Potential of cell tracking velocimetry as an economical and portable hematology analyzer

Before analyzing abnormal sickle RBC samples, a total of 31 healthy RBC samples were collected and the following parameters; Hb concentration, Hct, RBC count, MCV, RDW, MCH, MCHC and Hypo, were measured by using a combination of different methods explained above. The results of the measurements for these healthy RBC samples are presented in Table 1. As expected, most of the test results show a healthy blood status of the donors with their respective parameters within the normal ranges²⁶. However, some parameters related to the RBC volume (MCV, MCHC and RDW) and measured by CC or CTV are outside the normal range. These discrepancies are thoroughly analyzed and discussed in the following.

Mean corpuscular volume

The estimation of MCV (Eq. 8) calculated from the Hct obtained by the PCV method and the RBC count obtained from CC, gives an MCV range of 70–110 fL, which matches well with the literature values reported for healthy individuals²⁶. However, measurements from the Coulter Counter and CTV, using Eq. (5), give us values around 50 fL. The close overlap of size distribution data from the CC and the CTV is shown in Fig. 2. It can be seen from the bar graph that MCV_ave is greater than MCV_CC and MCV_CTV for all 31 healthy donor samples.

Comparison between the MCV_ave (calculated from Hct_PCV and RBC count from Coulter Counter), MCV_CC (direct measurment from Coulter Counter), and the MCV_CTV (calculated from sedimentation rate data) values for all 31 healthy blood samples.

As mentioned, for the determination of MCV_ave (Eq. 8), data obtained from PCV and CC are employed. Values from CC related to RBC count are within the normal range for healthy adults, as seen in Table 1²⁶. Values of Hct obtained from PCV method are also within the normal range (around 45%). The disagreement between the MCV obtained by Eq. (8) and the value obtained from CC and CTV may be due, in part, to the fact that the PCV method uses centrifugal force to pack the RBCs in a hematocrit capillary tube, which inevitably suffers from plasma entrapment, illustrated in Fig. 3. Thus, the discrepancy in MCV between the different methods is partly attributed to the plasma entrapment when measuring Hct in this manner. Nevertheless, it has been reported that less than 5% of plasma is trapped in the packed RBC layer during centrifugation²⁶. For example, Paterakis et al.²⁷ have suggested that the trapped plasma constitutes less than 3% for any normal or abnormal samples, including oxygenated sickle cells.

Schematic representation of plasma entrapment during Hct measurements by the PCV method.

Additionally, hematology analyzers routinely used to measure MCV such as the ADVIA treat RBCs with sodium dodecyl sulfate to make the cells spherical and glutaraldehyde to fix the membrane. Cell sizes are then measured using low angle and high angle light scattering while flowing in a single file order²⁸. Although the Mie scattering theory requires objects to be spherical and altering the morphology from a disc to sphere does not necessarily alter the MCV, osmotic pressure changes in the carrier fluid of sodium dodecyl sulfate would alter the cell volume considerably and may be a source of MCV overestimation^29,30.

Another plausible explanation for this disagreement is the fact that RBCs are deformable and some analyzers do not take into account the erythrocyte deformability. This could lead to an underestimation of the MCV measured on CC or CTV, as has been previously reported by others. For example, d’Onofrio et al.³¹ stated that electronic measures of MCV could be underestimated. In fact, these authors claim that electronic Hct_CC (measured on CC by using RBC count_CC and MCV_CC) is a redundant parameter that could be abandoned, and that there are significant discrepancies between the manual Hct (obtained by PCV method), with the artifacts of plasma trapping and cell shrinkage, and the automated measurements. Thus, they suggest that calculation of Hct from the mean or the sum of pulse sizes and RBC count leads to systematic underestimation of automated Hct (and MCV) compared to the manual PCV measurement. Finally, Brugnara et al.³² reported that individual MCV measurements on different hematology analyzers are dependent on the technology used and that impedance-based instruments (such as CC) might underestimate MCV in hypochromic RBCs. The differences in the technologies employed along with the fact that such impedance-based instruments might not account for the deformability of RBCs³³ could contribute to the disagreement on the MCV obtained with the different techniques. So far, no internationally accepted reference method has been published for measuring MCV³⁴.

Due to the previous explanations, and most importantly, because there is no reference method to measure MCV, we conclude that the CTV could potentially be employed for measuring MCV along with other hematological parameters as long as reference values or correction factors are established. Even though RBC deformability might not be considered on CTV, the values obtained from both CTV and Coulter Counter are very similar, and Coulter Counter is already considered as an accurate instrument for measuring cell and particle volume distributions.

A Coulter Counter calculates the volume of a cell by measuring the voltage difference between the inside and outside of an aperture tube (that is both filled with and submerged in an electrolyte solution) every time an object passes through the orifice of the tube and creates resistance in the circuit. The red cell volume distribution reported by the Coulter Counter, using electrical impedance, is in close agreement to the values reported by the CTV, which measures cell sedimentation to calculate volume. Figure 4a,b report the aggregate volume distributions of healthy and sickle RBC samples measured with both CC and CTV. These two methods show close agreement and high overlap between the histograms for both sets of subjects, yet CTV slightly underestimates MCV for healthy donors and overestimates for SCD subjects. The discrepancy in SCD data may be due to a number of factors including; increased density of sickle RBCs (due to dehydration), a lower sedimentation coefficient (due to sickling shape change), or an increased presence of large reticulocytes^35,36.

Volume histogram distribution (left y-axis) and cumulative curve (right y-axis) of RBCs measured on Coulter Counter (black) and CTV (red) for (a) all 31 healthy RBC samples and (b) all 15 sickle RBC samples.

Moreover, because of the disagreement in average MCV values between CC and CTV (51.8 and 49.6 fL) and the normal range (within 80–100 fL), and since RDW is calculated from the MCV, the values of RDW_CC and RDW_CTV are slightly higher than the normal value (generally below 15%). Specifically, RDW is obtained based on both the width of the cell volume distribution and the average cell size (i.e. it is calculated by dividing the standard deviation by the mean MCV). As can be seen in Table 1, the RDW values reported by CC and CTV are around 25%.

Hemoglobin mass and concentration in individual RBCs

The calculated MCHC_CTV scatterplots for all healthy donors and SCD subjects are plotted as a function of MCV_CTV in Fig. 5a,b. Black dots correspond to the RBCs above the hypochromic cutoff value of 28 g/dL and the red dots represent the subfraction of hypochromic RBCs. A high-volume hematology analyzer such as ADVIA uses light scattering and spectral absorption to determine MCHC while CTV measures induced velocity of RBCs that contain ferric Hb under a strong and constant magnetic energy gradient. Similarly, Fig. 6a,b plot MCH_CTV as a function of MCV_CTV for the same populations (healthy and sickle RBC samples, respectively) showing the MCHC hypochromic cutoff. Figures 5 and 6 demonstrate that our CTV system can measure the same parameters with detailed understanding of the population, as thousands of RBCs can be tracked per sample.

MCHC_CTV and MCV_CTV calculations for (a) all healthy RBCs and (b) sickle RBCs. Normal (black) cells and hypochromic (red) cells are shown with the cutoff value of 28 g/dL.

MCH_CTV and MCV_CTV calculations for (a) all healthy RBCs and (b) sickle RBCs. Normal (black) cells and hypochromic (red) cells are shown with the cutoff value of 28 g/dL.

By observing both Fig. 6 and Table 1, it can be concluded that the average MCH (expressed in pg of Hb per RBC) reported by CTV is around the normal range reported for healthy samples (27–33 pg) and the global average for healthy RBC donors is 29.62 pg. Moreover, our previous study assesed the performance of CTV on measuring mass of Hb on individual RBCs by comparing CTV results to the standard method based on spectrophotometry, and our instrument reported accurate measurements and analyses²⁴. Nevertheless, when comparing the MCHC measured by CTV to the reference range of 28–41 g/dL²⁶, CTV measured higher MCHC because MCV_CTV is smaller than the reference range (see Table 1).

Pearson’s correlation coefficients (r) between CTV-measured MCHC/MCV and MCH/MCV for healthy and SCD samples were calculated for a total of 17,222 healthy and 7655 sickle RBC tracks. An r value of + 1.0 indicates a perfect positive correlation between two variables while an r value of − 1.0 indicates perfect negative correlation. Thus, r² can be used to describe the correlation in either case. For the healthy RBC population, r_MCHC-MCV = − 0.249 and r_MCH-MCV = 0.442. For SCD samples, r_MCHC-MCV =  0.298 and r_MCH-MCV = 0.604. This represents the strength of fit between the intracellular iron content and its dependence on cell volume. The results suggest that larger cells have lower MCHC and higher MCH than smaller cells, and that the MCH dependence on volume is stronger than the trend of MCHC. Additionally, iron status in SCD subjects has a stronger dependence on volume than healthy ones; meaning, healthy individuals’ cells iron content is more independent of size than for those with SCD. The average values, standard deviation and coefficient of variation (CV) for MCV, MCHC and MCH in healthy and SCD populations are displayed in Table 2. Most notable from studying the variance is the wider distribution in MCV for SCD samples. CTV measured slightly higher iron content in healthy donors with similar variance between healthy and SCD samples. The small discrepancy between values reported in Table 1 may be due to the different number of cells measured between each donor. Analysis of variance and correlation between these parameters might be used to diagnose abnormal iron status in subjects. Provided a sufficiently large dataset for the subject, and baseline measurments of their gender, age, etc., CTV may be used to assess the snapshot of a subject’s iron status over the past 120 days (the approximate life of an RBC in vivo)³⁷.

Additionally, observing Table 1 also shows that besides MCH_CTV, the overall Hb_spec of healthy donor 6 is much lower compared to other healthy donors, and the hypochromic fraction is much greater compared to the samples obtained from healthy donors 1–5. We have represented the temporal evolution of Hypo (%) for all the healthy donors (blood was drawn from each healthy donor on a weekly basis) in Fig. 7. It is known that during recombinant human erythropoietin (rHuEPO) therapy for anemia, Hypo > 10% is an indication for iron supplement requirement¹⁶. The fraction of hypochromic RBCs for healthy donor 6 is as high as 55%, and as low as 3%, over the 4-week testing period while other healthy donors’ values stay around the cutoff value. Hypo has been considered as a very sensitive marker because small changes in the number of RBCs with inadequate hemoglobin can be measured before there is any change in the MCHC¹⁵. Thus, quantification of hypochromic and/or hyperchromic red cells is helpful in the diagnosis of anemia. In fact, in a population of young anemic females, the percentage of hypochromic RBCs had the highest accuracy in distinguishing IDA from other anemias with normal iron stores^29,32. Based on this, we concluded that our CTV can be very useful not only in measuring the mass of Hb in individual RBCs, but also on the diagnosis of the IDA.

Temporal evolution of the percentage of hypochromic RBCs measured on CTV for all healthy donors.

Finally, and for further analysis on this healthy donor samples (donor 6), the histogram of the weekly measured MCHC and MCH values has been presented in Fig. 8. On the left side of Fig. 8, the red line corresponds to the hypochromic RBCs and the black line presents normal RBCs in terms of Hb concentration. As expected, the normal RBC population decreases as the Hypo percentage increases and vice versa. One interesting fact is that while the overall Hb concentration stays around 12.7–12.9 g/dL, the ratio of the hypochromic RBCs and normal RBCs seems to have a dynamic fluctuation. For comparison, the histogram of Hb mass is presented on the right side of Fig. 8, showing a similar fluctuation to the MCHC presented on the left side of Fig. 8.

Weekly measurements of MCHC and MCH on CTV for healthy donor 6. Left panel presents MCHC_CTV (red line shows the hypochromic RBCs and the numbers inside the graph report Hb_spec) and right panel shows MCH_CTV (numbers inside the graphs report the average MCH_CTV).