We welcomed the Correspondence from Soares et al.1 regarding our thoughts2 on the inapplicability of the fractional moving blood volume (FMBV) concept to Virtual Organ Computer-aided AnaLysis (VOCAL) measurements. However, we believe there to be three flaws in their argument: 1) their opening sentence is slightly misleading regarding the results and conclusions of our paper; 2) a comparison between the methods of the two studies1, 2 indicates multiple differences in fixed power Doppler (PD) settings, and 3) there appear to be fundamental differences between our and their understanding of the physics of PD. Our cited paper concluded that three-dimensional (3D) FMBV using the proprietary software known as VOCAL is technically not achievable2. This is most likely to be a reflection of the fundamental properties of this particular software. We do not believe, and indeed have never claimed, that ‘fractional moving blood volume (FMBV) cannot be applied to three-dimensional power Doppler (3DPD)’. In fact, our method for generating 3D-FMBV from raw PD ultrasound signals was presented at the recent ISUOG world congress in Copenhagen3, 4. The work that is put forward to disprove our findings used different settings for frequency, line density, ensemble, smooth and balance, which may account for the variation in VOCAL index values1, 2. Whilst the underlying impact of these settings upon VOCAL indices is unknown, we would have expected Soares et al. to have used standardized settings as similar as possible to our own, in order to make appropriate comparisons. Most importantly, it appears that the authors fail to understand that PD is based upon amplitude or energy of ultrasound scattering rather than velocity5. Their Correspondence is based upon presumptions regarding Doppler shift and velocity, when it is well-established that PD is based upon amplitude. Whilst the authors are correct in stating that low pulse repetition frequencies (PRFs) ‘increase the sensitivity to detect low blood velocity’, they are incorrect in describing PD in relation to velocity of blood flow. PD is an amplitude-based technique. It quantifies all moving particles (blood) on the basis of signal strength, regardless of velocity. Doppler signals whose frequency is higher than the Nyquist limit will be aliased in spectral Doppler, but the power of the signal will not be affected by aliasing and so the PD measure should not be changed. Therefore, the assertion that high PRFs are required is irrelevant to PD, and indeed lower PRFs may at times be needed to interrogate deeper tissues. If anything, PD may be influenced by velocity at the lower rather than the higher end of the velocity range. This is because there needs to be a minimum Doppler shift detected by the transducer in order for a signal to be present that has measurable amplitude. The consequence of this is that at low flow velocities, the influence of the wall motion filter (WMF) may effectively ‘filter out’ some of the low velocity signal, which therefore cannot be measured. The authors miss this point when they describe emerging techniques to evaluate systolic and diastolic vascularity using spatiotemporal image correlation (STIC), which they erroneously describe as being ‘promising techniques to standardize 3DPD quantification’. When STIC is used to evaluate tissue vascularity it is able to produce 3D volumes from different phases of the cardiac cycle. The ratios of indices derived from volumes representative of systole and diastole may then be representative of 3D tissue resistance, as the WMF may result in more signal being filtered out in diastole than in systole6. This is the basis of novel volumetric indices such as the volumetric pulsatility index. It must be noted that these are indices that are assumed to represent impedance and they in no way act as standardization measures for 3DPD. In contrast, the FMBV technique achieves true internal standardization for vascular volume. Publication of Correspondence that erroneously applies velocity-based principles to PD imaging may mislead less experienced and emerging research groups. It is unfortunate that the derivation of indices such as those from VOCAL and their relationship to ultrasound physics are not made more explicit and readily available by the manufacturer. A. W. Welsh*†‡, S. L. Collins§¶, G. N. Stevenson** and R. Gill† †School of Women's and Children's Health, University of New South Wales, Randwick, NSW, Australia; ‡Department of Maternal-Fetal Medicine, Royal Hospital for Women, Locked Bag 2000, Barker Street, Randwick, NSW 2031, Australia; §The Nuffield Department of Obstetrics and Gynaecology, University of Oxford, Oxford, UK; ¶The Fetal Medicine Unit, John Radcliffe Hospital, Oxford, UK; **Institute of Biomedical Engineering, Department of Engineering Science, Oxford University, Oxford, UK *Correspondence. (e-mail: [email protected])
Read full abstract