This article was first published on Computer Vision News of March 2022.
Wiro Niessen is a man of many talents. He is a Professor of Biomedical Image Analysis at Erasmus MC in Rotterdam and Delft University of Technology. In the Netherlands, he works for Health-RI, an initiative to facilitate a health data infrastructure to enable optimal use of health data for research and intervention. Our readers also know him as the president of MICCAI from 2016 to 2019. Wiro speaks to us today about Quantib, a leading radiology artificial intelligence company, recently acquired by RadNet, where he is Founder and Scientific Lead.
Quantib was founded in 2012 as a spinoff from Erasmus MC. At the time, Wiro’s research group was working on multiple projects, including looking to develop quantitative imaging biomarkers to make radiology more objective. Several companies were interested in their work, and eventually, Quantib was founded with General Electric Healthcare as a launching customer.
Quantib’s first product analyzed MR brain images, quantifying atrophy or white matter lesions to support the diagnosis and prognosis of neurodegenerative diseases such as dementia. It received a very positive response from clinicians.
“It’s such a sense of achievement when you get feedback that these tools are helping clinicians in their clinical practice,” Wiro says proudly.
Compared to working in a university lab, where people are driven by curiosity, a commercial company needs to build a product that people want and are willing to pay for. It must be able to fulfill a proven clinical need and should be able to demonstrate how well it works, that it is of a certain quality, and what its limitations are. In a short space of time, it can go from research and proof of concept to a product that people use, with FDA approval and CE marking.
“What we brought from the academic side was the AI knowledge and a lot of talent,” Wiro explains.
“My research group was already within the hospital and we were well aware of where the potential was for AI in clinical practice, but we could not do it alone. When we started Quantib, we hired a new CEO, and we brought in expertise in quality, marketing and sales, and business intelligence. I stayed with the research and development side, and other people took over the business side. You have to collaborate with people from different backgrounds, experiences, and cultures to make something like this a success!”
Quantib has developed several products over the years. After the MR brain application, it investigated areas of urgent need where technological solutions were still missing. It identified prostate cancer as an interesting application. MRI has a non-invasive role in prostate cancer in monitoring and diagnosing patients and determining which patients need additional diagnostics. For example, some people will need a biopsy, while others will not, and MRI is invaluable in making those decisions.
“… A combination of high-quality imaging and radiologists supported by AI is the way forward for improving cancer diagnostics and screening.”
“With such a large user base, we’ll have access to much more data!”
“It’s complex imaging data to analyze even for an expert radiologist,” Wiro points out.
“We started to develop AI products that could assist the radiologist in this analysis. We got in touch with RadNet as it is the largest provider of radiological services in the US. They were interested in providing high-quality imaging, particularly for breast cancer, lung cancer, and prostate cancer, and were already active in AI for breast cancer. The idea being that a combination of high-quality imaging and radiologists supported by AI is the way forward for improving cancer diagnostics and screening.”
RadNet and Quantib have a symbiotic relationship. One is providing imaging services, with access to clinical expertise and plenty of data, but lacking the assistance of AI. The other is an AI company. Together, they can create a proactive rather than reactive approach, with imaging performed earlier so that disease can be detected early, before it has progressed, opening up the possibility for better treatments.
“That’s been my dream!”
With Quantib now aligned under the “RadNet family”, Wiro anticipates it having a significant impact on people’s lives.
“We’re going to deliver high-quality imaging for the main cancer types, supported by AI, to reduce the overall burden of disease,” he affirms.
“That’s been my dream! The number of people who will profit from the software developed at Quantib will increase enormously because of all the imaging centers run by RadNet. I’m even more excited about continuous product development. With such a large user base, we’ll have access to much more data! We’ll be able to retrain and optimize our algorithms, gather feedback from the clinical end-user, and repeat this cycle over and over to improve the quality of the end product.”
Quantib is now 40 people with talents ranging from advanced AI and deep learning to customer needs and quality.
“Individually, they’re very good people, but our uniqueness is in the team,” Wiro smiles.
“Constant improvements in technology have made it possible to have a massive impact on healthcare. It’s wonderful to see the community of medical imaging, medical image analysis, and AI impacting the health and healthcare of the future. It’s a fantastic time to work in this domain!”
Keep reading the Medical Imaging News section of our magazine.
Read about RSIP Vision’s R&D work on Medical Image Analysis.