The Zebra Medical Vision AI-powered tool that alerts radiologists of bleeding inside the skull has received clearance by the FDA, the third 510(k) cleared device for the deep learning medical imaging firm.
Called HealthICH, the product automatically detects high-risk internal brain bleeds based on standard, non-contrast head CT scans, which reduces the time for healthcare staff to respond with a diagnosis as well as increase their confidence.
Dr Stanley Lu, Monmouth Medical Centre’s head of neuroradiology and a radiologist, said: “We have deployed Zebra Medical Vision’s triage head CT solution that alerts our department of acute cases of intracranial hemorrhage.
“We have found the technology to be helpful as a second layer of analysis augmenting the clinical team’s detection and response time for acute cases.”
How Zebra Medial Vision uses AI to detect intracranial hemorrhages
Based in Israel, the healthcare imaging company uses a proprietary database of millions of imaging scans, along with machine and deep learning tools to create software that analyses data in real time with human level accuracy.
Intracranial hemorrhage accounts for around 10% to 20% of all strokes and can occur as a result of traumatic injury, ruptured arteries, stroke or cancer.
According to the National Institute of Health, the overall incidences of intracranial hemorrhage worldwide is 24.6 per 100,000 person per year with approximately 40,000 to 67,000 cases per year in the US.
The 30-day mortality rate ranges from 35% to 52%, with only 20% of survivors expected to have full functional recovery.
It is estimated that about 50% of this mortality occurs within the first 24 hours, emphasising the importance of early detection and effective treatment in the Emergency Department.
Zebra-Med’s intracranial hemorrhage triage solution can provide early detection of people who may have experienced a brain bleeding based on X-rays, CT scans and MRIs.
The algorithm comprises a unique neural network architecture designed to detect intracranial hemorrhage while tackling several challenges, such as relatively small bleed size and high variance within the brain.
Last month, the company announced it had received FDA clearance for its automated chest x-ray triage product that uses deep learning technology to analyze chest x-ray images in order to detect and prioritise lung trauma.
A recent study, titled Improved Intracranial Hemorrhage Classification using Deep Multi-task Learning, demonstrates how the AI-based product can achieve markedly high levels of intracranial hemorrhage detection and classification.
Dr Keith White, medical director of Imaging Services at Intermountain Healthcare, said: “We are working to improve our ability to quickly identify and respond to potentially dangerous clinical conditions through deployment of Zebra Medical Vision’s intracranial hemorrhage solution.”
