Artificial Intelligence (AI) is seen as the real boon for the twenty-first century regardless of any field it has its application in. Defining AI in simple words is the creation of intelligent machines that respond and function just like humans. Talking about its possible and implemented applications among tech, education, gaming, retail, domestic appliances, defense, banking, finance, and many others; healthcare has always been on the quest for discovering something extraordinary and moreover, something that aid mankind to find the optimum solution for the lethal diseases. In this context, Indian-origin doctors have successfully deployed AI in the quick detection of brain hemorrhage. It was declared on 24th of October 2019 by the authors of the study conducted to develop an AI system by teams from UC San Francisco and UC Berkeley.
Researchers for this purpose have molded AI so as to accurately detect even the most complex form of these hemorrhage cases. This will not only enable early detection of the brain hemorrhage but will also save lives in the smallest fragments of time. This AI system is so swift that it detects the blood flow in just a few seconds and moreover, delivers the much needed accuracy that otherwise is a major issue in hemorrhage detection. The detection becomes even more complex when there is a very tiny hemorrhage and yet, crucial to affect one’s health. Without a doubt, this high precision detection with quicker results will prove to be a major discovery in healthcare. According to Jitendra Malik, Ph.D., with UC Berkeley and one of the co-authors of the study, “Given the large number of people who suffer from traumatic brain injury every day and are rushed to the emergency department, this has very big clinical importance. That convinced me to work on this problem.”
The primary reason reported behind the sharp accuracy of this AI system is the training data that has been embedded in the algorithm. It is quite a known fact that better the quality of added data, the easier for the system to determine a pattern in the algorithm. This is how machine learning works. As far as this study is concerned, the data is contained of a convolutional neural network with more than 4,396 CT scans. Thus, these numerous inputs with special attention to details (as in, highlighting the abnormalities at a pixel level!) will allow AI not to be distracted by any distortion in the image so as to be confused by the actual hemorrhage. This is the reason for its much needed accuracy which lies all in the details. Adding more to its accuracy, the scanning was not conducted all at once but in parts that reduce the possible chances of any deviation or error.
