HeartFocus Link Expands AI Cardiac Ultrasound to Cart Systems

The intersection of artificial intelligence and medical imaging has long promised to bridge critical gaps in healthcare delivery. A recent development in cardiac diagnostics demonstrates how software architecture can adapt to existing hospital infrastructure rather than demanding complete hardware replacement. By decoupling advanced guidance algorithms from proprietary probes, a French medtech firm has introduced a pathway for broader clinical adoption. This approach addresses a persistent workforce deficit while navigating the complex regulatory environment that governs diagnostic software.

DESKi has launched HeartFocus Link, an AI-guided cardiac imaging product that connects via tablet and HDMI to any cart-based ultrasound system. The FDA-cleared software, previously available only on Butterfly Network handheld devices, targets the growing sonographer shortage that leaves 92% of eligible hospitalised patients without echocardiography.

What is HeartFocus Link and how does it function?

HeartFocus Link represents a strategic pivot for DESKi, the French medical technology company that originally developed the HeartFocus algorithm. Previously, the software operated exclusively within the Butterfly Network hardware ecosystem, which restricted its deployment to facilities that had already invested in specific handheld probes. The new Link product removes that hardware dependency by routing the tablet interface through a standard HDMI connection to any existing cart-based ultrasound machine. This architectural decision allows medical schools, residency programs, and training institutions to utilize the guidance overlay without purchasing entirely new imaging systems.

The software supports ten standard transthoracic echocardiographic views and relies on a patented three-dimensional guidance system. Instead of requiring clinicians to mentally map probe movements against a static reference guide, the system superimposes positioning instructions directly onto the live ultrasound feed. This overlay significantly reduces cognitive load by keeping the operator focused on a single visual field. The interface delivers real-time adjustments that help both experienced practitioners and trainees capture diagnostic-quality cardiac views with greater consistency.

Training environments receive specific functional enhancements designed to accelerate competency development. An automated recording feature captures image clips the moment predefined quality thresholds are met, eliminating the manual workflow of saving frames during practice sessions. A concurrent real-time quality scoring mechanism provides immediate, objective feedback on image acquisition. This shifts the training paradigm away from purely instructor-dependent assessment and toward measurable, data-driven performance metrics.

Why does the sonographer shortage matter for cardiac imaging?

The demand for cardiac ultrasound examinations has expanded at a pace that outstrips workforce growth over the past decade. Between 2011 and 2021, the volume of ultrasound procedures in the United States increased by fifty-five percent, climbing from approximately thirty-eight million to nearly sixty million. During that same period, the sonographer workforce expanded by only forty-four percent, while educational capacity grew by a mere twenty-three percent. This mathematical imbalance indicates that training pipelines cannot currently absorb retiring professionals, let alone accommodate rising clinical demand.

The consequences of this capacity gap are visible in hospital utilization rates. Clinical evidence consistently demonstrates that echocardiography reduces mortality and improves patient outcomes, yet the procedure is performed on only eight percent of eligible hospitalized patients. Sonographer vacancy rates peaked at sixteen point seven percent in 2023 and have only recently improved to twelve point four percent. These statistics highlight a structural bottleneck that limits diagnostic access across numerous healthcare systems.

AI-guided imaging tools aim to mitigate this deficit by lowering the barrier to entry for non-specialist clinicians. Research presented at the American College of Cardiology conference indicates that novice users operating with HeartFocus achieved greater than eighty-five percent agreement with expert assessments of echocardiographic parameters. The software enables practitioners to capture usable cardiac images after hours of focused training rather than requiring months of traditional apprenticeship. This acceleration of competency directly addresses the operational strain caused by staffing shortages.

How does the company plan to scale beyond its initial funding?

DESKI was established in 2016 by brothers Bertrand and Olivier Moal, who bring complementary expertise in clinical medicine and engineering. Bertrand serves as chief executive officer and holds a medical degree alongside a doctorate in biomechanical engineering, while Olivier possesses engineering qualifications from Berkeley and EPFL. The Bordeaux-based firm secured a six million dollar seed round in mid-2025, led by Racine², an impact fund managed by Serena and Makesense, with additional participation from BNP Paribas Développement. This capital injection supports ongoing algorithm refinement and market expansion efforts.

Regulatory strategy plays a central role in the company’s scaling roadmap. HeartFocus received Food and Drug Administration clearance in April 2025, accompanied by a Predetermined Change Control Plan. This regulatory framework permits DESKi to update the software’s artificial intelligence algorithms without undergoing full review cycles for every modification. The underlying models are trained on more than ten million data points, providing a robust foundation for continuous improvement within approved boundaries.

Commercial expansion also relies on credentialing partnerships that formalize clinical adoption. In April 2026, the company collaborated with Inteleos, a global healthcare certification body, to introduce the first AI cardiac point-of-care ultrasound certification. This credential allows clinicians and institutions to demonstrate verified competency in AI-assisted imaging. As hospitals begin establishing formal governance structures around diagnostic tools, standardized certification provides a necessary framework for procurement and departmental integration.

What are the technical trade-offs of the HDMI architecture?

The decision to route guidance through an external tablet and HDMI connection offers immediate compatibility but introduces specific technical limitations. Because the system operates as a visual overlay rather than a native integration into the ultrasound machine’s software stack, it cannot access or process the raw ultrasound data stream. This architectural choice restricts the depth of computational analysis the artificial intelligence can perform compared to fully integrated solutions that communicate directly with transducer hardware.

The overlay approach prioritizes rapid deployment across diverse institutional equipment. Hospitals rarely replace entire cart-based imaging systems, making a hardware-agnostic solution commercially viable. However, the inability to manipulate raw data means the algorithm relies entirely on the visual feed captured by the host machine. This dependency requires consistent calibration between the tablet display and the ultrasound monitor to maintain accurate spatial guidance for the operator.

Despite these constraints, the practical utility remains significant for training and preliminary screening workflows. The system successfully guides probe positioning and captures diagnostic frames without requiring deep technical integration. Medical institutions can implement the software immediately while evaluating whether deeper hardware partnerships or native software integration will be necessary for long-term clinical deployment.

How is the competitive landscape shifting around AI ultrasound?

DESKI operates within a rapidly evolving sector where multiple firms are pursuing regulatory approval for AI-guided cardiac diagnostics. UltraSight, an Israeli startup, recently secured its own FDA clearance to expand AI-guided echocardiography across various ultrasound systems. The broader health technology market is experiencing accelerated convergence between artificial intelligence software, sensor hardware, and clinical workflows. Companies ranging from wearable device manufacturers to specialized cardiac imaging developers are competing for clinician adoption and institutional contracts.

Navigating this competitive environment requires more than regulatory clearance. The company’s six million dollar funding base is modest relative to the scale of the diagnostic imaging market. Competing against established medical equipment manufacturers with extensive distribution networks will demand significant commercial traction or strategic partnerships with major ultrasound producers. The existing relationship with Butterfly Network provides an initial foothold, but sustained revenue generation remains the primary objective.

The challenge of translating FDA approval into widespread clinical adoption affects numerous diagnostic AI startups. Regulatory clearance validates safety and efficacy but does not guarantee procurement or workflow integration. Success will depend on demonstrating measurable improvements in training efficiency, diagnostic accuracy, and operational cost savings. Institutions will evaluate the technology based on tangible clinical outcomes rather than algorithmic capabilities alone.

What does the future hold for AI-assisted cardiac diagnostics?

The deployment of AI-guided imaging tools reflects a broader industry shift toward software-driven solutions that adapt to existing clinical infrastructure. By decoupling advanced diagnostic assistance from proprietary hardware, developers can address workforce shortages without demanding massive capital expenditures from healthcare facilities. The ongoing refinement of regulatory frameworks and credentialing standards will determine how quickly these systems transition from training environments to routine patient care. The coming years will likely reveal which architectural approaches and commercial strategies successfully bridge the gap between algorithmic promise and clinical reality.