Our program could have an innumerable amount of development variables in the ShouldIInfect function, each which would influence cancer cell growth in a different way. The ShouldIInfect function should definitely be expanded if we wished to make our program further more accurate and complex. Also, our current modeling screen is relatively small, so we could certainly enlarge the area that the cancer is allowed to grow in. In doing this however, we would be using up more system resources and this could possibly cause a slower computer to crash. So, if we did decide to enlarge the scope of the cancer modeling area, we should also insert code into our program that would write certain important statistics to file. In these statistics we would include, the number of cancerous cells versus the number of healthy cells, how any sweeps had been executed, and possibly the location of the most recently infected cells, to gain an idea of where the cancer was in relation to the rest of the graph.
Our results are an accurate representation of cancer movement but a very simplified version. So from here we would expand on our program rather than change the procedures because they work adequately. When we believed the number of variable factored in to the infect process were sufficient, it would be possible to make our program even more complex and closer to reality by modeling the movement of cancer through the lymphatic channels. Our current program only shows local growth but in actuality cancer cells travel through the lymphatic channels and infect other parts of the body. This would of course be something extra to add to the program, only in its final stages of development.
Yet another thing that we could work on if we had time would be a more advanced output system. It could be a 3-D representation in which the user would be able to travel through a gas-looking cloud that would be the skin. In it would be another color that represented the cancer growth. This would much more accurately show the skin and give an even better perception of what was really happening. Of course this would also be a complex and therefore system resource-intensive part of our program. Therefore, if we add many of these sorts of factors, soon we will have to test our program on a much more powerful computer that the home PCs that we have been testing it on.
