3 Ways AI is Changing the Medical Industry
Artificial Intelligence will never eliminate the need for doctors however has the ability to improve the health outcome and the quality of life for millions of people in the upcoming years. Implementing AI into the healthcare system means new methods of diagnosis, more accurate results, and the proper allocation of resources. Here are a few ways that AI is currently being used to better serve patients around the world.
Turning Selfies into Diagnostic Tools
Our everyday consumer-grade products have the ability to produce extremely high-quality photographs, photographs which are applicable for analysis by AI algorithms. The combination of smartphones and AI can play a huge role in the support of clinical quality imaging. Currently, researchers in the UK have created an algorithm which has the ability to analyze a photo of a child’s face and deliver a diagnosis of potential developmental diseases. The algorithm works by detecting peculiarities that might indicate an abnormality in the bones of the skull and face and is currently able to match images to over 90 disorders.
Predicting if a Patient Will Awake from a Coma
When a patient enters a coma, doctors perform a series of assessments and assign them a score out of 23. This score represents the likelihood that a patient will recover and often plays a significant role in decisions regarding life support. A research institution in China has developed a machine to better determine the likeliness of recovery. The AI system is able to detect neural activities which are too numerous and sophisticated to be visible to the human eye, while also uncovering unknown patterns from past cases.
Recently a 19-year-old patient suffering severe injury to his temple was given 7 points after multiple rounds of assessment, a score which would legally allow removal from life support. However, the system trained on brain scans analyzed the same data gave him a score of 20, and like the system’s prediction, the patient eventually recovered from the coma. The machine can also examine regions in charge of different functions such as verbal capacity, hearing and motion control to see how they interact with each other after physical damage. The AI system was able to get almost 90% of assessments correct and since has helped diagnose over 300 patients.
Detecting Skin Cancer
A new artificial intelligence system is able to diagnose skin cancer more accurately than doctors. Researches from France, Germany, and the United States taught a computer the difference between dangerous melanomas and harmless ones showing it over 100,000 images. The computer was tested against 58 dermatologists from 17 countries, each with over 5 years of experience. The doctors were able to detect 86.6% of diagnosis accurately, compared to a 95% accuracy from the AI system, which missed fewer melanomas and misdiagnosed fewer benign moles. The improved accuracy would mean fewer patients undergoing unnecessary treatment reducing overall wait times for those who need it and an easier, faster diagnosis enabling surgery before cancer spreads.