Originally computers were used to help humans to complete their tasks and improve their works. In general, they were based on computing skills and some others. But at the beginning of the 21st century and with the technological improvement, the developers decided to provide computers more privileges which lead to a new technology that allow those machines to learn by themselves and gain an intelligence that called artificial intelligence(AI) and could be compared to an expert human brain because the way that machines learn is similar to human brain functionality, it’s learning from past experiences and mistakes.
That means the machine with more experience should be more efficient. With this new technology, scientists start to find ways using it in many fields but the most important one was medicine. Today AI is widely used in medicine and no one could deny how much it added more improvements to such field. The real issue is how people can trust a machine that advances them instead of their doctors especially when it’s about their health. First of all, medical research takes a lot of time and very costly in many cases and especially when this research …8is about finding a cure for an urgent and specific disease. So, the best solution was using the new technology AI which could be more efficient and sometimes can achieve the same goal as normal research with less time, cost, and more precise. An example was giving (2018) by The Medical Futurist about an AI called Atomwise was made to decrease Ebola infectivity.
In 2015 this AI and by learning, analyzing, and studying a big data about molecular structure resulted in giving an accurate study which solves the problem within one day. On the contrary, a typical research would take months and maybe years by following old ways in medical research. More training is the key for a successful AI. Also, like humans, training means the ability to gain more experience an be more professional and expert. In fact, Hsieh (2017) wrote about an AI that studies hundreds of x-ray images that belong to patients that have Tuberculosis and some of them don’t have.
After the study, the result was an accurate diagnosis that reached 96% and with more training and studying algorithm improvements the 4% error could be reduced to less than that. Also, another AI that uses advanced learning algorithm can predict heart attacks 7.6% better than normal methods followed by doctors and these results was based only on studying about 300000 patient’s profiles. The accuracy can’t reach 100% in many cases but decreasing the error is the most important to give diagnosis compared to expert doctors which in many cases AI could be exact and detailed better than doctors.