In recent years, computers have had a spurt in popularity and in utility. These devices have gone from being able to only compute simple calculations and perform basic algorithms, to having the capabilities to analyse databases containing an indeterminable amount of information in a matter of seconds. The world as we know it thrives off computers; the most renowned companies in the world such as Google, Apple and Amazon rely on computing technology to maintain and organize all of their important data. Generally, computer programs are written manually by a user and tend to be very linear in terms of operation, this kind of programming is known as functional programming. The software will simply follow the commands, line by line, that the user had given it as input. On the other hand, there is an artificial intelligence, also known as machine learning or artificial neural networks. Instead of being restricted by the commands given by the programmer, the application will learn how to analyze data on its own. All the program needs is a general goal and a dataset that it can use to train itself with. Then based on the information that it had gathered, it makes an educated guess on what it believes the answer to the given problem is. That is a very generalized definition of machine learning, but it is a proper idea of what the software does. Engineers have been able to create systems that teach themselves how to play games like poker, blackjack, rock-paper-scissors and even video games. Although, artificial intelligence is not just used for games, people have been able to create applications that have educated themselves on real world scientific dilemmas, such as, the laws of physics by only using underlying data. One of the more recent revelations for machine learning has been it's conceivable potential in the medical world. Comparatively to many other industries, artificial intelligence has yet to have any kind of large impact on the progression of technology when it comes to this specific field. There are thousands of documents and reports that have been written regarding the application of machine learning to medical research although not many people have been able to use this information to create anything of great importance. Despite not being prominent in this area of analysis, there are a few successful projects that are used by doctors, cardiologists, radiologists, and many other professions around the world on a day-to-day basis.
	An example of machine learning being used for medical purposes is for the detection of diabetic retinopathy and macular edema; the former being a diabetic related issue caused by “a common and specific microvascular complication of diabetes, and remains the leading cause of preventable blindness in working-aged people.” (Cheung et al., 2010) and the latter being the formation of sores inside the macula (Bakri, 2016). Both of these diseases are eye related issues and, in the long run, are known to cause blindness. Considering these illnesses are very similar, the methods for diagnosing each of them are almost identical, making it difficult for doctors to pinpoint exactly which one they should treat the patient for. This is where machine learning comes into practical use. A team of researchers developed a neural network that was able to accurately identify these issues apart from each other:
	"[the] neural network was trained using a retrospective development data set of 128 175
retinal images, which were graded 3 to 7 times for diabetic retinopathy, diabetic macular edema, and image gradability by a panel of 54 US licensed ophthalmologists and ophthalmology senior residents" (Gulshan et al., 2016).
This specific system had a vast dataset to work with so it was able to train effectively. Once the team had finalized the software, they tested it on two completely unrelated sets of images compared to the one that had been originally used for training to see if it could accurately identify each disease. The first dataset contained 9963 images whilst the other had 1748. Each base respectively scored an accuracy of 93.4% and 93.9% (Gulshan et al., 2016). These numbers are quite accurate but not quite satisfactory enough so that the system could be used for clinical use. Regardless of its current benefit, there is always room for improvement. The neat thing about machine learning is that there is no limit to a neural networks potential. The software can always be trained to further extent by adding more information to its training dataset or by optimizing the small amount of code written by the user.
	Machine learning could also be implemented into radiology, and this would be a major technological breakthrough for specialists in the field. There are already thousands of extensive databases containing x-ray imagery and data. These sets of images are mostly formed of radiographs of various broken or fractured bone structures. There is an organization named Enlitic, currently based in Silicon Valley, who have created an artificial model that analyzes given radiographs and returns as output what type of damage had been done to the cartilage and displayed the exact location of where such damage had been done: “Using deep learning, a refined version of artificial neural networks, the computer developed rules that not only identified radiographs with fractures but highlighted the fractures.” (Jha et al., 2016). This is an example of a great medical discovery that has the potential to change the way professionals in the field conduct their everyday work. Generally, radiologists spend a large amount of time identifying the specific type of wound the patient is dealing with. This is very counterproductive considering the duration of discovering a remedy for peculiar cases may take a while. Admittedly, this system is indeed a neural network, therefore the accuracy of the software in its current state may not be satisfying enough for it to be considered for practical use. But, as mentioned previously in the case of the diabetic retinopathy network, new developments concerning artificial intelligence are being refreshed every single day.
	Machine learning is an extraordinary tool that has so much potential for pragmatic use in the medical field. Although, there are specific reasons as to why this technology is behind in development and has not been implemented clinically yet. In the case of