Scientists from Colorado State University and the University of São Paulo have overcome a challenge that has prevented medical ultrasound imaging from being used in intensive care and emergency room settings. This technological advancement could someday lead to improved critical care for patients.
Obtaining accurate medical images using ultrasound computed tomography (USCT) is challenging due to the uncertainty of sensor positioning, which changes with patient movement. The interdisciplinary team, including a mathematician, an engineer and a geophysicist, found a seismological solution to this problem and improved the resolution of ultrasound images for lung monitoring.
Their results, published in March in IEEE Transactions on Biomedical Engineering , bring USCT one step closer to being a safe, cost-effective and portable solution to radiation-emitting medical imaging.
Safer than CT scans
Ultrasound computed tomography does not emit ionizing radiation, so it is a safer alternative to medical imaging that uses X-ray computed tomography, or "CAT" scans. Ionizing radiation can damage living tissue, especially with repeated scans.
CT scanners also are generally large and stationary, so patients must be transported to the machines, regardless of their condition or life-saving equipment to which they might be connected. Moving patients with head or back trauma can be dangerous, and USCT offers a portable solution that can be brought to patients.
CSU Mathematics Professor Jennifer Mueller has been working to develop a USCT imaging option for bedside monitoring, so doctors can monitor and treat dangerous conditions quicker, without exposing patients to ionizing radiation.
"[USCT] has the potential to provide a real-time response to developing conditions that could be treated sooner rather than later when they're a big problem," Mueller said.
However, to use USCT in intensive care units and emergency rooms, sensors must be attached to the patient for ultrasound to be transmitted through the body. USCT sensors would be worn by the patient on a belt strapped around their body. Unlike the fixed parameters of CT machines, USCT boundaries vary with body size and shape. Patients also cannot remain perfectly still, so sensor positions would change constantly, with every breath.
Mueller was working to make USCT a viable option for bedside monitoring in early 2024 when Roberto Ceccato, a Ph.D. candidate at the University of São Paulo's Electrical Engineering School in Brazil, joined her project as a research intern.
Interdisciplinary collaboration and a little bit of serendipity
Ceccato had used CSU Geosciences Professor Rick Aster's textbook Parameter Estimation and Inverse Problems and sought out the geophysicist while working in Mueller's lab.
Sound-based seismic imaging is used to analyze Earth's interior. When Ceccato explained the sensor positioning problem to Aster, Aster suggested trying a seismic tomography technique used to accurately locate earthquakes and study Earth structure by correcting for irregular, near-surface geologic features – static correction.
Static corrections in seismology are estimates of how much a seismic wave slows down or speeds up close to a receiver during the final leg of its path through the Earth. The corrections account for variations such as elevation, local geology and weathering, which can delay or advance the signal's arrival.
In medical ultrasound imaging, the same mathematical method can be applied to find unknown changes in sensor position. Resolving this effect improves imaging so that more accurate and finer details can be distinguished.
"It provides adaptive sort of eyeglasses to correct for the distortion of not knowing exactly where your receivers and sources are located in this kind of a problem," said Aster, who co-authored the study.
Whether viewing internal organs or the interior of the Earth, the mathematics and physics are the same, he added. "USCT is applying seismology to the human body, basically."
Nearing clinical testing
Mueller and Ceccato developed an algorithm to accurately estimate the positions of ultrasound sensors and tested the approach in simulations and experiments using fake torsos made of ballistic gel, which is similar to human tissue. They consulted with medical experts to make sure the procedure would be useful in clinical settings.
In a previous study , Mueller's team ensured that they were using amplitudes and frequencies within the FDA safety ranges and tested the approach on pigs. She said their method is close to being tested on humans.
The new study demonstrably improved the resolution of the images by correcting the positioning problem.
"With this solution, continuous lung monitoring at the bedside becomes more practical, helping physicians make more informed decisions about mechanical ventilation strategies and track lung health in conditions such as COVID-19," Ceccato said.
Aster said this study is a good example of why it's important for students to learn the fundamentals of math and physics to achieve science and technology innovations. "It shows the power of science and mathematics to be applied across lots of different applications."
