Recent research by Google suggests it has developed a deep learning system which can predict a patient’s length of stay in the hospital as well as mortality rates. What’s more astounding is the fact it can predict the death of hospital patients, with 95% accuracy using Artificial Intelligence.
The study looked at a range of clinical problems in hospitalized patients. , recently published in the journal Nature cooperative Medicine (NPJ) Digital Medicine. Google applies AI technology to a large number of data from 216,000 patients at two medical centers who each stayed in the hospital for at least 24 hours. Researchers get a lot of data from electronic health records.
Can death be predicted?
Life, sickness and death are natural laws, and death is an inevitable process for human beings. Can death be predicted? As technology grows more sophisticated, the current craze for artificial intelligence may be able to do just that. However, facing such an outcome, it may feel a bit frightening.
In January, Stanford university in the United States developed an AI system to predict the death of humans. Based on nearly 2 million adults and children of the patient’s electronic health records data, combined with relevant medical diagnostic information, including: diagnosis, the number of days in the hospital, of the various treatment and medical prescription, get a big data statistics. The specific time of death can be predicted by data collection and systematic independent learning mechanism.
The research team, said the AI system to find the condition collected from 12 months to death of the patient data, they use the AI data training, let it can simulate human brain neurons organized, so you can judge the risk of patients die 3 ~ 12 months. The technology can greatly improve the year-end care for patients and their families: through more accurately predict when will the terminally ill or seriously ill patients died, caregivers can give priority to their wishes, and make sure that before it is too late for the important conversation.
Because of its reliance on massive amounts of information, healthcare is increasingly seen as the right field to use AI technology. In a speech earlier this year, FDA official Scott Gottlieb discussed the future of AI technology in the healthcare industry.
Therefore, there are still many problems to be solved to apply this technology to reality, such as AI, ethics and so on. Whether it is a blessing or a curse requires further attention.
References:
[1] Scalable and accurate deep learning with electronic health records
[2] Google Is Training Machines to Predict When a Patient Will Die
[3] Google says its AI is better at predicting death than hospitals
[4] Google’s AI Can Predict When A Patient Will Die