How Can Machine Learning Improve Health Care? Ask Google

Share |

Machine learning is the new buzzword in technology innovation. By giving computers the ability to constantly learn and adapt using past examples and precedents, it has the ability to significantly define innovation in the 21st century. Machine learning can provide a multitude of benefits to society. Self-driving cars, which ultimately exist because of machine learning, have the ability to reduce car accidents and make transportation more efficient. At the same time, speech recognition, driven by machine learning, allows individuals to communicate in ways unimaginable just years ago. Machine learning is also driving innovation in healthcare, by allowing doctors to better detect, predict and diagnose illnesses and injuries.

 

Technology’s role in healthcare innovation is not new. From the first successful transfusion of human blood in 1818, to handheld ultrasound machines, technology has played an integral part in healthcare development and transformation. Recently, the medical community has taken large strides towards electronic health systems in order to assess the plethora of information from medical records. Machine learning has already demonstrated the ability to be tremendously beneficial in analyzing all of this electronic medical information, and this trend will continue as it provides doctors with predictive analytics to better determine how long a patient will be hospitalized, how long they will receive treatment or even what kind of medicine a patient may need depending on their current condition.

 

This week, Google announced their partnership with world-class medical researchers and bioinformaticians at UC San Francisco, Stanford Medicine and University of Chicago Medicine to explore how machine learning combined with clinical expertise could improve patient outcomes, avoid costly incidents and save lives.

 

It is no secret that healthcare can be expensive and extremely burdensome to citizens and families. However, by better taking advantage of medical information, Google hopes to provide doctors with more insight to be able to predict patient outcomes. Every year, for example, problems with medications cause more than 770,000 injuries, deaths and unplanned hospitalizations. With predictive analytics from machine learning, this number and many other staggering medical-associated costs can be diminished substantially.

 

However, compiling massive amounts of medical information and data needed for machine learning is not an easy task. So by partnering with several institutions, Google hopes to standardize or harmonize data inputs and interoperability across organizations. Using their established deep learning technology on top of the open data standard FHIR, clinicians and researchers will be better equipped to gather and utilize this data. Also, in line with HIPPA privacy rules, Google strictly adheres to patient privacy and protects patient data in their secure Google Cloud infrastructure.

 

Though Google recognizes that the majority of their employees are not doctors or practitioners, they also recognize the ability of machine learning to provide tremendous technological breakthroughs in healthcare to not only reduce healthcare-related costs, but also improve the health of millions worldwide.

 

Nikola Nikola Marcich is an intern with the SIIA Policy team. He is currently an undergraduate student at the University of Virginia studying foreign affairs and international economics.