CoChat

Acoustic resolution photoacoustic Doppler velocimetry in blood-mimicking fluids

Open access
Brunker, Joanna, Beard, Paul · 2016 · Scientific Reports

The DOI matches its Crossref record; 21 of the 22 references it cites were found in a registry or an index.

Abstract

Photoacoustic Doppler velocimetry provides a major opportunity to overcome limitations of existing blood flow measuring methods. By enabling measurements with high spatial resolution several millimetres deep in tissue, it could probe microvascular blood flow abnormalities characteristic of many different diseases. Although previous work has demonstrated feasibility in solid phantoms, measurements in blood have proved significantly more challenging. This difficulty is commonly attributed to the requirement that the absorber spatial distribution is heterogeneous relative to the minimum detectable acoustic wavelength. By undertaking a rigorous study using blood-mimicking fluid suspensions of 3 μm absorbing microspheres, it was discovered that the perceived heterogeneity is not only limited by the intrinsic detector bandwidth; in addition, bandlimiting due to spatial averaging within the detector field-of-view also reduces perceived heterogeneity and compromises velocity measurement accuracy. These detrimental effects were found to be mitigated by high-pass filtering to select photoacoustic signal components associated with high heterogeneity. Measurement under-reading due to limited light penetration into the flow vessel was also observed. Accurate average velocity measurements were recovered using “range-gating”, which furthermore maps the cross-sectional velocity profile. These insights may help pave the way to deep-tissue non-invasive mapping of microvascular blood flow using photoacoustic methods.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/srep20902

Cites

21 of 21 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Experimental setup for pulsed photoacoustic Doppler flow measurement where the motion of micron-scale absorbing polystyrene microspheres suspended in water is used to represent blood flow. A pair of laser pulses separated by a time T are used to generate a corresponding pair of acoustic waveforms which are detected by an ultrasound receiver. The inset shows a typical distribution of the absorbers in the fluid phantoms. A large area (at least 5 mm diameter) of the absorbers is illuminated, but photoacoustic signals are only collected from the smaller region defined by the transducer focal spot in order to be representative of acoustic resolution PAF. The signals were amplified and then captured by an oscilloscope and downloaded to a PC, where they were processed to calculate the time shift ts between the two waveforms that arises due to motion of the absorbers between the two laser pulses.
Figure 1 Experimental setup for pulsed photoacoustic Doppler flow measurement where the motion of micron-scale absorbing polystyrene microspheres suspended in water is used to represent blood flow. A pair of laser pulses separated by a time T are used to generate a corresponding pair of acoustic waveforms which are detected by an ultrasound receiver. The inset shows a typical distribution of the absorbers in the fluid phantoms. A large area (at least 5 mm diameter) of the absorbers is illuminated, but photoacoustic signals are only collected from the smaller region defined by the transducer focal spot in order to be representative of acoustic resolution PAF. The signals were amplified and then captured by an oscilloscope and downloaded to a PC, where they were processed to calculate the time shift ts between the two waveforms that arises due to motion of the absorbers between the two laser pulses.CC BY · PubMed Central
Figure 2 Cross-correlation of entire waveforms (top row) and waveform segments (bottom row) according to Methods 1 and 2 respectively as described in the text. (a) One of 25 PA waveform pairs (p1(t), p2(t), each 5000 points, 1.25 μs long), acquired using the 30 MHz transducer (θ = 45°) following double pulse illumination (T = 0.5 ms) of a 9% suspension of red spheres flowing at an average velocity of V ± ΔV/2 = 58 ± 6 mm/s in a 2R = 390 μm tube. The mean C(t) of all 25 cross-correlation functions is shown in (b) and an enlargement of 100 points (sampling interval 0.25 ns) centred on the peak in (c). 62.5 ns (250-point) segments of p1(t) and p2(t) are enlarged in (d), and the corresponding mean cross-correlation function Cs(t), and 100 points around the peak are shown in (e) and (f) respectively. The dashed vertical lines in (b,c,e,f) mark the known time shift (11 ns). Compared to Method 1 (c), Method 2 (f) gives closer agreement between the known time shift and the measured value: the reason for this is discussed in the section on light penetration.
Figure 2 Cross-correlation of entire waveforms (top row) and waveform segments (bottom row) according to Methods 1 and 2 respectively as described in the text. (a) One of 25 PA waveform pairs (p1(t), p2(t), each 5000 points, 1.25 μs long), acquired using the 30 MHz transducer (θ = 45°) following double pulse illumination (T = 0.5 ms) of a 9% suspension of red spheres flowing at an average velocity of V ± ΔV/2 = 58 ± 6 mm/s in a 2R = 390 μm tube. The mean C(t) of all 25 cross-correlation functions is shown in (b) and an enlargement of 100 points (sampling interval 0.25 ns) centred on the peak in (c). 62.5 ns (250-point) segments of p1(t) and p2(t) are enlarged in (d), and the corresponding mean cross-correlation function Cs(t), and 100 points around the peak are shown in (e) and (f) respectively. The dashed vertical lines in (b,c,e,f) mark the known time shift (11 ns). Compared to Method 1 (c), Method 2 (f) gives closer agreement between the known time shift and the measured value: the reason for this is discussed in the section on light penetration.CC BY · PubMed Central
Figure 3 Comparison of the accuracy of velocity measurements made for suspensions of 3 μm red spheres in distilled water in concentrations ranging from 5% to 80% and flowing in a 600 μm tube. The measurements were acquired with T = 0.5 ms and the 80 MHz focussed transducer. (a) shows the measurements made for concentrations of 5%, 15%, 30% and 80% with a zero offset correction applied, and with the resolution (error bars) omitted for clarity. The mean fractional error for these and intermediate concentrations are shown in (b). The mean fractional error values were calculated for velocities |V| < 30 mm/s and the error bars represent the standard errors.
Figure 3 Comparison of the accuracy of velocity measurements made for suspensions of 3 μm red spheres in distilled water in concentrations ranging from 5% to 80% and flowing in a 600 μm tube. The measurements were acquired with T = 0.5 ms and the 80 MHz focussed transducer. (a) shows the measurements made for concentrations of 5%, 15%, 30% and 80% with a zero offset correction applied, and with the resolution (error bars) omitted for clarity. The mean fractional error for these and intermediate concentrations are shown in (b). The mean fractional error values were calculated for velocities |V| < 30 mm/s and the error bars represent the standard errors.CC BY · PubMed Central
Figure 4 Comparison of the accuracy of velocity measurements made for a suspension of 3 μm red spheres (concentration in distilled water approximately 14%) flowing in tubes with diameters of 250 μm, 600 μm and 800 μm. The measurements were acquired with T = 0.5 ms and using a focussed PVDF transducer with a centre frequency of approximately 25 MHz. The velocity measurements are shown in (a) with the resolution (error bars) omitted for clarity. The accuracy measurements in (b) are the mean fractional error values calculated for all velocity measurements |V| < 50 mm/s and the error bars represent the standard errors.
Figure 4 Comparison of the accuracy of velocity measurements made for a suspension of 3 μm red spheres (concentration in distilled water approximately 14%) flowing in tubes with diameters of 250 μm, 600 μm and 800 μm. The measurements were acquired with T = 0.5 ms and using a focussed PVDF transducer with a centre frequency of approximately 25 MHz. The velocity measurements are shown in (a) with the resolution (error bars) omitted for clarity. The accuracy measurements in (b) are the mean fractional error values calculated for all velocity measurements |V| < 50 mm/s and the error bars represent the standard errors.CC BY · PubMed Central
Figure 5 Comparison of photoacoustic (PA) signal time series (b) and normalised signal frequency content (c) for a low concentration (5%, first row) and a high concentration (80%, second row) of the red spheres. The plots in (1a) and (2a) show two-dimensional views of simulated 3 μm diameter spheres distributed at 5% and 80% concentrations respectively (average particle separations 26.1 ± 0.2 μm in and 8.42 ± 0.01 μm: see equation for Wigner-Seitz radius in Supplementary Information online) in a cube with a side length of 100 μm. The photoacoustic data in (b) were acquired using the 80 MHz transducer and for suspensions flowing in a 600 μm tube. The waveforms shown in black are the mean of over 2000 PA signals acquired for velocities ranging from 0 mm/s to 80 mm/s, and the example waveforms shown in grey are those with the median amplitude. The normalised frequency spectra in (c) are fast Fourier transforms (FFTs). The black line shows the mean of over 2000 FFTs calculated for the same set of signals that are evaluated in (b); the grey line is the FFT of the example signal shown in (b).
Figure 5 Comparison of photoacoustic (PA) signal time series (b) and normalised signal frequency content (c) for a low concentration (5%, first row) and a high concentration (80%, second row) of the red spheres. The plots in (1a) and (2a) show two-dimensional views of simulated 3 μm diameter spheres distributed at 5% and 80% concentrations respectively (average particle separations 26.1 ± 0.2 μm in and 8.42 ± 0.01 μm: see equation for Wigner-Seitz radius in Supplementary Information online) in a cube with a side length of 100 μm. The photoacoustic data in (b) were acquired using the 80 MHz transducer and for suspensions flowing in a 600 μm tube. The waveforms shown in black are the mean of over 2000 PA signals acquired for velocities ranging from 0 mm/s to 80 mm/s, and the example waveforms shown in grey are those with the median amplitude. The normalised frequency spectra in (c) are fast Fourier transforms (FFTs). The black line shows the mean of over 2000 FFTs calculated for the same set of signals that are evaluated in (b); the grey line is the FFT of the example signal shown in (b).CC BY · PubMed Central
Figure 6 (a) Weighted mean frequencies (WMFs) calculated from the fast Fourier transform (FFT) of signals acquired for different concentrations of 3 μm spheres. The WMF was calculated by summing the product of the amplitudes and the frequencies and normalising by the sum of the amplitudes. The black data points show the mean WMF of over 2000 FFTs calculated for a set of the same number of PA signals acquired for the relevant concentrations, and the error bars represent the standard deviation. The grey data points show examples of WMFs calculated from a single PA signal of median amplitude. Examples of the mean and median FFTs for the 5% and 80% concentrations are shown in Fig. 5. (b) Weighted mean frequencies as in (a) but plotted vs. the mean spacing between the spheres, calculated from the Wigner-Seitz radius which assumes a random distribution (see Supplementary Information online).
Figure 6 (a) Weighted mean frequencies (WMFs) calculated from the fast Fourier transform (FFT) of signals acquired for different concentrations of 3 μm spheres. The WMF was calculated by summing the product of the amplitudes and the frequencies and normalising by the sum of the amplitudes. The black data points show the mean WMF of over 2000 FFTs calculated for a set of the same number of PA signals acquired for the relevant concentrations, and the error bars represent the standard deviation. The grey data points show examples of WMFs calculated from a single PA signal of median amplitude. Examples of the mean and median FFTs for the 5% and 80% concentrations are shown in Fig. 5. (b) Weighted mean frequencies as in (a) but plotted vs. the mean spacing between the spheres, calculated from the Wigner-Seitz radius which assumes a random distribution (see Supplementary Information online).CC BY · PubMed Central
Figure 7 The effect of low-pass and high-pass filtering the photoacoustic data on velocity measurements made for a 9% concentration of red spheres flowing in a tube of diameter 390 μm. The PA signals were acquired using a PZT transducer with a centre frequency of 30 MHz. The velocity measurements for the unfiltered data are shown in (b) and the normalised mean FFT of all the signals is shown below. The measurements and corresponding FFT spectra in (a) and (c) are for PA signals after they were passed through a low-pass 5 MHz filter (a) or a high-pass 30 MHz filter (zero-phase Butterworth filter, order 2).
Figure 7 The effect of low-pass and high-pass filtering the photoacoustic data on velocity measurements made for a 9% concentration of red spheres flowing in a tube of diameter 390 μm. The PA signals were acquired using a PZT transducer with a centre frequency of 30 MHz. The velocity measurements for the unfiltered data are shown in (b) and the normalised mean FFT of all the signals is shown below. The measurements and corresponding FFT spectra in (a) and (c) are for PA signals after they were passed through a low-pass 5 MHz filter (a) or a high-pass 30 MHz filter (zero-phase Butterworth filter, order 2).CC BY · PubMed Central
Figure 8 Comparison of the accuracy of velocity measurements after passing the PA signals through one of various low-pass and high-pass filters. The velocity measurements were acquired using the 30 MHz transducer (θ = 45°) following double pulse illumination (T = 0.5 ms) of a 9% suspension of red spheres flowing in a 2R = 390 μm tube. The mean fractional error values were calculated for velocities |V| < 50 mm/s and the error bars represent the standard errors.
Figure 8 Comparison of the accuracy of velocity measurements after passing the PA signals through one of various low-pass and high-pass filters. The velocity measurements were acquired using the 30 MHz transducer (θ = 45°) following double pulse illumination (T = 0.5 ms) of a 9% suspension of red spheres flowing in a 2R = 390 μm tube. The mean fractional error values were calculated for velocities |V| < 50 mm/s and the error bars represent the standard errors.CC BY · PubMed Central
Figure 9 Illustration of the effect of laminar flow and light attenuation, resulting in generation of the largest amplitude photoacoustic signals in slow-moving absorbers at the near edge of the tube.
Figure 9 Illustration of the effect of laminar flow and light attenuation, resulting in generation of the largest amplitude photoacoustic signals in slow-moving absorbers at the near edge of the tube.CC BY · PubMed Central
Figure 10 Profiles of velocity measurements corresponding to signal segments shifted progressively along the photoacoustic signal time axis by 50 points (12.5 ns, or 18.45 μm). The measured velocity (y axis) is the mean value calculated for six individual velocity measurements acquired using the 30 MHz transducer (θ = 45°) following double pulse illumination (T = 0.5 ms) of a 9% suspension of red spheres flowing in a 2R = 390 μm tube. The dashed line shows the segment for which the velocity measurements represent the average flow velocity, as illustrated in Fig. 11. This part of the profile corresponds to the flow within the tube and is enlarged to illustrate laminar (parabolic) fits to the data points.
Figure 10 Profiles of velocity measurements corresponding to signal segments shifted progressively along the photoacoustic signal time axis by 50 points (12.5 ns, or 18.45 μm). The measured velocity (y axis) is the mean value calculated for six individual velocity measurements acquired using the 30 MHz transducer (θ = 45°) following double pulse illumination (T = 0.5 ms) of a 9% suspension of red spheres flowing in a 2R = 390 μm tube. The dashed line shows the segment for which the velocity measurements represent the average flow velocity, as illustrated in Fig. 11. This part of the profile corresponds to the flow within the tube and is enlarged to illustrate laminar (parabolic) fits to the data points.CC BY · PubMed Central
Figure 11 Mean velocity measurements made for the segment marked by the red dashed line in Fig. 10, acquired using Method 2 described in connection with Fig. 2. For comparison, the unfilled squares show the original data that was acquired using Method 1 and shown previously in Fig. 7(b).
Figure 11 Mean velocity measurements made for the segment marked by the red dashed line in Fig. 10, acquired using Method 2 described in connection with Fig. 2. For comparison, the unfilled squares show the original data that was acquired using Method 1 and shown previously in Fig. 7(b).CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 21 references

Research works printed but not linked: 3

  1. Patel, B. R. In Fundamentals of Fluid Mechanics (eds. Schetz, J. A. & Fuhs, A. E. ) 486 (John Wiley & Sons, 1999).
  2. Munson, B. R., Young, D. F. & Okiishi, T. H. In Fundamentals of Fluid Mechanics 327–329 (John Wiley & Sons, 2006).
  3. White, F. M. In Fluid Mechanics 341–342 (McGraw-Hill, 1999).

Cited by

See all 34

Metadata

Authors
Brunker, J., & Beard, P.
Affiliations
University College London (GB)
Corresponding author
Brunker, Joanna
Venue
Scientific Reports
Published
2016-02-19
Volume
6
Issue
1
Pages
20902-20902
ISSN
2045-2322 (ISSN-L)
OpenAlex
W2338716815
Cited by
32
References
22
Citations by year
2026: 1 · 2024: 5 · 2023: 2 · 2022: 4 · 2021: 5 · 2020: 3 · 2019: 2 · 2018: 4 · 2017: 2 · 2016: 4
FWCI
1.93
Type
journal article
Language
en
Update notices
erratum (2016-04-20)
Open access
Licence
Locations
  • Scientific Reports (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • UCL Discovery (University College London) (submittedVersion)
Funders
  • European Commission (FP7 ICT, 317744)
  • Engineering and Physical Sciences Research Council (EP/H005536/1, FP7 ICT, Contract 317744)
  • FP7 Information and Communication Technologies (317744)
Topics
  • Physical Sciences › Engineering › Biomedical Engineering › Photoacoustic and Ultrasonic Imaging
  • Physical Sciences › Engineering › Mechanics of Materials › Thermography and Photoacoustic Techniques
  • Physical Sciences › Engineering › Biomedical Engineering › Optical Coherence Tomography Applications
Keywords
blood flow measurement, blood-mimicking fluid, microspheres, detector bandwidth, spatial averaging, high-pass filtering, range gating, light penetration depth, microvascular blood flow, deep tissue imaging
MeSH terms
Acoustics, Algorithms, Blood Flow Velocity, Humans, Models, Cardiovascular, Laser-Doppler Flowmetry, Phantoms, Imaging, Photoacoustic Techniques

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.