IBM Watson to integrate MedyMatch's AI-based brain bleed detector

MedyMatch and IBM are teaming up to bring a decision-support tool that detects bleeding in the brain to emergency rooms. The artificial intelligence-driven application will be integrated in IBM Watson Health Imaging’s offerings.

The tool is designed to help physicians identify areas in the brain where there could be potential bleeding, said MedyMatch CEO Gene Saragnese. When a patient undergoes a head CT scan in a hospital, it is sent either to an onsite processor or an offsite processor through the cloud, he said. The application uses deep learning, other patient data and clinical insights to indicate where bleeding may be present.

While the tool itself is complex, the workflow is simple, Saragnese said. Once the case is interpreted, it is returned to the hospital’s information system, so multiple physicians may view it at the same time. The whole process is designed to take less than five minutes, he said.

Quickly identifying intracranial bleeding in the case of hemorrhagic stroke can minimize the long-term effects and could prevent death, according to the American Stroke Association. Stroke is the fourth leading cause of death in the U.S.

Launched early last year, Tel Aviv-based MedyMatch aims to deploy decision-support tools to improve healthcare, starting with the emergency room. Under the agreement, Watson Health’s imaging group will distribute the brain bleed detection tool through its established sales channels, according to a statement.

“[MedyMatch] is focused on creating new applications that help physicians make decisions in acute care settings,” Saragnese said. “… We’re not actually interested in creating a very, very large distribution network or sales force because they exist today.” The collaboration allows MedyMatch to leverage IBM’s existing infrastructure to bring its decision-support to more hospitals around the world.

Last summer, MedyMatch joined forces with New Jersey’s Capital Health to apply artificial intelligence to the analysis of stroke patients’ medical images.