Correction to: Comparison of Flow Estimators for Rotary Blood Pumps: An In Vitro and In Vivo Study

Anastasios Petrou, Daniel Kuster, Jong‐Seok Lee, Mirko Meboldt, Marianne Schmid DanersAnnals 2019 · DOI · OpenAlex

Abstract

This article serves as a formal correction to the 2018 study "Comparison of Flow Estimators for Rotary Blood Pumps: An In Vitro and In Vivo Study," originally published in the Annals of Biomedical Engineering. The correction addresses a clerical error that resulted in the omission of a critical parameter, $h_{2h}$, from the Pump-Inlet Pressure-Based (PIPB) flow estimator model. By restoring this coefficient, the mathematical framework for predicting blood flow in rotary pumps is completed, ensuring the reproducibility of the study’s comparative findings.

Introduction

Rotary Blood Pumps (RBPs), particularly Left Ventricular Assist Devices (LVADs), have become standard therapy for patients with advanced heart failure. To ensure safe and effective operation, clinicians must have real-time data on the blood flow rate through the pump. While physical flow sensors can provide this data, they are often avoided in clinical practice because they increase the device's footprint, introduce additional points of mechanical failure, and are prone to measurement drift over time.

To overcome these limitations, researchers have developed "virtual" flow estimators. These are mathematical models that calculate flow rates using readily available data, such as pump rotational speed (RPM) and pressure differentials. The 2018 study by the authors provided a comprehensive evaluation of various flow estimators. However, the published coefficients for the PIPB estimator were incomplete. This correction restores the missing $h_{2h}$ parameter, which is essential for the model's accuracy.

Methodology and Model Structure

The PIPB flow estimator relies on a system of static and dynamic coefficients to characterize the relationship between pressure, pump speed, and flow. The model uses a polynomial structure where specific coefficients represent different physical influences within the pump.

Static Coefficients

The static coefficients ($h_1, h_{2h}, h_{3h}, h_{2l}, h_{3l}$) define the steady-state relationship between the pressure head and the flow rate at constant speeds. These parameters are derived from empirical data and are crucial for establishing the baseline performance of the estimator.

Dynamic Coefficients

The dynamic coefficient ($L$) accounts for the inertia of the blood as it moves through the pump. This is particularly important in the human circulatory system, where flow and pressure are highly pulsatile. The inclusion of $L$ allows the PIPB estimator to maintain accuracy during rapid changes in flow or pump speed, rather than being limited to steady-state conditions.

Data Sources

The numerical values for these coefficients were identified using "Dataset A" from the original 2018 study. This dataset was generated through a combination of in vitro benchtop mock loop experiments and in vivo animal models, providing a robust foundation for validating the flow estimators across a range of physiological conditions.

Corrected PIPB Estimator Parameters

The primary purpose of this correction is to provide the missing $h_{2h}$ value and the complete set of parameters for the PIPB estimator. The missing parameter, $h_{2h} = 0.703 \text{ mmHg/(L/min)}^2$, represents a quadratic relationship between flow and pressure. In fluid dynamics, pressure losses are often proportional to the square of the flow rate; therefore, omitting this term significantly reduces the model's precision, especially at higher flow rates.

Corrected static and dynamic coefficients for the PIPB flow estimator.

As shown in the table above, the model utilizes both "high" ($h_{2h}, h_{3h}$) and "low" ($h_{2l}, h_{3l}$) subscript parameters. This suggests a sophisticated model that may switch or blend coefficients depending on the operating regime (such as specific speed or flow ranges) to maintain high precision across the pump's entire functional envelope.

Discussion and Conclusion

The restoration of the $h_{2h}$ coefficient is vital for the accurate implementation of the PIPB flow estimator. In the field of biomedical engineering, where these models may be integrated into the control loops of life-saving medical devices, precision is non-negotiable. An incomplete model could lead to significant flow estimation errors, potentially causing a pump to operate sub-optimally or in a manner that is dangerous to the patient.

While the specific numerical values provided here are unique to the pump hardware used in the original study, they provide a critical reference for researchers and engineers working on similar systems. This correction underscores the importance of rigorous data reporting and the necessity of maintaining accurate mathematical models in MedTech literature. Future work in this area continues to move toward self-calibrating estimators that can adapt to individual patient anatomy and changes in blood viscosity without requiring manual coefficient identification.

Reference: Correction to: Comparison of Flow Estimators for Rotary Blood Pumps: An In Vitro and In Vivo Study. Originally published in Ann Biomed Eng (2018) 46:1003–1016.

Summary updated

BibTeX
@article{petrou2019correction,
  author = {Anastasios Petrou and Daniel Kuster and Jong‐Seok Lee and Mirko Meboldt and Marianne Schmid Daners},
  title = {{Correction to: Comparison of Flow Estimators for Rotary Blood Pumps: An In Vitro and In Vivo Study}},
  journal = {Annals},
  year = {2019},
  doi = {10.1007/s10439-019-02295-0},
}
Figure