Ratan K. Saha

Detection and characterization of red blood cell (RBC) aggregation with photoacoustics

Red blood cells (RBCs) aggregate in the presence of increased plasma fibrinogen and low shear forces during blood flow. RBC aggregation has been observed in deep vein thrombosis, sepsis and diabetes. We propose using photoacoustics (PA) as a non-invasive imaging modality to detect RBC aggregation. The theoretical and experimental feasibility of PA for detecting and characterizing aggregation was assessed. A simulation study was performed to generate PA signals from non-aggregated and aggregated RBCs using a frequency domain approach and to study the PA signals dependence on hematocrit and aggregate size. The effect of the finite bandwidth nature of transducers on the PA power spectra was also investigated. Experimental confirmation of theoretical results was conducted using porcine RBC samples exposed to 1064 nm optical wavelength using the Imagio Small Animal PA imaging system (Seno Medical Instruments, Inc., San Antonio, TX). Aggregation was induced with Dextran-70 (Sigma-Aldrich, St. Louis, MO) and the effect of hematocrit and aggregation level was investigated. The theoretical and experimental PA signal amplitude increased linearly with increasing hematocrit. The theoretical dominant frequency content of PA signals shifted towards lower frequencies (<30 MHz) and 9 dB enhancements in spectral power were observed as the size of aggregates increased compared to non-aggregating RBCs. Calibration of the PA spectra with the transducer response obtained from a 200 nm gold film was performed to remove system dependencies. Analysis of the spectral parameters from the calibrated spectra suggested that PA can assess the degree of aggregation at multiple hematocrit and aggregation levels.

Characterization of Red Blood Cell Aggregation with Photoacoustics: A Theoretical and Experimental Study

The aggregation of red blood cells (RBCs) is a phenomenon that is governed by plasma fibrinogen concentration and the shear forces of flow. We propose the use of photoacoustics (PA) for the detection and characterization of RBC aggregation. A 2D simulation study was performed to investigate the dependence of the PA signal on hematocrit for non-aggregated cells and on the aggregate size for aggregated samples. Experimental confirmation of theoretical results was conducted using human RBC samples and the Imagio PA imaging device (Seno Medical Instruments Inc., San Antonio, TX). Samples were exposed to 1064 nm laser irradiation. PA signals were collected from the non-aggregated samples at varying hematocrit levels. Aggregation was induced by suspending the RBCs in various concentrations of Dextran-70. To account for the response of the imaging system, the PA spectra were calibrated by dividing by the response of the transducer measured using a needle hydrophone. Theoretical and experimental results show a monotonic increase in the PA signal amplitude with increasing hematocrit. As the size of aggregates increases, simulations demonstrate a shift towards lower frequencies in the PA power spectrum as well as enhancements as large as 11 dB compared to the non-aggregated sample. Experimental results confirm the ability of PA to detect and quantify the aggregation of RBCs. PA radiofrequency spectral analysis seems provides a quantitative means for distinguishing between aggregated and non-aggregated samples.

A new design of low pass filter by Gaussian derivative family

Exploiting the models of human visual system based on Gaussian derivatives and their non-localization property in spectral domain, a new design of low pass filter is proposed in this work. The filter is designed by a weighted combination of the Gaussian derivative family which makes the passband of this filter almost equiripple. This design principle can be extended for the bandpass filter also because of the bandpass nature of the Gaussian derivatives in spectral domain. This work is applied in the design of adaptive digital filters for sigma-delta based ultrasound beamformer

A simulation study on ultrasound backscattering by cell aggregates with poly-disperse cells

A Monte Carlo simulation study on ultrasound backscattering by cell aggregates with poly-disperse cells is discussed. The nuclei in a cell aggregate were assumed as scatterers and the Anderson model was used to obtain backscattering amplitude for each nucleus. The resultant backscatter echo from many particles was determined by using linear superposition of backscatter signals emitted by the nuclei. The random sequential adsorption (RSA) method was employed to generate spatial organizations of nuclei. The frequency dependent backscattering coefficient (BSC) and signal envelope statistics were obtained from tissue samples with different size distributions. For each poly-disperse sample the nuclear populations followed a Gaussian distribution with the nuclear packing fraction fixed at 50.36%. It was found that integrated backscattering coefficient (IBSC) computed between 10–30 MHz increased about 7 dB for the highest poly-disperse sample considered compared to that of a mono-disperse sample. A Gaussian input pulse was employed to investigate signal envelope statistics. It was found that envelope histograms followed the Rayleigh distribution. The Rayleigh fit parameter (σ) increased as dispersity increased. For example, for the highest poly-disperse sample, σ increased about 105% and 157% compared to mono-disperse sample for input pulses with 5 and 25 MHz as the center frequencies and 80% bandwidths. The present work shows that poly-dispersity contributes to ultrasound backscatter but the shapes of histograms did not vary with the size distribution of scatterers.

Ultrasound backscattering by three-dimensional distributions of aggregated red blood cells: A Monte Carlo study

We present a Monte Carlo study on ultrasound backscattering by three-dimensional distributions of aggregated red blood cells (RBCs). In this study, particles were allowed to interact attractively with neighbors to form aggregates. The interaction potential energy profile between a pair of particles was mimicked by the Morse potential. The strength of pair interaction energy was controlled by fixing depth and width of the potential. Some three-dimensional configurations of cells representing realizations of the system were simulated by employing the Metropolis algorithm. The frequency dependent backscattering coefficient (BSC) was determined at three hematocrits, H = 20%, 30% and 40% and at different aggregation levels. It was noticed that BSC increased and spectral slope (SS) decreased as the aggregating potential was raised and the effect was more pronounced at higher hematocrits, particularly at 40%. In conclusion, this model can generate three-dimensional configurations of aggregated cells and can capture the role of RBC aggregation on BSC and SS.

Realistic photoacoustic image simulations of collections of solid spheres using linear array transducer

A methodology for simulating photoacoustic (PA) images of samples with solid spherical absorbers acquired using linear array transducer is described. Two types of numerical phantoms (i.e., polystyrene beads suspended in agar medium) of two different size regimes were imaged with a 40 MHz linear array transducer utilizing this approach. The frequency domain features and statistics of the simulated signals were quantified for tissue characterization. The midband fit at 40 MHz was found to be about 35 dB higher for the sample with larger beads (radius ~7.36 μm) than that of the sample with smaller particles (radius ~ 1.77 μm). The scale parameter of the generalized gamma distribution function was found to be nearly 51 times greater for the former sample compared to the latter sample. The method developed here shows potential to be used a s a fast simulation tool for the PA imaging of collection of absorbers mimicking biological tissue.

Probing confined and unconfined hemoglobin molecules with photoacoustics

Photoacoustic (PA) measurements on confined and unconfined hemoglobin molecules are presented. In vitro experiments were performed with porcine red blood cells (RBCs) at 532 and 1064 nm at various laser fluences. Fluence was gradually changed from 8 to 21 mJ/cm2/pulse for 532 nm and 353 to 643 mJ/cm2/pulse for 1064 nm. PA signals from suspended RBCs (SRBCs) and hemolyzed RBCs (HRBCs) were measured using a needle hydrophone at hematocrits ranging from 10 to 60%. PA amplitude was found to be varied linearly with the laser fluence for each type of samples at the above two optical radiations. At 532 nm, PA signals from SRBCs and HRBCs were measured to be nearly equal, whereas, at 1064 nm, signal amplitude for SRBCs was approximately 2 times higher than that of HRBCs. The results suggest that it may be feasible to detect hemolysis with PAs.

Photoacoustic assessment of oxygen saturation: effect of red blood cell aggregation

The simultaneous photoacoustic assessment of oxygen saturation and red blood cell aggregation is presented. Aggregation was induced on porcine red blood cells using Dextran-70 at multiple hematocrit levels. Samples were exposed to 750 nm and 1064 nm for each hematocrit and aggregate size in order to compute the oxygen saturation. As the size of the aggregate increased, the photoacoustic signal amplitude increased monotonically. The same trend was observed for increasing hematocrit at each aggregation level. The oxygen saturation of aggregated samples was 30% higher than non-aggregated samples at each hematocrit level. This suggests that the presence of red blood cell aggregates impairs the release of oxygen to the surrounding environment. Such a result has important implications for detecting red blood cell aggregation non-invasively and making clinical decisions based on the simulatenous assessment of oxygen saturation.

A Monte Carlo study on intraerythrocytic staging of malaria parasite using photoacoustics

Malaria parasite during its intraerythrocytic development digests considerable amount of hemoglobin (Hb) of the host red blood cell (RBC) and converts into a polymerized form of heme, referred to as hemozoin (Hz). The molar extinction coefficient of Hz greatly differs from that of Hb at certain optical wavelengths. Consequently, light absorption and hence, photoacoustic (PA) emission of an infected cell would vary from that of a normal cell. The PA signals were calculated from simulated blood samples with various stages of infection (normal, ring, trophozoite and schizont) employing a theoretical model. A Monte Carlo algorithm was used to simulate tissue samples in 3D. The simulation results showed that the PA amplitude monotonically decreased and increased with parasite growth, when probed with 411 and 700 nm laser beams, respectively. The PA amplitude for an ensemble of infected cells (schizont stage) was nearly 2 times lower at 411 and 79 times higher at 700 nm than that of healthy RBCs. The PA spectral power also demonstrated similar trends. For example, at 7.5 MHz spectral power nearly decreased by 6 dB and increased by 38 dB between the same stages at those optical wavelengths, respectively. It is conjectures that it may be feasible experimentally to differentiate intraerythrocytic stages of malaria parasite when illuminated by a suitable optical radiation.

On the potential of using photoacoustic spectroscopy for monitoring red blood cell aggregation

In this paper we examine the potential of using photoacoustic (PA) spectroscopy for the monitoring of red blood cell (RBC) aggregation phenomena. The process of RBC aggregation has been shown to occur during periods of increased plasma fibrinogen concentration and periods of decreased blood flow (leading to diminished shear forces on the aggregates). Current techniques used to monitor RBC aggregation are invasive and do not provide an accurate assessment of the aggregation process in-vivo. We present a theoretical model for investigating the potential of PA spectroscopy for detecting and characterizing the aggregation phenomenon. We show that the signal strength increases with RBC aggregation. Experimental confirmation of the theoretical predictions is provided. Our theoretical and experimental results suggest the PA spectroscopy is capable of monitoring RBC aggregation and providing important information about changes that occur during the aggregation process as it pertains to the dynamics of aggregate formation.