Google is trying to answer when the patient will die

Google's AI system can now analyze patients' medical records to help doctors guess about health conditions, even estimating the time that patients can die.

According to Bloomberg, in one case, the doctor estimated a woman with terminal breast cancer to go to the hospital in a state of pleural effusion. After X-rays, the doctor's computer analyzed important information and put her risk of death at 9.3%. In turn, Google's algorithm, after reading 175,639 data points on her, gave a higher figure of 19.9%. The woman died a few days later.

Google's AI can now analyze patients' medical records.

It is Google's announcement in the study of the potential of neural networks in health care.

Google is one of many companies that are researching and applying AI in the field of health to help solve some difficulties when it is possible to analyze large amounts of data, perform many tasks that doctors must take many hours of work. .

In this case, using neural networks has proven effective in collecting and analyzing data, using them to study and improve analysis. According to Bloomberg, Google's technology could "predict the patient's condition, including hospital stay, re-admission rates and the risk of death."

Google's algorithm has the ability to retrieve "essential data in PDF files or papers, scripted charts to determine the problem to be solved, even to make a record that leads to conclusions to help doctors diagnose better guess ".

Hospitals, doctors and medical service providers have many solutions using electronic medical databases and patient data. But existing data mining techniques are expensive, complex and time-consuming, while Google's algorithms produce accurate and very fast results.

But this also raises concerns about the future that AI can control too many things with humans, including predicting death. A series of moral concerns and the way it is used also arise. The decision to apply insurance or to allocate beds to patients is situations that can be affected by this algorithm.

In the long run, Google wants to implement "a series of new AI tools" that accurately predict human symptoms and illnesses at the clinic.