Zebra Medical Vision unveiled new chest X-ray image analysis research, which uses artificial intelligence to identify 40 common clinical findings, alongside the completion of a $30 million funding round.
Zebra’s TextRay deep learning AI was trained on about 2 million images from frontal and lateral scans and clinical reports, which the company said will form the basis of a future automated chest X-ray analysis product currently under development.
In addition, the series C financing brings the total investment in the 2016 Fierce15 winner up to $50 million. The round was led by aMoon Ventures, with participation from Aurum, Johnson & Johnson Innovation and Intermountain Healthcare, as well as AI scientists Fei Fei Lee and Richard Socher.
Previous investors also joined the round, including Khosla Ventures, NVIDIA, Marc Benioff, OurCrowd and Dolby Ventures.
Due to shortages of trained radiologists, many radiographic technicians are called on for preliminary interpretations, particularly in Europe and Africa, the researchers wrote (PDF). In the U.S., nonradiology physicians often provide readings.
According to the researchers’ comparisons with a team of radiologists, TextRay’s interpretations were largely on par in diagnosing pulmonary edemas, abnormal aortas, pneumothorax, fluid buildup around the lung and other ailments.
“Medical experts worldwide are facing a critical challenge in handling overwhelming demand,” said Yair Schindel, managing partner at aMoon. “This investment aligns with our vision of backing scalable and sustainable innovations that will have a valuable impact on fundamental facets of global healthcare.”
The Shefayim, Israel-based Zebra has received CE regulatory marks for seven products, most recently in CT brain bleed and mammography lesion detection algorithms, as well as fatty liver, coronary artery calcium, emphysema and other disease indicators. Their systems are in use at more than 50 hospitals worldwide, the company said.
“We see significant interest in countries that have a problematic ratio of radiologists per capita. Providing tools that assist them in delivering better care is critical,” said Zebra CEO Elad Benjamin at the time.
Late last year, Zebra began offering physicians access to all its analytics algorithms for $1 per scan, including future systems under development for breast cancer, lung cancer, brain trauma and hypertension once they are available. In addition, the company began making their programs available on Google Cloud, alleviating hospitals from the burden of having to host the images on their own servers.