Ultrasound Localization Microscopy Precision of Clinical 3-D Ultrasound Systems

• Published in: IEEE transactions on ultrasonics, ferroelectrics, and frequency control ferroelectrics, and frequency control, 2024-12, Vol.71 (12: Breaking the Resolution Barrier in Ultrasound), p.1677-1689
• Main Authors: Dencks, Stefanie, Lisson, Thomas, Oblisz, Nico, Kiessling, Fabian, Schmitz, Georg
• Format: Article
• Language: English
• Subjects: 3-D; clinical; Image reconstruction; Imaging; In vivo methods and tests; Localization; Localization method; Location awareness; matrix transducer; microbubbles (MBs); Microscopy; phantom; Position measurement; precision; Scanners; super-resolution; Superresolution; Three-dimensional displays; Transducers; Ultrasonic imaging; ultrasound localization microscopy (ULM)
• ISSN: 0885-3010; 1525-8955
• DOI: 10.1109/TUFFC.2024.3467391

Ultrasound localization microscopy (ULM) is becoming well established in preclinical applications. For its translation into clinical practice, the localization precision achievable with commercial ultrasound (US) scanners is crucial-especially with volume imaging, which is essential for dealing with out-of-plane motion. Here, we propose an easy-to-perform method to estimate the localization precision of 3-D US scanners. With this method, we evaluated imaging sequences of the Philips Epiq 7 US device using the X5-1 and the XL14-3 matrix transducers and also tested different localization methods. For the X5-1 transducer, the best lateral, elevational, and axial precision was 109, 95, and 55~\mu m for one contrast mode, and 29, 22, and 19~\mu m for the other. The higher frequency XL14-3 transducer yielded precisions of 17, 38, and 6~\mu m using the harmonic imaging mode. Although the center of mass was the most robust localization method also often providing the best precision, the localization method has only a minor influence on the localization precision compared to the impact by the imaging sequence and transducer. The results show that with one of the imaging modes of the X5-1 transducer, precisions comparable to the XL14-3 transducer can be achieved. However, due to localization precisions worse than 10~\mu m, reconstruction of the microvasculature at the capillary level will not be possible. These results show the importance of evaluating the localization precision of imaging sequences from different US transducers or scanners in all directions before using them for in vivo measurements.