Mountain View, CA. – Google recently announced a new research paper they have completed. This paper showcases researchers using 46 billion data points to help predict medical outcomes within a hospital setting. Quartz.com reported the study included 216,221 adults over 11 years at two hospitals, University of California San Francisco Medical Center and the University of Chicago Medicine.
The biggest claim: being able to predict your death 24-48 hours before current methods – which in turn allows physicians the necessary time to save one’s life.
The biggest challenge: the diversity in how the data is collected. The written notes among health care providers differ as well as how they report the data within the various systems.
The solution: Google’s AI team built three complex deep neural networks to decode and decipher the data and built common outcomes based on the years of data.
You can read more about the study here and the original article by Quartz here.
Why This Matters:
Our team has been following AI trends in one capacity or another for years. And most recently, we released a new report based on our experts’ perspective on how AI impacts the most important member of any medical team: the patient.
This study above has a direct impact (obviously) on patient outcomes. If a machine can help predict what is needed to save your life after you’ve entered the hospital, would you question an earlier intervention? Neither would I. But alas, there are always questions on how reliable the data, unnecessary and costly interventions and of course, who’s to blame if the machine makes a mistake?
As Google, Amazon and others dip their toes into the healthcare space, this new study among other developments have us jazzed as to ‘what’s next’ on the horizon of AI and improving the patient experience. Read the AI related trends we’re covering this year and many more impacting healthcare here and reach out to us with questions around how we can help you better understand the shifts that are occurring in your market.
