Cyclic variations of high-frequency ultrasonic backscattering from blood under pulsatile flow

Diagnosis of diabetes mellitus using high frequency ultrasound and convolutional neural network

The incidence of diabetes mellitus has been increasing, prompting the search for non-invasive diagnostic methods. Although current methods exist, these have certain limitations, such as low reliability and accuracy, difficulty in individual patient adjustment, and discomfort during use. This paper presents a novel approach for diagnosing diabetes using high-frequency ultrasound (HFU) and a convolutional neural network (CNN). This method is based on the observation that glucose in red blood cells (RBCs) forms glycated hemoglobin (HbA1c) and accumulates on its surface. The study incubated RBCs with different glucose concentrations, collected acoustic reflection signals from them using a custom-designed 90-MHz transducer, and analyzed the signals using a CNN. The CNN was applied to the frequency spectra and spectrograms of the signal to identify correlations between changes in RBC properties owing to glucose concentration and signal features. The results confirmed the efficacy of the CNN-based approach with a classification accuracy of 0.98. This non-invasive diagnostic technology using HFU and CNN holds promise for in vivo diagnosis without the need for blood collection.

Parallelized ultrasound homodyned-K imaging based on a generalized artificial neural network estimator

The homodyned-K (HK) distribution model is a generalized backscatter envelope statistical model for ultrasound tissue characterization, whose parameters are of physical meaning. To estimate the HK parameters is an inverse problem, and is quite complicated. Previously, we proposed an artificial neural network (ANN) estimator and an improved ANN (iANN) estimator for estimating the HK parameters, which are fast and flexible. However, a drawback of the conventional ANN and iANN estimators consists in that they use Monte Carlo simulations under known values of HK parameters to generate training samples, and thus the ANN and iANN models have to be re-trained when the size of the test sets (or of the envelope samples to be estimated) varies. In addition, conventional ultrasound HK imaging uses a sliding window technique, which is non-vectorized and does not support parallel computation, so HK image resolution is usually sacrificed to ensure a reasonable computation cost. To this end, we proposed a generalized ANN (gANN) estimator in this paper, which took the theoretical derivations of feature vectors for network training, and thus it is independent from the size of the test sets. Further, we proposed a parallelized HK imaging method that is based on the gANN estimator, which used a block-based parallel computation method, rather than the conventional sliding window technique. The gANN-based parallelized HK imaging method allowed a higher image resolution and a faster computation at the same time. Computer simulation experiments showed that the gANN estimator was generally comparable to the conventional ANN estimator in terms of HK parameter estimation performance. Clinical experiments of hepatic steatosis showed that the gANN-based parallelized HK imaging could be used to visually and quantitatively characterize hepatic steatosis, with similar performance to the conventional ANN-based HK imaging that used the sliding window technique, but the gANN-based parallelized HK imaging was over 3 times faster than the conventional ANN-based HK imaging. The parallelized computation method presented in this work can be easily extended to other quantitative ultrasound imaging applications.

Ultrasound simulation of blood with different red blood cell aggregations and concentrations

BACKGROUND

Considerable progress of ultrasound simulation on blood has enhanced the characterizing of red blood cell (RBC) aggregation.

OBJECTIVE

A novel simulation method aims at modeling the blood with different RBC aggregations and concentrations is proposed.

METHODS

The modeling process is as follows: (i) A three-dimensional scatterer model is first built by a mapping with a Hilbert space-filling curve from the one-dimensional scatterer distribution. (ii) To illustrate the relationship between the model parameters and the RBC aggregation level, a variety of blood samples are prepared and scanned to acquire their radiofrequency signals in-vitro. (iii) The model parameters are determined by matching the Nakagami-distribution characteristics of envelope signals simulated from the model with those measured from the blood samples.

RESULTS

Nakagami metrics m estimated from 15 kinds of blood samples (hematocrits of 20%, 40%, 60% and plasma concentrations of 15%, 30%, 45%, 60%, 75%) are compared with metrics estimated by their corresponding models (each with different eligible parameters). Results show that for the three hematocrit levels, the mean and standard deviation of the root-mean-squared deviations of m are 0.27 ± 0.0026, 0.16 ± 0.0021, 0.12 ± 0.0018 respectively.

CONCLUSION

The proposed simulation model provides a viable data source to evaluate the performance of the ultrasound-based methods for quantifying RBC aggregation.

Classification of red blood cell aggregation using empirical wavelet transform analysis of ultrasonic radiofrequency echo signals

Grading red blood cell (RBC) aggregation is important for the early diagnosis and prevention of related diseases such as ischemic cardio-cerebrovascular disease, type II diabetes, deep vein thrombosis, and sickle cell disease. In this study, a machine learning technique based on an adaptive analysis of ultrasonic radiofrequency (RF) echo signals in blood is proposed, and its feasibility for classifying RBC aggregation is explored. Using an adaptive empirical wavelet transform (EWT) analysis, the ultrasonic RF signals are decomposed into a series of empirical mode functions (EMFs); then, dominant empirical mode functions (DEMFs) are selected from the series. Six statistical characteristics, including the mean, variance, median, kurtosis, root mean square (RMS), and skewness are calculated for the locally normalized DEMFs, aiming to form primary feature vectors. Random forest (RDF) and support vector machine (SVM) classifiers are trained with the given feature vectors to obtain prediction models for RBC classification. Ultrasonic RF echo signals are acquired from five groups of six types of porcine blood samples with average numbers of aggregated RBCs of 1.04, 1.20, 1.83, 2.31, 2.72, and 4.28, respectively, to test the classification performance of the proposed method. The best subset with regard to the variance, kurtosis, and RMS is determined according to the maximum accuracy based on the RDF and SVM classifiers. The classification accuracies are 84.03 ± 3.13% for the RDF classifier, and 85.88 ± 2.99% for the SVM classifier. The mean classification accuracy of the SVM classifier is 1.85% better than that of the RDF classifier. In conclusion, the machine learning method is useful for the discrimination of varying degrees of RBC aggregation, and has potential for use in characterizing and monitoring the RBC aggregation in vessels.

In vitro photoacoustic spectroscopy of pulsatile blood flow: Probing the interrelationship between red blood cell aggregation and oxygen saturation

We investigate the optical wavelength dependence in quantitative photoacoustic (QPA) assessment of red blood cell (RBC) aggregation and oxygen saturation (sO₂ ) during pulsatile blood flow. Experimentally, the pulsatile flow was imaged with a 700 to 900 nm laser using the VevoLAZR. Theoretically, the photoacoustic (PA) signals were computed based on a Green's function integrated with a Monte Carlo simulation of radiant fluence. The pulsatile flow created periodic conditions of RBC aggregation/nonaggregation, altering the aggregate size, and, in turn, the sO₂ . The dynamic range, DR (a metric of change in PA power) from 700 to 900 nm for nonaggregated RBCs, was 5 dB for both experiment and theory. A significant difference in the DR for aggregated RBCs was 1.5 dB between experiment and theory. Comparing the DR at different wavelengths, the DR from nonaggregated to aggregated RBCs at 700 nm was significantly smaller than that at 900 nm for both experiment (4.0 dB < 7.1 dB) and theory (5.3 dB < 9.0 dB). These results demonstrate that RBC aggregation simultaneously affects the absorber size and the absorption coefficient in photoacoustic imaging (PAI) of pulsatile blood flow. This investigation elucidates how QPA spectroscopy can be used for probing hemodynamics and oxygen transport by PAI of blood flow.

Quantitative Measurement of Erythrocyte Aggregation as a Systemic Inflammatory Marker by Ultrasound Imaging: A Systematic Review

This systematic review is aimed at answering two questions: (i) Is erythrocyte aggregation a useful biomarker in assessing systemic inflammation? (ii) Does quantitative ultrasound imaging provide the non-invasive option to measure erythrocyte aggregation in real time? The search was executed through bibliographic electronic databases CINAHL, EMB Review, EMBASE, MEDLINE, PubMed and the grey literature. The majority of studies correlated elevated erythrocyte aggregation with inflammatory blood markers for several pathologic states. Some studies used "erythrocyte aggregation" as an established marker of systemic inflammation. There were limited but promising articles regarding the use of quantitative ultrasound spectroscopy to monitor erythrocyte aggregation. Similarly, there were limited studies that used other ultrasound techniques to measure systemic inflammation. The quantitative measurement of erythrocyte aggregation has the potential to be a routine clinical marker of inflammation as it can reflect the cumulative inflammatory dynamics in vivo, is relatively simple to measure, is cost-effective and has a rapid turnaround time. Technologies like quantitative ultrasound spectroscopy that can measure erythrocyte aggregation non-invasively and in real time may offer the advantage of continuous monitoring of the inflammation state and, thus, may help in rapid decision making in a critical care setup.

Simultaneous assessment of red blood cell aggregation and oxygen saturation under pulsatile flow using high-frequency photoacoustics

We investigate the feasibility of photoacoustic (PA) imaging for assessing the correlation between red blood cell (RBC) aggregation and the oxygen saturation (sO2) in a simulated pulsatile blood flow system. For the 750 and 850 nm illuminations, the PA amplitude (PAA) increased and decreased as the mean blood flow velocity decreased and increased, respectively, at all beat rates (60, 120 and 180 bpm). The sO2 also cyclically varied, in phase with the PAA for all beat rates. However, the linear correlation between the sO2 and the PAA at 850 nm was stronger than that at 750 nm. These results suggest that the sO2 can be correlated with RBC aggregation induced by decreased mean shear rate in pulsatile flow, and that the correlation is dependent on the optical wavelength. The hemodynamic properties of blood flow assessed by PA imaging may be used to provide a new biomarker for simultaneous monitoring blood viscosity related to RBC aggregation, oxygen delivery related to the sO2 and their clinical correlation.

Measurement of the Doppler power of flowing blood using ultrasound Doppler devices

Measurement of the Doppler power of signals backscattered from flowing blood (henceforth referred to as the Doppler power of flowing blood) and the echogenicity of flowing blood have been used widely to assess the degree of red blood cell (RBC) aggregation for more than 20 y. Many studies have used Doppler flowmeters based on an analogue circuit design to obtain the Doppler shifts in the signals backscattered from flowing blood; however, some recent studies have mentioned that the analogue Doppler flowmeter exhibits a frequency-response problem whereby the backscattered energy is lost at higher Doppler shift frequencies. Therefore, the measured Doppler power of flowing blood and evaluations of RBC aggregation obtained using an analogue Doppler device may be inaccurate. To overcome this problem, the present study implemented a field-programmable gate array-based digital pulsed-wave Doppler flowmeter to measure the Doppler power of flowing blood, in the aim of providing more accurate assessments of RBC aggregation. A clinical duplex ultrasound imaging system that can acquire pulsed-wave Doppler spectrograms is now available, but its usefulness for estimating the ultrasound scattering properties of blood is still in doubt. Therefore, the echogenicity and Doppler power of flowing blood under the same flow conditions were measured using a laboratory pulser-receiver system and a clinical ultrasound system, respectively, for comparisons. The experiments were carried out using porcine blood under steady laminar flow with both RBC suspensions and whole blood. The experimental results indicated that a clinical ultrasound system used to measure the Doppler spectrograms is not suitable for quantifying Doppler power. However, the Doppler power measured using a digital Doppler flowmeter can reveal the relationship between backscattering signals and the properties of blood cells because the effects of frequency response are eliminated. The measurements of the Doppler power and echogenicity of flowing blood were compared with those obtained in several previous studies.

In vivo observation of the hypo-echoic "black hole" phenomenon in rat arterial bloodstream: a preliminary Study

The "black hole," a hypo-echoic hole at the center of the bloodstream surrounded by a hyper-echoic zone in cross-sectional views, has been observed in ultrasound backscattering measurements of blood with red blood cell aggregation in in vitro studies. We investigated whether the phenomenon occurs in the in vivo arterial bloodstream of rats using a high-frequency ultrasound imaging system. Longitudinal and cross-sectional ultrasound images of the rat common carotid artery (CCA) and abdominal aorta were obtained using a 40-MHz ultrasound system. A high-frame-rate retrospective imaging mode was employed to precisely examine the dynamic changes in blood echogenicity in the arteries. When the imaging was performed with non-invasive scanning, blood echogenicity was very low in the CCA as compared with the surrounding tissues, exhibiting no hypo-echoic zone at the center of the vessel. Invasive imaging of the CCA by incising the skin and subcutaneous tissues at the imaging area provided clearer and brighter blood echo images, showing the "black hole" phenomenon near the center of the vessel in longitudinal view. The "black hole" was also observed in the abdominal aorta under direct imaging after laparotomy. The aortic "black hole" was clearly observed in both longitudinal and cross-sectional views. Although the "black hole" was always observed near the center of the arteries during the diastolic phase, it dissipated or was off-center along with the asymmetric arterial wall dilation at systole. In conclusion, we report the first in vivo observation of the hypo-echoic "black hole" caused by the radial variation of red blood cell aggregation in arterial bloodstream.

High spatial and temporal resolution observations of pulsatile changes in blood echogenicity in the common carotid artery of rats

Previous studies have found that ultrasound backscatter from blood in vascular flow systems varies under pulsatile flow, with the maximum values occurring during the systolic period. This phenomenon is of particular interest in hemorheology because it is contrary to the well-known fact that red blood cell (RBC) aggregation, which determines the intensity of ultrasound backscatter from blood, decreases at a high systolic shear rate. In the present study, a rat model was used to provide basic information on the characteristics of blood echogenicity in arterial blood flow to investigate the phenomenon of RBC aggregation under pulsatile flow. Blood echogenicity in the common carotid arteries of rats was measured using a high-frequency ultrasound imaging system with a 40-MHz probe. The electrocardiography-based kilohertz visualization reconstruction technique was employed to obtain high-temporal-resolution and high-spatial-resolution time-course B-mode cross-sectional and longitudinal images of the vessel. The experimental results indicate that blood echogenicity in rat carotid arteries varies during a cardiac cycle. Blood echogenicity tends to decrease during early systole and reaches its peak during late systole, followed by a slow decline thereafter. The time delay of the echogenicity peak from peak systole in the present results is the main difference from previous in vitro and in vivo observations of backscattering peaks during early systole, which may be caused by the very rapid heart rates and low RBC aggregation tendency of rats compared with humans and other mammalian species. The present study may provide useful information elucidating the characteristics of RBC aggregation in arterial blood flow.

The effect of flow acceleration on the cyclic variation of blood echogenicity under pulsatile flow

It has been shown that the echogenicity of blood varies during a flow cycle under pulsatile flow both in vitro and in vivo. In general, the echogenicity of flowing whole blood increases during the early systole phase and then reduces to a minimum at late diastole. While it has been postulated that this cyclic variation is associated with the dynamics of erythrocyte aggregation, the mechanisms underlying this increasing echogenicity with flow velocity remain uncertain. The effect of flow acceleration has also been proposed as an explanation for this phenomenon, but no specific experiments have been conducted to test this hypothesis. In addition, the influence of ultrasonic attenuation on the cyclic variation of echogenicity requires clarification. In the present study, a Couette flow system was designed to simulate blood flowing with different acceleration patterns, and the flow velocity, attenuation, and backscattering coefficient were measured synchronously from 20%- and 40%-hematocrit porcine whole blood and erythrocyte suspensions using 35-MHz ultrasound transducers. The results showed ultrasonic attenuation exerted only minor effects on the echogenicity of blood under pulsatile flow conditions. Cyclic variations of echogenicity were clearly observed for whole blood with a hematocrit of 40%, but no variations were apparent for erythrocyte suspensions. The echogenicity did not appear to be enhanced when instantaneous acceleration was applied to flowing blood in any case. These findings show that flow acceleration does not promote erythrocyte aggregation, even when a higher peak velocity is applied to the blood. Comparison of the results obtained with different accelerations revealed that the cyclic variation in echogenicity observed during pulsatile blood flow may be jointly attributable to the effect of shear rate and the distribution of erythrocyte on aggregation.

Design and implementation of a smartphone-based portable ultrasound pulsed-wave Doppler device for blood flow measurement

Blood flow measurement using Doppler ultrasound has become a useful tool for diagnosing cardiovascular diseases and as a physiological monitor. Recently, pocket-sized ultrasound scanners have been introduced for portable diagnosis. The present paper reports the implementation of a portable ultrasound pulsed-wave (PW) Doppler flowmeter using a smartphone. A 10-MHz ultrasonic surface transducer was designed for the dynamic monitoring of blood flow velocity. The directional baseband Doppler shift signals were obtained using a portable analog circuit system. After hardware processing, the Doppler signals were fed directly to a smartphone for Doppler spectrogram analysis and display in real time. To the best of our knowledge, this is the first report of the use of this system for medical ultrasound Doppler signal processing. A Couette flow phantom, consisting of two parallel disks with a 2-mm gap, was used to evaluate and calibrate the device. Doppler spectrograms of porcine blood flow were measured using this stand-alone portable device under the pulsatile condition. Subsequently, in vivo portable system verification was performed by measuring the arterial blood flow of a rat and comparing the results with the measurement from a commercial ultrasound duplex scanner. All of the results demonstrated the potential for using a smartphone as a novel embedded system for portable medical ultrasound applications.

The acute effects of smoking on the cyclic variations in blood echogenicity of carotid artery

The objective of this research is to study the cyclic variations in echogenicity (CVE) as an acute response to smoking. CVEs, caused by the aggregation of red blood cells (RBC) were measured from the cross-sectional images of the common carotid artery using coded harmonic imaging of a commercial ultrasound system. The amplitude of the CVE (A(cve)) was analyzed among 28 smokers before and after smoking. A(cve) was increased in 22 smokers and decreased in six smokers after 1-2 cigarettes were smoked. Heart rate (HR) was also estimated from the ultrasonic images before and after smoking. The smokers were optimally divided into two clusters with respect to the change in A(cve) and the intrinsic characteristics of smokers (i.e., daily consumed cigarettes and smoking years) through a two-step cluster analysis (TSCA). The increase in A(cve) after smoking was significantly higher in the heavy smoker cluster compared with the light smoker cluster. The results suggest that the acute changes in A(cve) in response to smoking are different between heavy smokers and light smokers. This preliminary study demonstrates the potential application of coded harmonic ultrasound imaging to detect or characterize RBC aggregation. In addition, the results may be useful for understanding the acute physiologic changes caused by smoking.

Ultrasonic attenuation and backscatter from flowing whole blood are dependent on shear rate and hematocrit between 10 and 50 MHz

Ultrasonic backscatter has recently been used extensively to investigate erythrocyte aggregation, which is an inherent hematological phenomenon in the blood circulation system. The size of rouleaux can be estimated by measuring certain parameters of signals backscattered from flowing blood. However, most measurements of backscatter from blood use a constant value for the attenuation coefficient to compensate for the loss of ultrasound energy. This correction may be inaccurate because the attenuation varies with the blood properties, which prompted us to explore the effects of hemodynamic properties on ultrasonic attenuation and backscatter to better understand the blood rheological behaviors. Experiments were performed on porcine whole blood in a Couette flow apparatus. Ultrasonic attenuation and the backscattering coefficient of blood were measured at various frequencies (from 10 to 50 MHz), hematocrits (from 0 to 60%), and shear rates (from 0.1 to 200 s⁻¹). The results indicated that the attenuation and backscattering coefficients of blood are highly variable, depending in a complex manner on shear rate, hematocrit, and the measurement ultrasound frequency. The attenuation of blood decreased rapidly with increasing shear rates, eventually reaching a steady state asymptotically, and increased linearly with the hematocrit from 10 to 50 MHz at various shear rates, and also with the ultrasound frequency. The effect of erythrocyte aggregation means that the change in ultrasonic attenuation in blood with shear rate may be attributed to the absorption mechanism, which is enhanced by the increased blood viscosity at lower shear rates. Compensating the measured backscattering coefficients of blood for the shear-rate-dependent attenuation coefficient increased the accuracy of erythrocyte aggregation assessments. Together, the experimental results suggest that the shear-rate-dependent attenuation coefficient should be considered in future developments of ultrasonic technologies for characterizing blood rheology when the ultrasound frequency is higher than 20 MHz.

High-frequency attenuation and backscatter measurements of rat blood between 30 and 60 MHz

There has recently been a great deal of interest in noninvasive high-frequency ultrasound imaging of small animals such as rats due to their being the preferred animal model for gene therapy and cancer research. Improving the interpretation of the obtained images and furthering the development of the imaging devices require a detailed knowledge of the ultrasound attenuation and backscattering of biological tissue (e.g. blood) at high frequencies. In the present study, the attenuation and backscattering coefficients of the rat red blood cell (RBC) suspensions and whole blood with hematocrits ranging from 6% to 40% were measured between 30 and 60 MHz using a modified substitution approach. The acoustic parameters of porcine blood under the same conditions were also measured in order to compare differences in the blood properties between these two animals. For porcine blood, both whole blood and RBC suspension were stirred at a rotation speed of 200 rpm. Three different rotation speeds of 100, 200 and 300 rpm were carried out for rat blood experiments. The attenuation coefficients of both rat and porcine blood were found to increase linearly with frequency and hematocrit (the values of coefficients of determination (r(2)) are around 0.82-0.97 for all cases). The average attenuation coefficient of rat whole blood with a hematocrit of 40% increased from 0.26 Nepers mm(-1) at 30 MHz to 0.47 Nepers mm(-1) at 60 MHz. The maximum backscattering coefficients of both rat and porcine RBC suspensions were between 10% and 15% hematocrits at all frequencies. The fourth-power dependence of backscatter on frequency was approximately valid for rat RBC suspensions with hematocrits between 6% and 40%. However, the frequency dependence of the backscatter estimate deviates from a fourth-power law for porcine RBC suspension with hematocrit higher than 20%. The backscattering coefficient plateaued for hematocrits higher than 15% in porcine blood, but for rat blood it was maximal around a hematocrit of 20% at the same rotation speed, and shifted to a hematocrit of 10% at a higher speed. The backscattering properties of rat RBCs in plasma are similar to those of RBCs in saline at a higher rotation speed. The differences in attenuation and backscattering between rat and porcine blood may be attributed to RBCs' being smaller and the RBC aggregation level being lower for rat blood than for porcine blood.