AI-powered assistance for hospitals
Mayur Saxena and his colleagues at Droice Labs aim to offer tailor-made treatment recommendations that yield faster results, reduce complications, and optimize hospital workflow.
“We work with hospitals to treat patients better,” he says. “We use artificial intelligence and machine learning to predict how prescribed treatments perform for individual patients.”
In just over a year, the Droice team has developed a system that’s already been piloted at hospitals in the United States, the Netherlands and Russia. Next, they want to test it in more hospitals in United States and identify several hospital partners that can help take the project to the next level.
Saxena founded Droice Labs with Aleksandr Makarov and Harshit Saxena. The trio of Columbia University graduate students set out to help hospitals navigate a sea of treatment data and make better-informed clinical and operational decisions.
One application for their technology is treatment plan optimization platform. The system combs scientific papers, patient records, drug-performance data, and other high-quality sources to deliver patient-specific predictions for which treatments (drugs, medical devices, or surgeries) might work best and which might trigger complications.
But it’s one thing to offer a prediction, another to deliver it in a way physicians will use.
So the team embedded their tool in all major electronic medical records systems. The result: At-a-glance recommendations that become part of existing clinical workflows.
This Droice Labs system also provides on-demand access to relevant papers and trial data, boosting physician confidence in its recommendations.
With its European pilots showing strong results, the company started testing the system in U.S. settings. Participation in the Digital Health Breakthrough Network (DHBN) will help them do that better over the next few months.
“Having a well-connected partner is very useful,” Saxena says. “HITLAB does a lot of due diligence on companies selected to join the network, which lends us additional credibility. It helps a lot in this space.”
The team is already planning what comes after DHBN studies. “In the next few months we are focusing on finding more U.S. hospital partners which can help us create more applications for the underlying artificial intelligence engine,” Saxena says.