Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

In Vivo Venous Assessment of Red Blood Cell Aggregate Sizes in Diabetic Patients with a Quantitative Cellular Ultrasound Imaging Method: Proof of Concept

  • Julien Tripette,

    Affiliation Laboratory of Biorheology and Medical Ultrasonics, Research Center, University of Montreal Hospital (CRCHUM), Québec, Canada

  • Linh-Chi Nguyen,

    Affiliations Laboratory of Biorheology and Medical Ultrasonics, Research Center, University of Montreal Hospital (CRCHUM), Québec, Canada, Institute of Biomedical Engineering, University of Montreal, Québec, Canada

  • Louise Allard,

    Affiliation Laboratory of Biorheology and Medical Ultrasonics, Research Center, University of Montreal Hospital (CRCHUM), Québec, Canada

  • Pierre Robillard,

    Affiliations Department of Radiology, University of Montreal Hospital (CHUM), Québec, Canada, Department of Radiology, Radio-Oncology and Nuclear Medicine, University of Montreal, Québec, Canada

  • Gilles Soulez,

    Affiliations Institute of Biomedical Engineering, University of Montreal, Québec, Canada, Department of Radiology, University of Montreal Hospital (CHUM), Québec, Canada, Department of Radiology, Radio-Oncology and Nuclear Medicine, University of Montreal, Québec, Canada

  • Guy Cloutier

    guy.cloutier@umontreal.ca

    Affiliations Laboratory of Biorheology and Medical Ultrasonics, Research Center, University of Montreal Hospital (CRCHUM), Québec, Canada, Institute of Biomedical Engineering, University of Montreal, Québec, Canada, Department of Radiology, Radio-Oncology and Nuclear Medicine, University of Montreal, Québec, Canada

In Vivo Venous Assessment of Red Blood Cell Aggregate Sizes in Diabetic Patients with a Quantitative Cellular Ultrasound Imaging Method: Proof of Concept

  • Julien Tripette, 
  • Linh-Chi Nguyen, 
  • Louise Allard, 
  • Pierre Robillard, 
  • Gilles Soulez, 
  • Guy Cloutier
PLOS
x

Abstract

Background

Diabetic patients present higher level of red blood cell (RBC) aggregation contributing to the development of vascular complications. While it has been suggested that this hematology/rheology parameter could bring additional prognostic information for the management of those patients, RBC aggregation screening is not included as a clinical practice. Most medical centers are not equipped to measure properly this parameter, although sedimentation tests can bring some indication. Here, we aimed at evaluating the feasibility of using ultrasound to assess in-vivo hyper-aggregation in type 2 diabetic patients.

Research design and methods

Seventeen diabetic patients and 15 control subjects underwent ultrasound measurements of RBC aggregation in both cephalic and great saphenous veins. Non-invasive in-vivo ultrasound measurements were performed using a newly developed cellular imaging technique, the structure factor size and attenuation estimator (SFSAE). Comparisons with an ex-vivo gold standard rheometry technique were done, along with measurements of pro-aggregating plasma molecule concentrations.

Results

In-vivo RBC aggregation was significantly higher in diabetic patients compared with controls for cephalic vein measurements, while a trend (p = 0.055) was noticed in the great saphenous vein. SFSAE measurements were correlated with gold standard in-vitro measures, fibrinogen and C-reactive protein plasma concentrations.

Conclusion

RBC aggregation can be measured in-vivo in diabetic patients using ultrasound. Prospective studies are needed to determine whether the SFSAE method could help clinicians in the early management of vascular complications in this patient population.

Introduction

Red blood cell (RBC) aggregation is a normal and reversible phenomenon influencing blood flow throughout the circulation. The basic structure of RBC aggregates (or “rouleaux”) is an alignment of several cells, very similar to a “stack of coins” [1]. RBC aggregates are likely to form in venules and veins, where flow shear forces are low, and they are dispersed in regions with more important hemodynamic stress. Despite RBC aggregation is considered as a normal phenomenon, hyper-aggregation (which describes bigger aggregates in the form of clumps) is potentially deleterious in a wide range of pathologies, including diabetes, as it impedes blood circulation and decreases tissue perfusion [2, 3].

While the exact mechanisms are still not clearly known, RBC hyper-aggregation is usually attributed to some changes in the plasma concentration of fibrinogen (Fb) and other inflammatory acute proteins (e.g., immunoglobulin G, haptoglobin, C-reactive protein) [47], which modify the interaction between RBC [8]. RBC hyper-aggregation affects flow dynamics and RBC distribution in the microvascular network (e.g., microvessel plugging and shunting), promoting long term impairments of the vasomotor control [9]. For these reasons, hyper-aggregation has been thought to participate in the development of vascular complications such as diabetic angiopathies [1, 10], and some authors suggested that its measure could bring early prognostic information for the management of vascular complications in those patients [1113].

Despite the fact that several in-vitro laboratory techniques can be used to measure RBC aggregation (including blood smear microscopic observation, sedimentation rate, low shear viscosity measurement, laser-assisted optical methods, etc…), they all present some limitations and require blood sampling. The current gold standard method is a laser-assisted scattering technique associated with a Couette flow system, which is able to perform ex-vivo measurements of RBC aggregation kinetics and adhesion forces of aggregates [14, 15]. While this method is able to provide a complete rheological profile of the patient’s blood sample (i.e., aggregation time and aggregate strength), it is still unavailable in most clinical centers and cannot be used to measure RBC aggregation in real time at the bedside.

The present study proposes a new cellular ultrasound imaging technique, the structure factor size and attenuation estimator (SFSAE) [1618], to estimate in-vivo RBC aggregate sizes in diabetic patients non-invasively and in real time. The aim of this paper is to test the hypothesis that the SFSAE technique can be used to detect diabetes-related hyper-aggregation in-situ within superficial veins accessible for ultrasound monitoring purpose.

Methods

Population and protocol

Seventeen diabetic patients (DIAB group) and 15 healthy subjects (CONT group) participated in the study. Diabetic patients had all been diagnosed with type 2 diabetes mellitus for more than 5 years and had oral medication to regulate their glycaemia. Both DIAB and CONT groups were matched in age (53.4 ± 1.7 years vs. 55.2 ± 2.3 years) and gender (6F and 9M vs. 8F and 9M). However, participants presented different average body mass indexes (BMI) (31.4 ± 1.3 vs. 24.2 ± 0.9 kg/m2 for DIAB and CONT, respectively). Among diabetic patients, 10 subjects were diagnosed with Rutherford stage 1 or 2 lower limb peripheral arterial disease [19]. All subjects were nonsmokers and provided a written informed consent. The human ethic committee of the University of Montreal Hospital approved the study.

Recruited subjects came at the Department of Radiology of the University of Montreal Hospital for data collection. After an overnight fast, 30 mL of venous blood was drawn using EDTA, sodium-citrated and dry vacutained tubes. Blood analyses were performed within two hours after collection. Ultrasound measurements of RBC aggregation began about 30 min after subjects took a breakfast.

Biochemical analyses

Plasma concentrations of glycated hemoglobin (HbA1c; G7 HPLC analyzer, Tosoh Bioscience Inc., South San Francisco, California), fibrinogen (BCS XP System, Siemens Healthcare, Erlangen, Germany), immunoglobulin G (IgG; BN II System, Siemens Healthcare), haptoglobin (Hp; BN II System, Siemens Healthcare) and C-reactive protein (CRP; BN II System, Siemens Healthcare) were measured from blood samples with standard laboratory tests.

Hematocrit and ex-vivo RBC aggregation measurements

The hematocrit (Hct) was measured by micro-centrifugation. RBC aggregation time (TA), RBC aggregation kinetics (S10, which reflects the amplitude of the aggregation phenomenon), and the disaggregation threshold (γthr, which reflects aggregate strength) were determined using an ex-vivo laser-based laboratory instrument (Regulest, Florange, France). This method based on a Couette flow formed by two concentric cylinders allows applying shear forces to the blood sample to assess aggregation indices under different hemodynamic conditions. To date, this method is still considered as the gold-standard to measure RBC aggregation in human subjects [14]. Blood sampled in EDTA tubes were used to perform these tests. All measurements were done at 37°C.

In-vivo RBC aggregation assessment

Subjects were in supine position during the whole examination. The cephalic vein in the proximal portion of the forearm (CEP) and the great saphenous vein in the distal portion of the leg (GSV) were scanned with a high frequency ultrasound system (Vevo 770, Visualsonics, Toronto, Canada) equipped with a mono element oscillating probe (RMS-710 transducer, central frequency of 25 MHz). A new cellular imaging mode based on the SFSAE was employed to characterize RBC aggregation, as already used in-vivo in animal models [17, 18]. Venous monitoring was preferred because low shear rates offer favorable conditions for the formation and maintenance of aggregates. The SFSAE utilizes a spectral model that allows extracting the mean aggregate diameter D, from the analysis of radio-frequency (RF) echoes backscattered by blood. D is the ratio of the diameter of a fractal isotropic aggregate to the diameter of one RBC, and it therefore increases proportionally with RBC aggregation. A value of D < or = 1 indicates disaggregated RBC [20]. The SFSAE model compensates for skin and tissue attenuations allowing D to be independent of subject adiposity. RF ultrasound echoes from longitudinal views of each vessel were acquired using a Panametrics receiver (5900 PR, Waltham, MA, USA) and a digital oscilloscope (Gagescope 8500CS, Montreal, QC, Canada), as described in Fig 1A. The RF data corresponds to the raw ultrasound signal that has not been modified by any signal processing normally used to obtain classical B-mode images. The proposed cellular imaging technique is described in details elsewhere [20, 21]. For each subject, a sequence of 25 consecutive vascular images of D was spatially averaged in both the CEP (DCEP) and the GSV (DGSV). A schematic presentation of the experimental setting is described in Fig 1B–1E. This new imaging technique has already been validated under in-vitro and in-vivo conditions (not in human) [17, 18, 2022].

thumbnail
Fig 1. Top: view of the experimental set-up.

(a) Raw radio-frequency (RF) data are acquired with a high-frequency ultrasound device (Vevo 770, Visualsonics, Toronto, Canada). (b) Analog RF data are digitalized and recorded on a computer. (c) B-mode images can be reconstructed for display purpose, and (d) a parametric image of D representing local level of red blood cell (RBC) aggregation can be computed with the SFSAE cellular imaging method. (e) The parametric image is superimposed on the B-mode image after segmentation and calculation of the D index. Finally, local D values temporally and spatially averaged are computed for diagnosis. Bottom: example of images obtained at different levels of aggregation. (f) Low level of RBC aggregation in a control cephalic vein. (g) High level of RBC aggregation in a diabetic great saphenous vein.

https://doi.org/10.1371/journal.pone.0124712.g001

Flow shear rate measurement

Because RBC aggregation is dependent on the shear condition associated with the blood flow, we measured this potential confounding variable at the same locations as RBC aggregation measurements using a clinical ultrasound scanner (Aixplorer, Supersonic Imagine, France), as done elsewhere [18, 22]. B-mode images and Doppler mean velocities (Vmean) were acquired in both CEP and GSV. Vmean was computed by time-averaging the maximum center line velocity measured in pulse-wave Doppler mode over a few cardiac cycles (typically 5). The vein diameter (Di) was measured from B-mode images. Then, the mean shear rate within the vein (γ) was calculated as [23]: where r is the radius of the vessel and n is a constant related to the flow velocity profile. In this study, we supposed that velocity profiles in the veins were parabolic, which is consistent with steady Newtonian flow (i.e., n = 2). The presence of RBC aggregation likely blunted velocity profiles; estimated shear rates could thus be slightly overestimated.

For very low velocities in veins (i.e., Vmean < 0.6 cm/s), the L15-4-38 ultrasound probe (Supersonic Imaging, France) could not detect the time-varying velocity due to the wall filter (the wall filter is an instrument setting avoiding low velocity vibrations from surrounding tissues to corrupt the flow waveform). Since a velocity of 0 cm/s seems unlikely to occur in living veins, we assumed Vmean = 0.6 cm/s for all velocities not detected by the probe.

Statistical analyses

Student t-tests were used to compare measures between groups. One-tail analyses were performed for blood markers as one-way variations were expected. Linear regressions and Pearson tests were performed to determine relations between DCEP and DGSV, and between ultrasound RBC aggregation parameters and blood parameters (i.e., plasma proteins and ex-vivo measurements of RBC aggregation). Analyses were conducted using R (R.app GUI 1.40-devel, R Foundation for Statistical Computing, Vienna, Austria) and Sigma Stat (v. 3.1, San Jose, California, USA) packages. Significant differences referred to p < 0.05. Values are expressed as mean ± SEM.

Results

Biochemical analyses

As shown in Table 1, the DIAB group presented significantly higher HbA1c values compared to the CONT group (7.38 ± 0.29 vs. 5.62 ± 0.05%; p < 0.001). No significant difference was found between groups for IgG. Diabetic patients exhibited higher haptoglobin (CONT: 1.41 ± 0.18 vs. DIAB: 1.96 ± 0.20 g L-1; p = 0.026), CRP (CONT: 2.50 ± 0.85 vs. DIAB: 5.36 ± 0.92 mg L-1; p = 0.017) and fibrinogen concentrations (CONT: 2.92 ± 0.24 vs. DIAB: 3.93 ± 0.12 g L-1; p = 0.039).

Hematocrit and ex-vivo measurements of RBC aggregation

Hct values were statistically similar between groups (Table 1), whereas gold standard laser-assisted optical measurements indicated a higher RBC aggregation in the DIAB group (i.e., lower TA: 1.92 ± 0.10 vs. 2.41 ± 0.16 s in controls, p = 0.016; higher S10: 28.6 ± 1.1 vs. 25.1 ± 1.2 (no unit), p = 0.047; and higher γthr: 193 ± 16 vs. 150 ± 12 s-1, p = 0.046, see Fig 2).

thumbnail
Fig 2. Comparison between control subjects (CONT) and diabetic patients (DIAB) for ex-vivo red blood cell aggregation parameters (TA: aggregation time in second; S10: RBC aggregation kinetics (no unit); γthr: disaggregation threshold in second-1).

Data are presented as Tukey boxplots. ○ describes outliers. * p < 0.05, mean ± SEM.

https://doi.org/10.1371/journal.pone.0124712.g002

In-vivo measurements of RBC aggregation and flow shear rates

As presented in Fig 3, we observed higher ultrasound-based in-vivo and in-situ RBC aggregation in the CEP of diabetic patients (DCEP: 3.20 ± 0.44 vs. 1.49 ± 0.35 (no unit), p = 0.004). CEP shear rates measured with ultrasound were found higher in the DIAB group than in the CONT group (p = 0.024; Table 2). Because higher shear rates tend to disaggregate RBC, this is noticeable that we could nevertheless observe higher aggregation in diabetes. DGSV exhibited a trend for higher RBC aggregation in the DIAB group (DGSV: 3.64 ± 0.49 vs. 2.30 ± 0.35 (no unit); p = 0.055). In-situ shear rates in that vessel were also significantly higher in the DIAB group compared to the CONT group (p = 0.018; Table 2). As also shown in that Table, no differences were found between groups for both CEP and GSV diameters. Flow velocities were higher in the DIAB group for both CEP and GSV (p = 0.020 and p = 0.030, respectively, see Table 2).

thumbnail
Fig 3. Comparison between control subjects (CONT) and diabetic patients (DIAB) for ultrasound in-vivo measurements of RBC aggregation in the cephalic vein CEP (DCEP) and in the great saphenous vein GSV (DGSV).

Data are presented as Tukey boxplots. ○ describes outliers. * p < 0.05, mean ± SEM.

https://doi.org/10.1371/journal.pone.0124712.g003

Correlations between in-vivo RBC aggregation and laboratory tests

A significant correlation was found between DCEP and laser aggregometry γthr values (r = 0.47, p = 0.011), but not between DGSV and γthr (r = 0.32, p = 0.11). No significant correlation was noticed between DCEP and other ex-vivo rheological parameters (TA and S10). DCEP correlated with fibrinogen (r = 0.54, p = 0.003) and CRP plasma concentrations (r = 0.53, p = 0.004). DGSV exhibited a significant correlation with fibrinogen plasma concentrations as well (r = 0.43, p = 0.028).

Discussion

In the present study, we observed higher levels of in-vivo RBC aggregation in diabetic patients compared to control subjects. Bigger aggregates were indeed imaged in the DIAB group under natural venous flow conditions. This study constitutes the first in-vivo quantitative assessment of RBC aggregate sizes in human subjects, and a fortiori in diabetic patients. These in-vivo results correlated with 1) the ex-vivo aggregate strength measurements and 2) pro-aggregating plasma molecule concentrations (cf. fibrinogen and CRP).

Higher RBC aggregation in diabetic patients

No surprisingly, higher RBC aggregation was noted in diabetic patients compared to the control group. The ex-vivo gold standard laser-assisted optical assessments also showed higher RBC aggregation in the DIAB group. The lower aggregation time (cf. TA) and higher magnitude of S10 observed in this group indeed indicate that diabetic erythrocytes have a higher propensity to form aggregates. Moreover, the ex-vivo laser-assisted method showed stronger binding energy between aggregates of diabetic patients compared to control subjects. While these observations have been reported many times under ex-vivo conditions [1012, 24], the present study confirmed for the first time that this phenomenon does actually exist in-vivo. In the cephalic vein, D was higher in diabetic patients indicating larger flowing aggregate sizes. We also noted a strong trend (p = 0.055) in the GSV for the same phenomenon.

Characteristics of in-vivo RBC aggregation in diabetes mellitus

The correlation between DCEP and γthr (r = 0.47, p = 0.011) indicates that the aggregate strength may be the most important determinant for observing in-vivo RBC aggregation (compared to other ex-vivo parameters reflecting the propensity to form aggregates: i.e., TA and S10), at least in the diabetic population. While the interpretation of in-vivo RBC aggregation data traditionally involved analyses of indirect laboratory test indices: e.g., sedimentation rate, low-shear blood viscosity or laser aggregometry measures TA, S10 and γthr (or other similar indices depending on the manufacturer of the instrument) [15], our results suggest that future direct assessment of in-situ RBC aggregation is possible, especially in the diabetic population.

In the present study, DCEP correlated with fibrinogen and CRP plasma concentrations (r = 0.54, p = 0.003; r = 0.53, p = 0.004, respectively), which are known promoters of RBC aggregation [57]. DGSV exhibited a significant correlation with fibrinogen plasma concentration as well (r = 0.43, p = 0.028). The relation between high levels of fibrinogen and stronger RBC aggregates has already been shown [25], supporting the present association between high inflammatory acute-phase protein concentrations, elevated ex-vivo RBC aggregate strength, and higher in-vivo RBC aggregation in diabetic patients. The higher in-vivo RBC aggregation found concomitantly with higher shear rates (favoring the breaking down of reversible aggregates) in the diabetic population, for both cephalic and great saphenous veins, also support the hypothesis that enhanced RBC aggregating energies were determinant for explaining reported observations. The exact mechanism for the higher shear rate in veins of diabetic patients compared to the control population is unknown but might be associated with 1) a hemodynamic compensation phenomenon to avoid the deleterious effect of large flowing aggregates (inducing high blood viscosity) on the venous return, or 2) an arteriovenous shunting phenomenon that has already been described elsewhere [26]. We indeed observed higher mean velocities in the diabetic population (Table 2).

The validity of the SFSAE cellular imaging measures is supported by previous ex-vivo microscopic observations performed under dynamic conditions, which highlighted the spherical shape of compact aggregates in diabetes mellitus compared to normal cylindrical “rouleaux” observed in healthy subjects [27, 28]. Indeed, SFSAE images are based on a model assuming flowing RBC aggregates of spherical forms [16]. Therefore, from a physical point of view, elevated values of D indicate larger circulating spherical clusters. In the case of flowing rouleaux in normal subjects (likely rotating with the flow), a biased overestimation of the aggregate size is expected with this model. Consequently, the basic assumption behind the SFSAE imaging method did not affect the interpretation of results. To summarize, the SFSAE-based in-vivo measurements of RBC aggregation revealed that circulating diabetic RBC aggregates were bigger and presented a stronger resistance to shear rates compared to their counterpart in age-matched control subjects.

Clinical perspectives

Hyper-aggregation of erythrocytes has been described in several pathologies, including HIV infection, myocardial infarction, sepsis, stroke, venous thrombosis and diabetes mellitus, to name a few examples [3]. However, it remains unclear whether it is caused by the pathology (through the action of inflammatory factors, oxidative stress agents or vascular alteration) or whether it contributes to the etiology of the pathology, and its associated vascular complications. Indeed, an increase in RBC aggregation may impede blood circulation and decreases tissue perfusion, subsequently sustaining the pathological process. The in-vivo hyper-aggregation observed in diabetic patients in both CEP and GSV might reflect some blood rheology disturbances in the upstream microvasculature. In diabetes mellitus, early hemorheological alterations have been linked to the development of vascular complications, like retinopathy, nephropathy, lower limb ischemia, brain ischemia, hypertension and atherosclerosis [11, 12, 29, 30]. A prospective clinical trial suggested that the level of RBC aggregation offers a high positive prediction for diabetic foot syndrome deterioration comparable to that associated with transcutaneous oxygen pressure, which is a widely used biomarker in clinical practice [11]. In light of these results, it was proposed to include hemorheological parameters in the screening of patients who presented a risk for foot ulceration, in order to obtain additional prognostic information. Because ex-vivo measurements of RBC aggregation require specific equipments that are unavailable in most clinical centers, the SFSAE cellular imaging technique, which can be implemented on a clinical ultrasound scanner, could represent a promising alternative to introduce rheological parameters in the management of diabetic patients. In this case, the targeted population would rather be patients in the early phase of diabetes mellitus development rather than patients with already confirmed vascular complications like in the present pilot study. Further larger scale and prospective studies have to be conducted in non-severe patients in order to address the predictive effectiveness of the proposed in-vivo RBC aggregation measurements.

Limitations and summary

The SFSAE cellular imaging method allows reliable, non-invasive and in-vivo measurements of RBC aggregation using ultrasound. It has been conclusively used in preclinical animal studies on superficial or surgically exposed veins [17, 18]. According to the physics of ultrasound, biological tissue attenuation at 25 MHz limits measurements to superficial vessels (at approximately 1 cm deep or smaller). This may be seen as a limitation; nevertheless, because no differences in the RBC aggregation level were noticed between the cephalic and great saphenous veins for the studied population, measurements over the arm would be possible if severe edema precludes measurements over the patient’s feet. In-vivo hemorheological profile at the bedside thus seems feasible with the proposed imaging method.

Another limitation of the present study is that DGSV only presented a trend (p = 0.055) for differences between groups. Specific flowing conditions could have impaired measurements in this foot area. Indeed, arteriovenous shunting has been proposed as a key factor of foot gangrene development in diabetic patients [26]. Such phenomena may have promoted unfavorable high shear flow conditions for the formation of RBC aggregates in this group. Moreover, a complex hemodynamic environment with several venous bifurcations was observed in our study and likely impacted RBC aggregate formation by disturbed flow at branching points, thus reducing in-vivo SFSAE measures of the aggregate size in diabetic patients. At the opposite, CEP segments were rectilinear, and could thus favor aggregate formation by inflammatory plasma proteins. Further studies will have to address this point further to clarify the measurement methodology.

To summarize, the present study confirmed for the first time the presence of larger RBC aggregates in diabetic patients under in-vivo flowing conditions. The SFSAE imaging method might be included in the management of diabetic patients to add prognostic information about the development of vascular complications. Further larger scale prospective studies would be required to address this latter point.

Acknowledgments

Funding of this study was provided by the Canadian Institutes of Health Research (MOP-84358). Authors have special thanks to Jocelyne Lavoie, Zhao Qin, Andrée Cliche, Hasti Valiallah, Michel Gouin and Boris Chayer for their technical assistance.

Author Contributions

Conceived and designed the experiments: JT LCN LA GS GC. Performed the experiments: JT LCN PR. Analyzed the data: JT LCN LA. Contributed reagents/materials/analysis tools: PR GC. Wrote the paper: JT LCN LA PR GS GC.

References

  1. 1. Baskurt OK, Meiselman HJ. RBC aggregation: More important than RBC adhesion to endothelial cells as a determinant of in vivo blood flow in health and disease. Microcirculation 2008;15: 585–590. pmid:18608991
  2. 2. Le Dévéhat C, Vimeux M, Khodabandehlou T. Blood rheology in patients with diabetes mellitus. Clin Hemorheol Microcirc 2004;30(3–4): 297–300. pmid:15258379
  3. 3. Meiselman HJ, Baskurt OK. Hemorheology and hemodynamics: Dove andare? Clin Hemorheol Microcirc 2006;35: 37–43. pmid:16899904
  4. 4. Khodabandehlou T, Le Devéhat C, Razavian M. Impaired function of fibrinogen: Consequences on red cell aggregation in diabetes mellitus. Clin Hemorheol 1996;16(3): 303–312.
  5. 5. Weng X, Cloutier G, Beaulieu R, Roederer GO. Influence of acute-phase proteins on erythrocyte aggregation. Am J Physiol 1996;271(6 Pt 2): H2346–H2352. pmid:8997292
  6. 6. Weng X, Roederer GO, Beaulieu R, Cloutier G. Contribution of acute-phase proteins and cardiovascular risk factors to erythrocyte aggregation in normolipidemic and hyperlipidemic individuals. Thromb Haemost 1998;80(6): 903–908. pmid:9869158
  7. 7. Saldanha C. Fibrinogen interaction with the red blood cell membrane. Clin Hemorheol Microcirc 2013;53(1–2): 39–44. pmid:23246814
  8. 8. Neu B, Meiselman HJ. Red blood cell aggregation. Handbook of Hemorheology and Hemodynamic. Edited by Baskurt OK, Hardeman MR, Rampling MW, Meiselman HJ. IOS Press, 2007: 114–136.
  9. 9. McHedlishvili G, Maeda N. Blood flow structure related to red cell flow: determinant of blood fluidity in narrow microvessels. Jpn J Physiol 2001;51(1): 19–30. pmid:11281993
  10. 10. Le Devéhat C, Vimeux M, Bondoux, Khodabandehlou T. Red blood cell aggregation in diabetes mellitus. Int Angiol 1990;9(1): 11–14. pmid:2212794
  11. 11. Khodabandehlou T, Le Dévéhat C. Hemorheological disturbances as a marker of diabetic foot syndrome deterioration. Clin Hemorheol Microcirc 2004;30: 219–223. pmid:15258346
  12. 12. Khodabandehlou T, Vimeux M, Le Dévéhat C. Measurements of transcutaneous oxygen pressure and changes in blood rheology as markers of prognosis of critically ischemic limb in diabetes mellitus patients. Int J Low Extrem Wounds 2003;2(1): 13–18. pmid:15866822
  13. 13. Mantskava M, Momtselidze N, Pargalava N, Mchedlishvili G. Hemorheological disorders in patients with type 1 or 2 diabetes mellitus and foot gangrene. Clin Hemorheol Microcirc 2006;35(1–2): 307–310.
  14. 14. Baskurt OK, Boynard M, Cokelet GC, Connes P, Cooke BM, Forconi S, et al. International Expert Panel for Standardization of Hemorheological Methods. New guidelines for hemorheological laboratory techniques. Clin Hemorheol Microcirc 2009;42(2): 75–97. pmid:19433882
  15. 15. Hardeman MR, Goedhart PT, Shin S. Methods in hemorheology. Handbook of Hemorheology and Hemodynamic. Edited by Baskurt OK, Hardeman MR, Rampling MW, Meiselman HJ. IOS Press, 2007: 242–266.
  16. 16. Franceschini E, Yu FT, Cloutier G. Simultaneous estimation of attenuation and structure parameters of aggregated red blood cells from backscatter measurements. J Acoust Soc Am 2008;123(4): EL85–EL91. pmid:18396926
  17. 17. Yu FT, Armstrong JK, Tripette J, Meiselman HJ, Cloutier G. A local increase in red blood cell aggregation can trigger deep vein thrombosis: evidence based on quantitative cellular ultrasound imaging. J Thromb Haemost 2011;9(3): 481–488. pmid:21143377
  18. 18. Tripette J, Denault AY, Allard L, Chayer B, Perrault LP, Cloutier G. Ultrasound monitoring of RBC aggregation as a real-time marker of the inflammatory response in a cardiopulmonary bypass swine model. Crit Care Med 2013;41(8): e171–e178. pmid:23648566
  19. 19. Rutherford RB, Baker JD, Ernst C, Johnston KW, Porter JM, Ahn S, et al. Recommended standards for reports dealing with lower extremity ischemia: revised version. J Vasc Surg 1997;26: 517–538. pmid:9308598
  20. 20. Yu FTH, Cloutier G. Experimental ultrasound characterization of red blood cell aggregation using the structure factor size estimator. J Acoust Soc Am 2007;122: 645–656. pmid:17614521
  21. 21. Franceschini E, Yu FT, Destrempes F, Cloutier G. Ultrasound characterization of red blood cell aggregation with intervening attenuating tissue-mimicking phantoms. J Acoust Soc Am 2010;127: 1104–1115. pmid:20136231
  22. 22. Yu FT, Franceschini E, Chayer B, Armstrong JK, Meiselman HJ, Cloutier G. Ultrasonic parametric imaging of erythrocyte aggregation using the structure factor size estimator. Biorheology. 2009;46(4): 343–63. pmid:19721194
  23. 23. Cloutier G, Qin Z, Durand LG, Teh BG. Power Doppler ultrasound evaluation of the shear rate and shear stress dependences of red blood cell aggregation, IEEE Trans Biomed Eng 1996;43(5): 441–450. pmid:8849457
  24. 24. Cloutier G, Zimmer A, Yu FT, Chiasson JL. Increased shear rate resistance and fastest kinetics of erythrocyte aggregation in diabetes measured with ultrasound. Diabetes Care 2008;31(7): 1400–1402. pmid:18375419
  25. 25. Maharshak N, Arbel Y, Shapira I, Berliner S, Ben-Ami R, Yedgar S, et al. Increased strength of erythrocyte aggregates in blood of patients with inflammatory bowel disease. Inflamm Bowel Dis 2009;15(5): 707–13. pmid:19137610
  26. 26. Gabbay IE, Gabbay M, Gabbay U. Diabetic foot cellular hypoxia may be due to capillary shunting—a novel hypothesis. Med Hypotheses 2014;82(1): 57–9. pmid:24280559
  27. 27. Schmid-Schonbein H, Malotta H, Striesow F. Erythrocyte aggregation: Causes, consequences and methods of assessment. Tijdschr NVKS 1990;15: 88–97.
  28. 28. Foresto P, D’Arrigo M, Carreras L, Cuezzo RE, Valverde J, Rasia R. Evaluation of red blood cell aggregation in diabetes by computerized image analysis. Medicina (Buenos Aires) 2000;60: 570–572.
  29. 29. Momtselidze N, Mantskava M, Mchedlishvili G. Hemorheological disorders during ischemic brain infarcts in patients with and without diabetes mellitus. Clin Hemorheol Microcirc 2006;35(1–2): 261–264.
  30. 30. Rosenson RS, Fioretto P, Dodson PM. Does microvascular disease predict macrovascular events in type 2 diabetes? Atherosclerosis 2011;218(1): 13–18. pmid:21763654