Why do Aidoc AI tools appeal to radiologists?

Since its founding nine years ago, Israeli clinical decision support startup Aidoc has become a major player in AI in healthcare, raising $370 million and signing more than 150 contracts with health systems such as Mount Sinai, Yale New Haven Health and Sutter Health. These providers use the company’s platform to help radiologists quickly identify critical medical exam findings, as well as to ensure those findings are acted upon.

Timing has been a key factor in Aidoc’s success, noted Tom Valent, the company’s chief commercial officer, in an interview earlier this month at the Radiological Society of North America’s annual conference in Chicago.

He said Aidoc benefited from entering the market as a “second generation” healthcare AI company, after earlier startups like Arterys and Zebra Medical had already familiarized providers with the technology.

This allowed Aidoc to focus more on product innovation and delivery rather than educating the market from scratch, Valent explained.

“The first generation started to raise awareness and open the enterprise to AI, and we were able to leverage that and focus more on execution,” he said.

The other main reason for Aidoc’s success was its early focus on acute clinical use cases, which allowed the startup to quickly demonstrate its value and build credibility with vendors, Valent said.

“These conditions are often a matter of life and death, so the immediate clinical outcome is very clear, versus something that is long-term and requires extensive clinical studies to prove,” he noted.

He also highlighted Aidoc’s R&D culture as a differentiator.

Valent noted that the company’s focus on research and development — rather than aggressive marketing — has allowed Aidoc to create AI tools that easily integrate into clinical workflows and address the real-world complexity of healthcare environments without creating additional burden.

As the startup continues to develop new AI models, patient safety and quality will remain a key priority, he noted, adding that accuracy is a key factor.

“We need to make sure our products have high levels of accuracy. Why? Because if they have low sensitivity, that means they’ll miss important things – and if doctors start relying on it to triage their workflow and we have low specificity, that means we could create a lot of false positives, which would effectively lead doctors to ignore the AI,” Valent explained.

Transparency is also an important part of AI security, he stressed. He said transparency is key before and after deployment, with tools such as model cards that clearly explain how Aidoc’s algorithms are trained, where they can fail and what clinicians should expect.

He also noted that continuous monitoring of real-world usage and performance helps ensure that AI tools are used appropriately for clinical decision support.

Looking ahead, Valent believes Aidoc’s commitment to patient safety and transparency will be key to the company’s continued success.

Photo: Tom Werner, Getty Images