In this edition of our MedTech Spotlight Series, Melody Lam, Head of BioTalent West Coast, spoke with Joseph Sokol, CEO & Founder of iCardio.ai, about how AI is changing the future of echocardiography and cardiovascular imaging. Based in Los Angeles, iCardio.ai is developing deep learning technology designed to automate the interpretation of heart ultrasounds.
The company’s AI products have been trained on more than 200 million images, with partnerships across major ultrasound vendors including Abbott, Butterfly Network, and UltraLinQ. In 2024, iCardio.ai received FDA Breakthrough Device Designation for its automated capability to detect aortic stenosis from b-mode ultrasound imagery, followed by 510(k) FDA clearance for its AI system for automated echocardiography.
Joseph shared that he originally came from a finance background but wanted to work on something more meaningful and impactful. When the opportunity to work with the original iCardio.ai dataset came up, he shifted full-time into the company nearly seven years ago.
“In MedTech, beyond career success, you have the opportunity to directly improve and save lives.”
At the core of iCardio.ai’s work is automating a wide range of tasks in echocardiography, which Joseph described as the most fundamental and widely used imaging modality in cardiology. He explained that improving interpretation in echo has the potential to impact clinical decision-making at scale across cardiovascular care.
One of the company’s biggest differentiators is its dataset. iCardio.ai developed one of the largest echocardiography datasets in the world and recently expanded it further through a collaboration with Cedars-Sinai, incorporating additional imaging data into its training resources.
“That gives us one of the most valuable and unique training resources in cardiac ultrasound AI today.”
When discussing AI implementation in healthcare, Joseph explained that trust remains one of the biggest challenges. While he believes AI will become the new standard for interpretation, especially in imaging, physicians are understandably confident in their own interpretations. He also pointed out that echocardiography has some of the highest inter-reader variability of any imaging modality.
Joseph believes AI can improve consistency, accuracy, and efficiency while reducing cost and fatigue-related variability. The remaining hurdle, he said, is adoption confidence.
Looking ahead, he sees trust in clinical AI developing gradually, similar to the evolution of autonomous driving. He explained that adoption will likely begin with assistive capabilities such as measurements, then progress into disease detection, and eventually expand into full report generation.
“Over time, once clinicians see that AI outputs are more reliable and consistent, trust naturally increases and adoption accelerates.”
As AI adoption continues to grow across healthcare, Joseph’s perspective reflects how companies like iCardio.ai are focused on using technology to support clinicians with more consistent and scalable interpretation in cardiovascular imaging.