Just when you thought Google couldn’t offer any more radical surprises, it comes up with an AI that has started playing the role of the prophet of death. The company has apparently developed an AI that has the ability to predict a person’s death with incredible accuracy.
The death predicting AI
Google’s new AI is aimed for use in hospitals as a tool to determine how long a patient will live in cases of serious medical conditions. Based on the initial trials, the accuracy of the AI in predicting the death of the patients has been 95 percent.
The AI uses a large amount of patient data to arrive at its results. This includes everything from the gender, age, medical history, patient records, etc. The algorithm scans through thousands of such variables and then outputs the time of death. And it is the ability to crunch a large amount of patient data that enables the AI to be as accurate as it is.
“In general, prior work has focused on a subset of features available in the EHR [electronic health record], rather than on all data available in a HER… which includes clinical free-text notes, as well as large amounts of structured and semi-structured data”, IFL Science quotes the people who worked on this project.
In the trails, the AI was 86 percent successful in accurately predicting the length of time a patient would stay in the hospital. Traditional methods only had an accuracy of 76 percent. Inpatient mortality prediction was accurate up to 95 percent of the time for the AI compared to only 86 percent using the traditional methodology. Plus, the AI predicted unexpected patient readmissions with 77 percent accuracy when the existing methods could only hit 70 percent accuracy.
Google’s massive health strategy
The death predicting AI is not the only focus area of Google in the healthcare industry. In fact, the tech giant already has many projects in place aimed at making the company a global leader in health technologies.
Google’s Medical Brain has already developed AIs aimed at niche medical fields like cardiology, radiology, and ophthalmology. They are also reportedly working on an AI for the dermatology industry. In India, the company is testing out an AI that analyzes eye images to determine the onset of a condition called diabetic retinopathy.
Industry experts explain that Google could license these AI systems and start selling them to hospitals and clinics very soon. The company might also package the AI as a service and deliver it through the cloud, which is what is more likely given that Google is already a major player in the cloud industry.
In addition to Google, a major tech company who is betting high on healthcare AI is IBM. In collaboration with the Cleveland Clinic Lerner College of Medicine of Case Western Reserve University, IBM has launched a project called WatsonPaths.
The AI algorithm of WatsonPaths is expected to help physicians make faster and more accurate decisions regarding patient treatment. In addition, the AI also enables extraction of new insights from the patient’s electronic medical records that were hitherto not possible or too time-consuming for an average human.
Pharmaceutical companies are also investing in AI that will help them cut down the drug development process from decades to just a few years. This will not only ensure cheaper drugs, but also faster medications for newer diseases.