Researchers in South Korea have reported that ultrasound signals passing through an ECMO circuit vary in a predictable way with the size of a clot inside it, suggesting a path toward measuring clots rather than only detecting them. The work was led by Hyun Kwan-yong of the Department of Cardiovascular and Thoracic Surgery at Catholic University's St. Vincent's Hospital and Lim Gong-min of the same department at CHA Bundang Medical Center, according to 병원신문. The findings appeared in September 2026 in Bioengineering, an international journal in the bioengineering field with an impact factor of 4.4, and were supported by a basic research grant from St. Vincent's Hospital's Clinical Medicine Research Institute.
ECMO, or extracorporeal membrane oxygenation, circulates a patient's blood outside the body to add oxygen and remove carbon dioxide when the heart or lungs are failing badly. It is a life-sustaining therapy for critically ill patients. Clots that form inside the circuit during treatment can impair the equipment or cause embolic complications in the patient, which is why early detection and ongoing monitoring are considered important.
Why it matters
The research group had already shown in 2024 that an ultrasound sensor mounted on the outside of an ECMO circuit could pick up clots moving through the tubing without touching the blood. The new study takes that a step further by asking whether the signal itself carries information about how large a clot is. If the approach holds up in real clinical conditions, it could give clinicians a non-invasive way to judge clot burden during ECMO runs, a setting where the alternative is often limited to watching for signs of trouble.
What the experiment measured
The team built cylindrical clot mimics in different diameters from acrylic, ABS and stainless steel (SUS304). They passed 5 MHz and 10 MHz ultrasound through the samples and analyzed the characteristics of the received signals.
As the diameter of the mimic grew, three signal features all rose: the maximum height of the signal, the area under the waveform, and the sum of squared amplitudes. Waveform area stood out. Across six experimental conditions covering the different materials and frequencies, it showed a linear relationship with a coefficient of determination (R²) between 0.904 and 0.997. That result is what led the authors to propose waveform area as a candidate metric for estimating clot size in future work.
Lim Gong-min, the study's first author, framed the advance as a shift from a yes-or-no answer to a quantitative one. Earlier work established that ultrasound could spot a clot, he said; this study showed that the signal changes consistently as the mimic's size changes, laying groundwork for a technique that assesses size as well as presence.
Hyun Kwan-yong, the corresponding author, said clots arising during ECMO can degrade circuit performance and cause embolic complications, making early detection and continuous observation important. He said a technology capable of assessing actual clot size, built on this research, would help improve the safety of ECMO treatment. He also noted that further validation is needed with real blood, biological clots and the flow conditions found in ECMO circuits before the method could reach clinical use.
Limits and next steps
The study used manufactured mimics rather than blood clots, and the testing took place outside a working ECMO circuit with its characteristic flow. Those gaps are the ones the researchers say they intend to close. Whether waveform area remains as reliable a marker in flowing blood, and in clots formed from a patient's own blood rather than acrylic or steel, is the question that will decide if the technique moves from bench work toward the bedside.
