soldered a new top to the normal E5 Shield, attaching resistor pins to a circuit board with number-designated output pins. There were a total of 13 places to wire our LEDs to; our final design used 11. An inherent property of the E5Shield was that pins 3,5,6,9,10, and 11 were compatible with the a.analogWrite matlab function described above, but the rest of the pins were not. This meant that we attached the dendrite-region LEDs (the ones we wanted to vary the brightness of) to aforementioned pins, and the rest of the cell body LEDs could be assigned random pin numbers. Once all the soldering was done, and the code written, all that remained was testing it out!
Ultimately, we were able to model a neuron in Solidworks (if not quite to the scale we had planned), manipulate the necessary circuits and wiring to perform our simulation. The code worked well, and after one glitch of a malfunctioning LED, all the wiring proved to be working well. The final demonstration of our project had its flaws however - most importantly, the somewhat ineffective lighting. The simulation of the action potential traveling down the cell as demonstrated by sequenced LEDs was not as easily visible as we had hoped; our printed model blocked to much of the light. Below are some video demonstrations of the LED neuron in action, as well as pictures of our final design.
-Our project was not an experiment, but it was definitely a learning-and-growing experience for both of us. It forced us to learn more about SolidWorks and MATLAB, and to find alternate solutions when our original plan did not work out very well. If we could do the project over again from the start we would have planned for the block of wood to be our base the whole time, allowing us to drill 3.9 millimeter holes through all of the wood. This would allow us to cut holes in the model only 1 millimeter in diameter, which we could have fit in the model if it were only slightly bigger. That said, another issue we would have to fix is the confusion that resulted from scaling multiple times. Faced with the same situation we encountered then in the future, if we were unhappy with the first scale we would simply erase the action using the project manager toolbar, setting only one relation between the sketch and the model. One other improvement would be to hollow out most of the cell body using the shell tool. We are happy with the way that the lights were animated; we just wish they could shine through more and in a more-lit room.
-One thing we could try to do if we worked further with this idea is to create individual parts for each part of the cell, connected by hinges so that the cell can bend and move.
1. Matlab Code
Sadava, D., Heller, H.C., Orians, G.H., Purves, W.K., Hillis, D.M. 2006. Life: The Science of Biology. 8th ed. Sunderland, MA: W. H. Freeman. Print. 1266