by Amirah Al Idrus |
Intensix, which is harnessing machine learning in the ICU, reeled in an $8.3 million Series A. The funds are pegged for the expansion of its sales and marketing ops in North America as well as further development of its predictive analytics platform.
The Intensix platform applies machine learning to the early detection of life-threatening complications in intensive care. A proliferation of structured and unstructured ICU data—from vital signs to historical and demographic data—goes into the system, is run through a set of models and results in predictions, CEO Gal Salomon said.
ICU staff and management could use these predictions to head off deterioration before it happens. Studies have shown that the tech could potentially save lives, reduce the length of hospital stays and lower costs, the company said in a statement.
“The explosion of patient data, via electronic health records, sensors, and medical devices, provides physicians with an untapped wealth of new information,” said Ittai Harel, managing general partner at Pitango Venture Capital, which led the round. “Machine learning and predictive analytics have the power to harness this data, offering great value to the healthcare industry, particularly in the area of critical care.”
While it is easy to recognize that something has gone wrong when, for example, a patient’s blood pressure drops, signs of deterioration can come too late, Salomon said. Results from a prospective study at Tel Aviv Medical Center showed that the platform could recognize deterioration before care teams notice it might occur. The study started out modeling sepsis, the leading cause of death in the ICU, and its results have become the starting point of an interventional study, Salomon said.
The company teamed up with the Mayo Clinic last year to look into the feasibility of using its platform to predict deterioration associated with infection. The results will be published later this month.
Intensix is also in discussions with different institutions in the U.S. to conduct more studies of the platform, Salomon said. The company will focus on collecting information and verifying the accuracy of the technology in order to be ready for the commercial phase next year, he said.
As more players start using the tech, its predictive ability will only improve.
“In the domain that we’re facing—machine learning—the bigger the data that you have, the better the accuracy of the model,” Salomon said.