The sky is no longer the limit to what machines can do in service to humans, ranging from examining a rock on Mars to intricate medical procedures on humans, themselves. Now Artificial Intelligence AI has managed to estimate the possibility of a heart attack or strokes in a subject. The system was more precise in prediction than Human specialists.
Doctors have to study data from patients thoroughly to estimate the possibility of a stroke in patients. However, AI system, using medical record data, predicted 355 cases correctly on top of what doctors predicted. 355 more lives saved by timely predicting the disease, reported Futurism.
The cause of this leap of advancement is the ability of AI system to analyze a huge set of data, and being able to identify risks in patients that would otherwise be marked healthy. The Science Mag covered the research performed by Nottingham University Scientists, who were able to impart the wisdom to the AI system. Between 2005-2015, stats from nearly 400,000 people were collected and then 75% of the data was handed to the AI system. The factors that were a risk to the heart were examined. Later, the rest of the 25% data was fed to evaluate the generative algorithm of the AI system. IEEE Spectrum claims that where doctors, under the current guidelines of medicine, could predict 72% patients correctly, the AI predicted 76% correct predictions. This 4% makes around 15 thousand patients that only the AI system was able to identify.
Among the 7,404 heart attack cases, 4998 patients were correctly predicted by AI System, 355 more cases than doctors would have. Looking more closely at these results regarding lives saved, the margin does get a little bit closer. There was a total of 7,404 actual cases of heart attack and stroke in the data, with the AI system accurately predicting 4,998 patients. The traditional method of diagnosis would have found 355 fewer cases, resulting in what is essentially a theoretical value of 355 lives saved through the AI.
The challenge at this point is to introduce more and more accuracy to the algorithm. There may be a time when 4% becomes 40% and we hope it does become much more accurate.