Israeli startup Zebra Medical Vision Ltd. has partnered with Johnson & Johnson’s Depuy Synthes to develop artificial intelligence (AI)-based algorithms to reduce the costs and radiation risks associated with imaging to prepare for orthopedic surgery. The idea is to replace repeated imaging, such as CT scans or MRI imaging, that are used for preoperative orthopedic surgical planning.
The technology would enable the creation of 3D models from X-ray images, which require less radiation exposure than CT and are less expensive and more commonly available than other imaging modalities. It is slated to be integrated into Warsaw, Ind.-based Depuy’s existing Velys Digital Surgery product.
The deep learning medical imaging startup already has more than 15 strategic partnerships with companies including Amsterdam-based imaging giant Royal Philips, Chicago-based medical image workflow and management company Intelerad Medical Systems Inc. and Burlington, N.J.-based radiology platform Nuance. Zebra Medical has more than 150 hospital customers worldwide. Its Triage AI products, which helps by prioritizing patient images for radiologist review, have gained the most traction thus far.
“Zebra-Med's Triage AI solutions gains a wide attraction having been the first multi-modality AI Triage solution for mammograms, CTs and X-rays, that alerts at the point of care,” Zebra Medical co-founder and CEO Eyal Gura told BioWorld MedTech. “Zebra-Med's world's first FDA approved AI chest X-ray triage product was shown to save physicians more than 85% of the time taken to reach the acute condition.”
“Sites from Europe and Australia see great value with the deployment of Zebra-Med's AI Vertebral Compression Fracture product, as it increases the number of detected patients by 400%,” he continued. “Based on the Royal Osteoporosis Society data and calculations from June 2019, this would translate to the U.K.’s NHS a savings of £3.3 million just for the Oxfordshire trust.”
Specifically in bone health, Zebra Medical the CE-marked AI Vertebral Compression Fracture product employs AI to assess osteoporosis risk and bone mineral density (BMD) values. It uses machine learning to assess the risk, as well as the existence of osteoporotic fractures by classifying and correlating various bone density scores, emulating DEXA (dual energy x-ray absorptiometry) scores and analyzing bone structure. It is part of the Nuance partnership.
Zebra Medical’s Imaging Analytics Engine examines brain, lung, liver, cardiovascular and bone disease in CT scans as well as 40 different conditions in X-ray scans, and breast cancer in 2D mammograms to detect abnormalities and prioritize images for further radiologist review. Most recently, in November it gained its fourth 510(k) clearance from the FDA for the identification and triage of pleural effusion in chest X-rays.
In the Depuy deal, Zebra Medical is taking a slightly different approach – focusing on enhancing basic technology using artificial intelligence, rather than its existing emphasis on detection and triage.
“In this specific application we aim to bridge this gap by introducing an AI-based technology that allows a creation of three-dimensional (3D) models from simple, 2D plain X-ray images,” explained Zebra Medical strategic partnership product lead Asher Kuper in a blog post. “Such a technology will revolutionize the way in which surgeons worldwide are conducting their pre-surgical planning and provide a more accurate and precise planning method.”
Zebra Medical is one of a slew of well-funded startups focused on the application of AI to medical imaging analytics and enhancement, which is shaping up to be one of the earliest and most useful applications of machine learning within health care thus far. Founded in 2014, the company has raised roughly $52 million from high-profile venture investors including Khosla Ventures, Marc Benioff, Intermountain Investment Fund, Ourcrowd Qure, Aurum, Amoon, Nvidia, Johnson & Johnson Innovation – JJDC Inc. and Dolby Ventures.
“We have the opportunity to impact and improve orthopedic procedures and outcomes in areas including the knee, hip, shoulder, trauma, and spine care,” summed up Gura. “We share a common vision surrounding the impact we can have on patients’ lives through the use of AI, and we are happy to initiate such a meaningful strategic partnership, leveraging the tools and knowledge we have built around bone health AI in the last five years.”