In Focus
Towards better analytical tools for blood quality assessment
Emerging technologies to the rescue
Current approaches to red blood cell (RBC) quality assessment are largely inferential with respect to their relevance to cell function.
At the level of the blood component, measurements such as hemolysis, hematocrit, and hemoglobin content per unit provide indirect information on product integrity and RBC dose. At the level of the cell population, routine hematology indices similarly provide surrogate measurements of function. For example, hemoglobin content offers an estimate of oxygen-carrying potential whereas cell size indices provide only limited insight into morphology. Correlations between important metabolites indicative of energetic homeostasis, such as ATP and 2,3-DPG, or redox homeostasis, such as NAD(P)+/NAD(P)H, and the functional parameters they influence—including morphology, deformability, and oxygen offloading capacity—are often weak.
The efficacy of a transfusion, as determined by the unit properties, can be considered along two distinct but interrelated functional axes: the capacity of RBCs to load and unload oxygen, and their ability to persist in circulation before splenic sequestration and clearance. Although these properties are often implicitly treated as concordant, their relationship is not inherent; an RBC may retain adequate oxygen-exchange capacity yet be rapidly cleared from circulation, or conversely persist despite impaired oxygen delivery capacity. Compounding this is the marked heterogeneity of stored RBC populations, wherein deterioration occurs along a continuum and a relatively small subpopulation of severely compromised cells may be obscured by measurements that report only population-wide averages.1 A comprehensive assessment of RBC quality therefore requires not only direct evaluation of both functional dimensions, but also sufficient analytical resolution to identify clinically meaningful subpopulations that would otherwise be masked within the bulk population. We now have strategies capable of directly interrogating RBC morphology, mechanics, and oxygen-exchange function at increasingly high throughput and, in some cases, at single-cell resolution, creating an opportunity to move beyond bulk surrogates toward more functionally informative measures of blood quality.
Among the emerging strategies that seek to operationalize this shift toward more direct and scalable functional assessment, perhaps the simplest to implement is the Sysmex® FlowScore, because it derives functionally relevant information from light-scatter data already obtainable using conventional hematology instrumentation.2 The method applies a nonlinear regression model to map forward- and side-scatter parameters onto the time constant of oxygen unloading. As RBCs undergo metabolic deterioration during storage, loss of biconcavity and progressive spherical remodeling alter these scatter properties. FlowScore captures this transition through changes in their relative relationship. Its principal appeal is therefore pragmatic: it extracts a functionally relevant estimate from measurements that are already routine, offering a potentially straightforward bridge between conventional hematology data and oxygen-delivery function. The limitation, however, is that such a bulk scatter-derived metric remains fundamentally compressive. FSC and SSC are composite signals influenced by cell size, shape, orientation, and optical properties, and morphologically distinct RBC populations can therefore overlap substantially in scatter space. FlowScore may consequently capture broad remodeling of the erythrocyte population while failing to resolve the heterogeneity that becomes most important when only a minority of cells are severely compromised. This is particularly relevant to the second functional axis of transfusion efficacy - post-transfusion persistence - because storage-induced microerythrocytes, including type III echinocytes, spheroechinocytes, and spherocytes, are preferentially and rapidly cleared after transfusion, and their abundance within a unit is associated with reduced post-transfusion recovery.3 A unit-level scatter metric may therefore obscure the very subpopulation most likely to undergo splenic sequestration and early clearance. Furthermore, FlowScore does not directly measure RBC function; instead, it estimates oxygen onloading kinetics using flow cytometry light-scatter parameters (cell geometry).
An alternative strategy is to assess morphology directly at the single-cell level by pairing label-free imaging flow cytometry with deep learning.4 Such approaches address many of the limitations of historical morphology assessment, which relied on labor-intensive and subjective inspection of blood films and was vulnerable to both observer variability and preparation artifacts. Automated analysis of single-cell brightfield images can instead classify RBC morphology at near-expert performance while providing a more objective and scalable view of morphological deterioration.4 More importantly, it shifts quality assessment from a single population-average descriptor toward the distribution of phenotypes present within a unit, allowing rare or disproportionately damaged subpopulations to be identified rather than averaged away.
Figure 1: The evolution of red blood cell quality assessment
While geometry and morphology are associated with oxygen-carrying capacity and deformability, direct assessment of these properties avoids the potentially pernicious consequences of relying on correlative inference alone. The LORRCA® (RR Mechatronics) remains the established reference platform for RBC deformability analysis, generating elongation index–shear stress curves under iso- and anisotonic conditions. However, it is low throughput and limited to population-level averages. Newer platforms offer either greater practicality or greater phenotypic resolution. RheoSCAN® (RheoMeditech Inc) provides rapid, low-volume deformability measurements suitable for routine screening, butlike LORRCA,does not resolve mechanical heterogeneity within a unit.5 In contrast, Rivercyte’s Naiad® and Zellmechanik’s AcCellerator® apply high-throughput deformability cytometry to characterize the mechanical and morphological properties of individual cells.6,7 These approaches therefore extend deformability assessment to permit mechanical assessment of the compromised subpopulations that exist within an otherwise apparently acceptable unit.
A complementary set of technologies addresses the other major functional axis: oxygen exchange. HEMOX® (TCS Scientific) is the established reference platform for generating oxygen dissociation curves (ODC) and determining p50, but it shares many of the practical limitations of LORRCA, namely relatively low throughput and population-level readouts OxyDial has launched their Oxygen-Hemoglobin Equilibrium device[DM1.1] that directly measures the ODC by recording oxygen partial pressure (pO2) and hemoglobin saturation at many points during controlled deoxygenation.8 This provides direct p50 measurement through a faster and more automated platform, making it more suited for routine RBC quality assessment than the HEMOX analyzer. Newer approaches offer either greater throughput or greater physiological resolution. BMG Labtech’s oxygen dissociation assay [DM2.1] on their SPECTROstar Nano spectrophotometer enables high-throughput, 96-well screening of haemoglobin deoxygenation kinetics, although the published workflow uses purified haemoglobin and therefore excludes the effects of RBC geometry, diffusion distance, membrane properties, and cellular heterogeneity.9 In contrast, BloodSpeed® [DM3.1] utilizes a microfludic, microscopic imaging technique to measure oxygen loading and unloading kinetics directly in intact RBCs at single-cell resolution, allowing heterogeneity in oxygen-exchange function to be resolved within the unit itself.10
Routine adoption of these more comprehensive evaluation techniques has remained limited, largely because these approaches lack the efficiency required for routine implementation. Historically, limited recognition of the disconnect between conventional quality metrics and RBC function reduced the urgency to invest in such direct analytical approaches. This is now changing, as emerging technologies increasingly aim to make comprehensive and functionally relevant assessment of RBC quality faster, more scalable, and more compatible with routine blood-bank workflows.
References
- Mykhailova O, Olafson C, Turner TR, D’Alessandro A, Acker JP. Donor-dependent aging of young and old red blood cell subpopulations: Metabolic and functional heterogeneity. Transfusion. 2020.
- Rabcuka J, Smethurst PA, Dammert K, Saker J, Aran G, Walsh GM, et al. Assessing the kinetics of oxygen-unloading from red cells using FlowScore, a flow-cytometric proxy of the functional quality of blood. eBioMedicine. 2025.
- Roussel C, Morel A, Dussiot M, Marin M, Colard M, Fricot-Monsinjon A, et al. Rapid clearance of storage-induced microerythrocytes alters transfusion recovery. Blood. 2021..
- Doan M, Sebastian JA, Caicedo JC, Siegert S, Roch A, Turner TR. et al. Objective assessment of stored blood quality by deep learning. Proc Natl Acad Sci U.S.A.. 2020.
- RheoMeditech Inc. RheoSCAN. 2026.
- Rivercyte. Naiad 1.0. 2026.
- Zellmechanik Dresden. Real-time deformability cytometry. Available from: https://www.zellmechanik.com/technology.html. 2026.
- Oxydial. Available from: https://oxydial.com/. 2026.
- Patel MP, Siu V, Silva-Garcia A, Xu Q, Li Z, Oksenberg D. Development and validation of an oxygen dissociation assay, a screening platform for discovering, and characterizing hemoglobin-oxygen affinity modifiers. Drug Des Devel Ther. 2018.
- BloodSpeed. Next-generation haematology analysis. Available from: https://bloodspeed.org/. Accessed July 6, 2026.
