With an estimated 22 million veterans from service in the US’ 5 main branches – the Army, Navy, Air Force, Marine Corps and Coast Guard – the US Department of Veterans Affairs (VA) offers a comprehensive range of benefits to those who have served, mostly during combat. One of these benefits is enrolment in the country’s largest integrated health system with care that is personalised, proactive and driven by the needs of the patient.
Flow Health is a precision medical platform that combines artificial intelligence (AI) with machine learning to produce an operating system that can sift through data and provide insight into patient care and treatment. In a five-year deal with the VA, Flow Health will work with the US Department to help develop personalised healthcare plans with deep learning at the heart. Deep learning, an offshoot of machine learning, will be used to create the world’s largest knowledge graph, a comprehensive database that will even contain genome and phenotype information. When looked at from a high-level, this can help to identify those at risk of disease and manage treatment pathways.
Using AI requires an incredible amount of computational power, much more than your typical server or rack. This unique partnership will open up access to VA records for more than 20 million veterans in towns, cities and states across the country – an estimated 30 petabytes (30,000,000,000,000,000 bytes) of information.
The provision of healthcare is a topic that reaches far beyond the States. Google’s AI initiative, DeepMind, has recently signed a five-year partnership with the Royal Free London NHS Foundation Trust, a much larger commitment than the initial pilot earlier in 2016. DeepMind’s Streams will run off the NHS’ patient data, alerting clinicians should their patients show indicators for Acute Kidney Injuries. This AI will do more than just analyse data, it could eventually form the foundation for treatment paths. Running from a smartphone, the app will be rolled out next year with hopes that the AI could help power devices that can detect other life-threatening illnesses.
As has been found in automation and manufacturing, machine learning is a tool that can highlight improvements across the board. When you apply this learning to the healthcare system, the applications can reach far beyond national infrastructure. Similar to the VA’s project with Flow Health, we could soon see wearable-driven health insurance, AI powered healthcare providers and long-distance surgeries via robotics – all powered by data, data that could one day save lives.