AI now embedded in nearly every cardiovascular IT system
Artificial intelligence (AI), once seen as something from the far-off future, is now a standard feature across all major cardiovascular information systems (CVIS) and cardiac picture archiving and communication systems (PACS). Amy Thompson, research manager for healthcare IT with Signify Research, discussed this trend with Cardiovascular Business at ACC.25 in Chicago.
Thompson said all the health IT vendors she spoke with at the conference now have AI integrations. In fact, she could not think of any big players in the market that do not have AI features designed to assist clinicians with workflows. With that progress in mind, she said hospitals looking for new systems need to be aware of how well these different AI features integrate into the technology in question. That integration is often seamless, but that is not always the case.
"With AI, you can have the most fantastic tool available in the market, but also you need to think about how it does integrate into the wider ecosystem and the workflow, which is where the CVIS, the cardio PACS vendors come into the equation," Thompson said.
She explained that many cardiovascular IT vendors are relying on partnerships with third-party AI companies rather than developing their own AI due to competing priorities and development costs. This has fueled a wave of collaborations between AI developers and imaging IT vendors in the past few years.
“We’ve seen much more with partnerships so far with imaging IT, mostly on the echo front,” she said.
In addition to exam workflow and post-processing automation seen on flagship AI-driven cardiac ultrasound systems like the Siemens Origin, startup vendors such as US2.AI are offering AI driven point-of-care ultrasound (POCUS). This type of technology was demonstrated throughout the exhibition floor at ACC.25, including the Viz.AI and Aidoc booths.
While AI in echocardiography has seen the fastest uptake due to its high study volume, AI for cardiac CT angiography (CCTA) is still emerging. “It’s a newer technology in terms of advancements,” Thompson noted. “We’ve yet to see big formal sales distribution or integration partnerships between CCTA AI vendors and cardio PACS or CVIS systems, but I expect those strategies to come into play in 2025 or early 2026.”
Partnerships are also extending beyond imaging. Thompson pointed to collaborations focused on other areas, including electrophysiology and structural heart imaging.
Ultimately, she emphasized that workflow remains the key to unlocking AI’s value. This includes a lot of AI that can scan images for unrelated CT or echo scans to look for incidental cardiac findings for more proactive preventive care.
“A lot of the preventative stuff we talk about, like risk prediction, is all well and good,” she said. “But you need the workflow to make sure that patient is appropriately followed up. Flagging a risk six or 12 months early, but not being able to do anything with it, hasn’t actually improved patient quality or care.”