24 Gennaio 2024

Better imaging technology for the vascular system

A research team from the University of Trento has developed a promising ultrasound technique

With almost 100,000 kilometres of veins and capillaries (two and a half times the circumference of the Earth), the circulatory system travels through the human body to carry blood and fluids to organs. It has been studied since ancient times, but was precisely described only in the seventeenth century ("Exercitatio anatomica de motu cordis et sanguinis in animalibus" by William Harvey, 1628), and it is still the subject of investigation from multiple points of view. As a treasure map, the vascular system can provide valuable diagnostic and therapeutic information and, in recent years, ultrasound has become one of the main tools for vascular medical imaging. This technique however is constrained by the diffraction limit, which limits the spatial resolution to the scale of the wavelength: the resolution improves at higher frequencies, but the penetration depth diminishes.
The journal "Ieee Transactions on Medical Imaging" has recently published a study by the ultrasound laboratory Ultra of the Department of Information Engineering and Computer Science (Disi) of the University of Trento, that promises to significantly improve the quality of vascular system imaging through the interpolation of images acquired through low-frame rate ultrasound.
"Our goal – explains Giulia Tuccio, PhD student in Industrial Innovation and first author of the article – was to bring ultrasound localization microscopy (Ulm) to the clinical field. The new technique – called "Time Efficient Ultrasound Localization Microscopy" (Teulm) – makes it possible to significantly improve the imaging of the human vascular system, while continuing to use the frame rate of hospital scanners."
"Until now, this technique was mainly used in preclinical studies," continues Libertario Demi, professor of Electronic Bioengineering at Disi, principal investigator of the study and corresponding author of the article. "It is based on the use of contrast agents, micro-bubbles the size of red blood cells that are injected into the blood vessels. By following them, we collect geometric and dynamic information on tissue vascularisation."
Sajjad Afrakhteh, research fellow at Disi, more than others, was in charge of interpolation, that is, the generation of high frame rate ultrasound data from low frame rate acquisitions: "Before implementing our method for the generation of Ulm images, we discussed the difficulties of cardiac imaging. Our approach allows us to use an extremely low frame rate while maintaining the ability to accurately reconstruct vascular structures. To do this, we use both temporal and spatial information, then use radial basis functions to estimate the missing values in the estimation process."
What are the applications of this new technology? Tuccio replies: "The technique we have developed can be very useful to treat cancers or chronic diseases such as Alzheimer's, which lead to vascular alterations. This information is crucial for diagnosis and prevention, but also for more precise interventions during therapy."
The tests will continue in the coming months thanks to a grant from Fondazione Vrt. The vascular structure of an organ will be replicated by 3D printing in collaboration with the Prom Facility centre in Rovereto. The goal is to create a highly controlled model to be used to further develop the method and improve the results obtained on preclinical models.
The article "Time Efficient Ultrasound Localization Microscopy Based on A Novel Radial Basis Function 2D Interpolation", by Giulia Tuccio, Sajjad Afrakhteh, Giovanni Iacca and Libertario Demi of the Department of Information Engineering and Science of the University of Trento was published, on 25 December 2023 in "Ieee Transactions on Medical Imaging". It is available at


Image descriptions (©UniTrento - Ultrasound Lab):

  1. Preclinical datasets: reconstructed maps of the vascular system obtained by combining thousands of images per second.
  2. Preclinical datasets: a comparison of ULM- and TEULM-generated scans. As can be seen, Teulm retains vascular tissue information even at low frame rates, close to scans generated in clinical settings (100Hz).
  3. Preclinical datasets: a comparison of Teulm- and Ulm-generated scans. At the same frame rate, the Teulm-generated scan is much more defined.
  4. Preclinical datasets: the image compares Teulm- and Ulm-generated scans. Despite the much lower frame rate (50 Hz versus 500 Hz), the Teulm-generated image has a resolution comparable to that obtained by Ulm.