Google AI Has Accuracy 95% Predict People Will Die
It’s not
something unusual that AI can read abundance of data well but when it comes to
prediction, AI might be ranked working beyond human proficiency. To make
prediction, a team of researcher from Stanford, the University of Chicago and
UC San Francisco have developed Google AI to analyze patient condition in hospital.
AI predicts a patient whether will die in hospital after 24 hours admission including
how many days will be in the hospital and the possibility of being readmitted with
accuracy 95%. It is said that the accuracy could also help doctors to improve
medical treatment for patients. Capability of AI to predict related with the
process of machine learning that has been running for years. According to researchers,
the AI has been taught since 2016 with the 221 adults participants from two US
medical centers for analyzing data amount to 46 billion. Data that AI read
contains patient’s age, ethnicity and gender. Some other information that
support algorithm to analyze like patient’s vital medical history, charts and
doctor’s hand written. (TI01/FN/20/6/June/2018)
No comments:
Post a Comment