RadNet leaders discuss rationale behind blockbuster $103M iCAD AI acquisition
Leaders with RadNet Inc. on Wednesday discussed why they decided to pull the trigger on the company’s blockbuster acquisition of breast artificial intelligence vendor iCAD Inc.
The Los Angeles-based imaging center operator first announced the deal Tuesday, with investors receiving 0.0677 shares of RadNet common stock for each share of iCAD held at closing. This represents a transaction value of $103 million or $3.61 per share of iCAD common stock.
RadNet provides 1.9 million or about 4.4% of mammograms per year in the U.S., along with 50,000 breast biopsies. Adding iCAD’s installed customer base would boost the total to a combined 10 million annual mammograms impacted by its AI offerings, across 1,700 sites in 50 countries. The Nashua, New Hampshire-based vendor also brings $19.6 million in additional annual revenue (based on 2024 figures) and a “strong” workforce of 66 FTEs with expertise in AI research and development.
RadNet believes the acquisition will accelerate its ability to scale AI offerings to meet growing demand, “creating the industry’s most comprehensive portfolio” of breast imaging solutions. Leaders also expect the company to have a positive adjusted “EBITDA (earnings before interest, taxes, depreciation and amortization) run rate”—turning a profit from its core operations—by the end of 2026.
“We are looking at this as an opportunity to further solidify our position as the leader in the outpatient imaging space,” CEO Howard Berger, MD, told investors during a conference call April 16. “And we're getting more and more inbound calls—unsolicited, I might say—from hospital systems that are looking for an outpatient imaging strategy. And this is an integral part of that strategy, which ultimately dovetails into their inpatient strategy, too. This [acquisition] sends the message that we wanted into the community and the marketplace about our commitment to digital health, not only for RadNet, but as a transformative part of what radiology will be adopting over the next several years to transform it into truly a tech services- and tech-enabled business.”
Company history and products
The vendor’s history dates to 1984 when it launched as R.H. Research and was later renamed as Howtek Inc. It initially focused on printing and scanning technologies, helping convert hard-copy images into digital formats for industries such as radiology and graphic arts, according to a recent iCAD social media post.
By the early 2000s, the vendor acquired Intelligent Systems Software Inc., Qualia Computing and a subsidiary company, CADx Systems Inc., rebranding as iCAD in 2001. These acquisitions added the company’s namesake computer-aided detection (CAD) technology for breast cancer into its portfolio. ICAD aimed to leverage these mammographic film-scanning and pre-AI CAD capabilities to “assist radiologists in detecting breast cancer more effectively.”
By 2016, iCAD released the “industry first” 3D AI mammography solution for digital breast tomosynthesis, scoring a U.S. Food and Drug Administration greenlight in 2017. It later released another industry first with the 3D AI Risk tool in 2021 and earned FDA clearance for its Profound Detection (version 4) offering last year.
RadNet leaders discussed some of these products, how they plan to deploy them, and the differences between iCAD’s Profound Breast Health Suite and its own SmartMammo solution from subsidiary company DeepHealth. The latter is only compatible with 3D mammography in the U.S. market and on systems from GE HealthCare and Hologic. Meanwhile, iCAD’s Profound works with both 2D and 3D mammo and on Siemens Healthineers and Fujifilm breast imaging systems.
“The idea is that the AI-only product can address a larger installed base and a larger segment of the market, which is a significant product gain now that we have in our portfolio,” Sham Sokka, PhD, a former Philips executive who joined RadNet in 2023 as chief operations and technology officer in its digital health division, told investors during Wednesday's call.
RadNet touted the ability to offer radiology groups and hospitals a “cascade” of increasingly sophisticated AI services. This would start with the basic “AI-only” detection software from iCAD, which is approved in over 50 countries, to aid radiologists with spotting breast cancer on mammograms. The second tier would incorporate AI detection and SmartMammo workflow efficiency tools—including smart alerts to prioritize exam reviews, with “intelligent worklist sorting.” The software assists radiologists by overlaying AI findings on images and automatically populating AI data into reports.
And then the third tier—RadNet’s Enhanced Breast Cancer Detection Program (EBCD), which charges patients an extra $40—would include the first two items plus an additional “safeguard review.” So, in the event of a disagreement between AI and the primary reading radiologist, a second reader is brought in to review the discrepancy and reduce the need for recalls, Sokka noted. RadNet presented a study about the program at RSNA 2024, which found that 36% of women opted for the AI add-on, with this group 21% more likely to have their breast cancer detected.
RadNet also touted iCAD capabilities that it does not have in its current AI portfolio. These include a risk tool that allows radiologists to predict a woman’s potential of developing breast cancer, based on current medical images. A second product focuses on breast arterial calcification, allowing providers to assess cardiac risk factors via mammograms. Neither are approved in the U.S. yet, but RadNet sees opportunities with these capabilities. Leaders said they’ll eventually seek to cross-sell DeepHealth products to thousands of iCAD customers across the globe.
“So, taking those models, bringing them into the RadNet ecosystem where we can test and validate and use the RadNet data and clinical environment to accelerate those risk products is another part of the product synergies that we're looking to harness as well,” Sokka said.
Leaders highlighted other potential synergies that come with the acquisition. RadNet expects to cut about $7 million in costs once the organizations combine. This would include $4 million in direct savings from combining their boards of directors, sharing expenses that come with being a publicly traded company, and only having to perform one financial audit.
Another $3 million in synergies would come from satisfying planned DeepHealth hires through the added iCAD team. The vendor’s roster has 14 FTEs working in customer support, 26 focused on sales and other commercial activities, and 26 more in research and development. RadNet praised iCAD’s “strong” R&D track record, with global reach to develop, market, deploy and service AI solutions “at scale.”
“They've had this demonstrated experience of being in multiple markets and really putting a very thoughtful and accelerated strategy to enter new markets, and we're very excited to combine that capability with the rest of the product portfolio,” Sokka said.
More acquisitions on the horizon
Wednesday’s announcement continues RadNet’s recent buying spree in the AI space. The company started by acquiring Cambridge, Massachusetts-based DeepHealth in 2020 for around $44 million. In the process, it added founder, radiologist and former Siemens Healthcare North America CEO Greg Sorensen, MD, to the team. RadNet later acquired Dutch AI firms Quantib and Aidence in 2022 for a total of nearly $100 million, and it bought U.K.-based Kheiron Medical Technologies for $1 million last year. At the time of the initial DeepHealth AI acquisition, leaders emphasized the desire to improve radiologists’ efficiency and accuracy while materially benefiting shareholders.
RadNet boasts a sizable cash balance north of $700 million, and leaders in recent months have emphasized their desire to spend it on acquisitions. However, RadNet is funding the iCAD deal with stock, which prompted questions from investors during the call. The purchase figure includes an $85 million enterprise value for iCAD, with RadNet adding an additional $17.2 million in cash from the vendor’s reserves.
“We did not use any of our balance sheet, if you will, for this acquisition,” Berger told investors. “And I think the importance of that is that the stockholders of iCAD saw the benefit in this combination, this merger and being a part of the growth story in artificial intelligence, particularly for breast imaging, and therefore were confident about the long-term benefit and opportunity of having RadNet stock.”
He said the company continues to have conversations with hospitals and other provider organizations about the possibility of growing through acquisition on the imaging center side of its business. RadNet currently operates about 400 outpatient centers across Arizona, California, Delaware, Florida, Maryland, New Jersey, New York and Texas. Chief Financial Officer Mark Stolper said last month that the company continues to explore expansion outside those eight states.
“I'll emphasize that we do have an active pipeline of imaging center acquisition opportunities,” Stolper told investors Wednesday. “I'm highly confident that we will put some capital to work this year with some of the dislocation in the capital markets, both on the equity side as well as the debt side. It's putting more and more pressure on some of the smaller operators, some of the private equity-owned operators, and that bodes well for more opportunities on the core imaging center side, going forward. We have been extremely disciplined with regards to price and valuation. We'll continue to be that way. We feel we have a very good sense of what businesses are worth in our industry, but I'm also highly confident that we'll get some transactions done this year.”
Berger added that he expects consolidation to continue on the AI side of the industry after closing the iCAD deal in the coming months. The publicly traded vendor reported its 2024 financial results last month, which included a 13% increase in revenue (up to $19.6 million) and net loss of $5.6 million (an improvement from a $7 million net loss in 2023).
“This is the beginning of what we believe will be further consolidation in the AI industry within radiology, and healthcare for that matter, and the importance of that can't be overstated,” Berger said. “Many of these companies have been struggling for quite some time to establish themselves as viable companies that will be able to continue to invest and improve their artificial intelligence offerings. Clearly, most of them have been unable to do that or may fail. So, in a very real sense, this will help preserve the long-term stability in the AI space for breast imaging, which perhaps at this juncture may be one of the most important tools that AI has evolved for earlier detection and better outcomes. So, in a way, I think many of the people who are using iCAD products, as well as others, should be relieved that a company such as RadNet now is a long-term provider of the services and committed to the industry.”
Marty Stempniak has covered healthcare since 2012, with his byline appearing in the American Hospital Association's member magazine, Modern Healthcare and McKnight's. Prior to that, he wrote about village government and local business for his hometown newspaper in Oak Park, Illinois. He won a Peter Lisagor and Gold EXCEL awards in 2017 for his coverage of the opioid epidemic.