Predicting kidney injury 48 hours before it happens with DeepMind’s new Artificial Intelligence
Streams is the new solution for kidney Injuries killing 100000 inhabitants in the UK every year.
In 2017 collaborating with British health care institutions, DeepMind started testing an app called Streams with the purpose to predict kidney injury.
At the beginning Streams was more about data storage -regrouping health and bloods test results before notifying clinicians in case of problem- than Artificial Intelligence as pitched by DeepMind. Then DeepMind created a machine learning algorithm capable of predicting AKI up to two days before it happened.
Since kidney function can be restore fairly easily with medications, DeepMind’s clinical lead, Dominic King, hopes using AI to predict patient deterioration would allow medical professionals to intervene earlier.
For the app to be optimized, DeepMind still have to deploy it in real world settings: making it effective for both genders, managing patient data in real-time, respecting data protection law, improving algorithms, delivering an improvement in kidney health…
However now the results are encouraging. In fact in getting healthcare providers on board with technology, Deepmind is trying to leverage three key arguments/ that its technology improves patient outcomes, that clinicians and patients actually find it useful and that it doesn’t increase per-patient healthcare costs.
DeepMind Health’s integration into Google in November 2017 suggests that the company will start to place greater emphasis on selling its products. Moreover Google offers great opportunities in terms of information collecting and exciting projects.