Medical imaging is a comprehensive term that includes various specialized technologies. After developing AI-enhanced tools for X-rays and mammograms, the French startup Gleamer is now focusing on improving magnetic resonance imaging (MRI).
Rather than starting anew, Gleamer has acquired two startups—Pixyl and Caerus Medical—that have been working on AI-driven MRI analysis. Gleamer is part of a second wave of startups aiming to enhance medical imaging through artificial intelligence. Many tech entrepreneurs launched ventures in this field around 2014 or 2015, but while most struggled to gain traction, some consolidation has occurred. Notably, companies like Zebra Medical Vision and Arterys were acquired by Nanox and Tempus, respectively. Founded in 2017, Gleamer has developed an AI assistant designed for radiologists, functioning as a supportive tool in medical imaging. This technology aims to enhance diagnostic accuracy in interpreting medical images. To date, Gleamer has successfully onboarded 2,000 institutions across 45 countries, processing a total of 35 million examinations. The company has also received CE and FDA certifications for its bone trauma interpretation product, and it offers certified solutions in Europe for chest X-rays, orthopedic assessments, and bone age measurements. “Unfortunately, a one-size-fits-all approach to radiology falls short,” said Christian Allouche, co-founder and CEO of Gleamer, in an interview with TechCrunch. “Creating a comprehensive model that addresses all facets of medical imaging while meeting doctors’ performance expectations is incredibly challenging.”
To tackle this, Gleamer has formed small internal teams dedicated to specific areas like mammograms and CT scans. “Just three weeks ago, we launched our mammography product, which we’ve been developing for 18 months,” Allouche noted. This product is built on a proprietary AI model trained with 1.5 million mammograms. “We have partnered with Jean Zay, the French government’s GPU cluster,” Allouche added, and the company is also working on CT scans focused on cancers.
When it comes to MRI, Allouche explained, “MRI encompasses a distinct technological realm. There are numerous tasks involved, including detection, segmentation, characterization, classification, and multi-sequence imaging.” This complexity is why Gleamer is acquiring the two smaller startups to expedite its progress in this area. The financial terms of these acquisitions have not been disclosed. “These two companies will serve as our MRI platforms, with the clear goal of addressing all use cases within the next two to three years,” Allouche stated. ### Preventive Medical Imaging While Gleamer’s models demonstrate encouraging results, they are not flawless. For instance, the startup claims that its new mammography model can detect four out of five cancers, compared to a typical human radiologist, who identifies cancer in three out of five cases without AI support. However, the productivity improvements offered by a tool like Gleamer could significantly transform medical imaging. A tumor that goes unnoticed may only be detected in a follow-up examination months later. “In the near future, I believe we will all receive routine whole-body MRIs covered by our insurance, as they do not involve radiation,” Allouche remarked. Yet, in certain urban areas, there is already a shortage of radiologists to meet the demand for reactive imaging. If the industry shifts towards preventive imaging, AI tools will become essential.
Allouche envisions AI serving as an “orchestrating and triaging” tool in medical imaging. Most examinations are conducted to exclude certain diagnoses, creating a pressing need for automation powered by robust AI models that exhibit significantly higher sensitivity than human radiologists.
