Google has developed an artificial intelligence algorithm that could predict when you’ll die with up to 95 percent accuracy, according to the tech giant’s researchers.
The research, which tackles a range of clinical issues among hospital patients, was recently published in the journal Nature. Google applied artificial intelligence to a vast amount of data from more than 216,000 adult patients hospitalized for at least 24 hours each in two medical centers.
The research tapped into data from Electronic Health Records.
“We were interested in understanding whether deep learning could produce valid predictions across wide range of clinical problems and outcomes,” researchers explain, in the journal. “We therefore selected outcomes from divergent domains, including an important clinical outcome (death), a standard measure of quality of care (readmissions), a measure of resource utilization (length of stay), and a measure of understanding of a patient’s problems (diagnoses).”
The proof-of-concept study found that the algorithm could accurately predict risk of mortality, hospital readmission, prolonged hospital stay and discharge diagnosis. “In all cases, the method proved more accurate than previously published models,” it said.
The AI was 95 percent accurate at predicting patient mortality based on data from the University of California, San Francisco health system and 93 percent accurate using data from the University of Chicago Medicine system, according to the research.
This was significantly more accurate than the traditional predictive model, the augmented Early Warning Score, which uses a number of factors to help physicians determine how ill a patient is. This method was 85 percent accurate at the University of California, San Francisco system and 83 percent accurate at the University of Chicago Medicine system, the study says.
The Google research comes at a time when the potential benefits and risks of applying artificial intelligence are being hotly debated. From cybersecurity risks and so-called ‘doomsday’ machines that could wreak devastation to the technology’s potential to drive economic growth, experts are weighing the possible long-term impact of AI.
Healthcare, which relies on a bewildering array of information, is increasingly being touted as a good fit for utilizing AI. FDA Commissioner Scott Gottlieb, for example, discussed the promise of AI in a speech delivered earlier this year.
The technology, however, also poses major challenges.
Speaking on Fox and Friends Tuesday, Family Medicine Physician Dr. Mikhail Varshavski said that, while connecting vast quantities of health information could be beneficial for patients, data privacy is key. “The thing that is worrying for me is what happens with this data and who owns this data?” he said. “I hope, as a doctor, that these companies use the data to benefit the patients, not the companies themselves.”
“Machines make mistakes and sometimes they make mistakes based on faulty data,” Varshavski added. “There needs to be oversight of what these things do.”
Even within Google, the application of AI has proved controversial. The search giant recently ended its involvement in Project Maven, a controversial military program that uses AI to improve drone targeting. Project Maven has been a source of tension within Google. In April, over 3,100 Google workers signed a letter addressed to the company’s CEO, Sundar Pichai, asking him to pull the tech giant out of the project.
Follow James Rogers on Twitter @jamesjrogers