Intrinsic biophysical cell properties hold an enormous potential for cell class and state classification in microfluidics, allowing to avoid the need of cost intensive fluorescence labelling. Several methods can accomplish cell identification, while convolutional neural networks show an outstanding performance compared to other state-of-the-art classification methods, regarding accuracy and speed. In fact, neural networks show high performance for known image class prediction but struggles when unknown (out of distribution) image classes need to be identified. In such a scenario no prior knowledge of the unknown cell class can be used for the model training, which inevitably results in image misclassification. In fact, to distinguish unknown cell classes, a neural network must first construct an in-distribution of known images to afterwards detect out of distribution as unknowns, which is also called open-set classification assumption. Ones, a new cell class is identified, the neural network can be retrained with the obtained knowledge to dynamically update its cell class database. This process can be simply repeated for each new detected cell class. We applied this open-set idea to scattering pattern snapshots of different classes of living cells obtained in microfluidics. Our outcome shows a proof-of-concept for open-set based convolutional neural network for cell image classification, which can be applied to a wide range of single cell classification approaches to reduce uncertainty in machine learning based technologies.
We investigated subclasses of living peripheral blood cells in a microfluidic-based system, with the aim to characterize their morphometric and optical properties, and to track their position in flow in a label-free modality. We employed two coherent imaging techniques: a scattering approach of precisely aligned single cells, and a digital holography approach to achieve optical cell reconstructions in flow. Cells were first 3D-aligned in round shaped capillary and subsequently measured in a following square shaped channel. Results were obtained at two fixed measurement positions, the first one was chosen close to the entrance of the measurement channel to ensure 3D cell alignment for scattering investigations; the second was placed 15 mm after to study additional cell properties by digital holography and to investigate possible variations of axial cell positions. First, the refractive index, ratio of the nucleus over cytoplasm, and cell dimension were investigated from scattering investigations. Further quantitative phase-contrast reconstructions by digital holography were employed to calculate surface area, dry mass, biovolume and positions of cells using the scattering outcomes as input parameters. The precise cell alignment at the first measurement position could be confirmed. At the second measurement position a full label-free characterization of cell classes in distinct vertical positions was realized and supported by applied microfluidic force calculations, which can be used to align, deform and/or separate cells. Our results confirm the possibility to differentiate cell classes in flow, thus avoiding chemical cell staining or labeling, which are nowadays used.
Cell of human blood stream are divided into two groups: Red Blood Cells (RBC) and White Blood Cells (WBC). RBC have a peculiar biconcave disk shape and they are responsible for the delivering of O2 and CO2 through the body. WBC are a more widespread class of cell ensuring immunity against pathogens. They can be divided in two main classes: granulocyte cells and A-granulocyte cells. Neutrophils, basophils and eosinophils belong to the granulocyte cell class, while lymphocytes and monocytes belong to A-granulocyte. Both in RBC and WBC, the intrinsic physical properties of a cell are indicators of cell condition and, furthermore, of the overall human body state. Thus, the accurate comprehension of the physiological structure of WBCs is fundamental to recognize diseases. Here we show the possibility to simple and straightforwardly characterize the physical properties of individual RBC and mononuclear WBC in a microfluidic context, using a wide angle light scattering apparatus and a corresponding theoretical simulation of Optical Signature (OS). A non-Newtonian polymer alignment solution for cell is used to ensure an individual cell alignment in the microfluidic flow, thus permitting a precise investigation. Additionally, Quantitative Phase Imaging (QPI) holographic measurements are performed to estimate cell morphometric features, such as their refractive index. We analyzed more than 200 WBCs and 100 RBCs of three different probands. Results showed distinct cell populations according to their measured dimensions and shape, which can be associated to the presence of RBC, lymphocytes and monocytes.
The Light Scattering Profile (LSP) of an individual cell provides a fast and accurate characterization of its morphological properties. By combining a camera-based small angle light scattering apparatus with a microfluidic-induced particle migration technique, it is possible to characterize cells in microfluidic flows. The scattering profile of an individual cell can be fully characterized by our optimized optical light collection system. Viscoelastic-induced particle migration by polyethylene oxide implemented in a low-cost microfluidic device composed of an alignment section and a measuring section opens the possibility of precise, label-free, individual cell analysis. We have studied living cells in microfluidic flows by our light scattering apparatus and by a Digital Holographic Microscope (DHM) system. Our DHM measurements provided an accurate 3D position tracking even in multiple cell conditions.
A CCD-camera based small angle light scattering (SALS) apparatus has been used to characterize single micrometric particles flowing in a micro-channel. The measured scattering vector spans the range 2x10-2 - 6:8x101μm-1. The incident laser light is collimated to a spot of about 50 μm in diameter at the sample position with a divergence lower than 0.045 rad. Such small collimated laser beam opens the possibility to perform on-line SALS of micron-sized particles flowing in micro-channels. By properly designing the micro-channel and using a viscoelastic liquid as suspending medium we are able to realize a precise 3D focusing of the target particles. The forward scattering emitted from the particle is collected by a lens with high numerical aperture. At the focal point of that lens a homemade beam stop is blocking the incident light. Finally, a second lens maps the scattered light on the CCD sensor, allowing to obtain far field images on short distances. Measurements with mono-disperse polystyrene particles, both in quiescent and in-flow conditions have been realized. Experiments in-flow allow to measure the single particle scattering. Results are validated by comparison with calculations based on the Lorenz-Mie theory. The quality of the measured intensity profiles confirms the possibility to use our apparatus in real multiplex applications, with particles down to 1 μm in radius.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.