Classification of Red Blood Cell Rigidity from Sequence Data of Blood Flow Simulations Using Neural Networks

Abstract

Numerical models for the flow of blood and other fluids can be used to design and optimize microfluidic devices computationally and thus to save time and resources needed for production, testing, and redesigning of the physical microfluidic devices. Like biological experiments, computer simulations have their limitations. Data from both the biological and the computational experiments can be processed by machine learning methods to obtain new insights which then can be used for the optimization of the microfluidic devices and also for diagnostic purposes. In this work, we propose a method for identifying red blood cells in flow by their stiffness based on their movement data processed by neural networks. We describe the performed classification experiments and evaluate their accuracy in various modifications of the neural network model. We outline other uses of the model for processing data from video recordings of blood flow. The proposed model and neural network methodology classify healthy and more rigid (diseased) red blood cells with the accuracy of about 99.5% depending on the selected dataset that represents the flow of a suspension of blood cells of various levels of stiffness.

Computational Study of Methods for Determining the Elasticity of Red Blood Cells Using Machine Learning

“…Our recent research goal is therefore not to compare the consistency of real and simulation experiments, but to focus on the differences in the capabilities of 3D models and their, video-recording-based, 2D descriptions. In [17], we showed that different results can be yielded by whether the video footage was acquired at the top or at the side of a channel with a flowing fluid.…”

“…Several factors influence the migration of particles across streams in confined-space flow suspensions, with reduced elasticity of a certain fraction of cells generally considered to be a key factor. In a previous study by [17], we worked with only two levels of elasticity of RBCs, healthy and damaged (stiffer). In the present work, RBCs with three and nine elasticity levels, respectively, are included in the individual simulations, which both represents a shift towards a more realistic model of blood flow and creates the need to address the issue from a more holistic perspective.…”

“…In total, we simulated a total of 54 red blood cells in both cases, Sim3 and Sim9. For example, we used a similar procedure in the study [17]. For each internal simulation step of the ESPreSso module, the current position of each non-stationary point in the blood flow, and hence all red blood cell triangulation points, is calculated.…”

“…The problem of determining a RBC's elasticity using simulation experiments was investigated in [17], where the main goal was to classify a RBC's elasticity into predefined categories. In this paper we extended the solution to the problem of categorisation of values of elastic parameters to a regression task, where we determine for each cell the value of the regression function determining the elasticity of the given cell.…”

Classification of Red Blood Cells Using Time-Distributed Convolutional Neural Networks from Simulated Videos

“…The determination of RBC elasticity through simulation experiments has been investigated in prior works [16,17]. However, this paper proposes a novel approach capable of identifying various levels of RBC elasticity associated with different diseases from video recordings.…”