Image sensing with multilayer nonlinear optical neural networks

Nature Photonics, Journal Year: 2023, Volume and Issue: 17(5), P. 408 - 415

Published: March 23, 2023

Language: Английский

---

Label-free microfluidic cell sorting and detection for rapid blood analysis

Lab on a Chip, Journal Year: 2023, Volume and Issue: 23(5), P. 1226 - 1257

Published: Jan. 1, 2023

Blood tests are considered as standard clinical procedures to screen for markers of diseases and health conditions. However, the complex cellular background (>99.9% RBCs) biomolecular composition often pose significant technical challenges accurate blood analysis. An emerging approach point-of-care diagnostics is utilizing "label-free" microfluidic technologies that rely on intrinsic cell properties fractionation disease detection without any antibody binding. A growing body evidence has also reported dysfunction their biophysical phenotypes complementary hematoanalyzer analysis (complete count) can provide a more comprehensive profiling. In this review, we will summarize recent advances in label-free separation different components including circulating tumor cells, leukocytes, platelets nanoscale extracellular vesicles. Label-free single morphology, spectrochemical properties, dielectric parameters characteristics novel blood-based biomarkers be presented. Next, highlight research efforts combine microfluidics with machine learning approaches enhance sensitivity specificity studies, well innovative solutions which capable fully integrated sorting Lastly, envisage current future outlook platforms high throughput multi-dimensional identify non-traditional diagnostics.

Language: Английский

---

Neural network-enhanced real-time impedance flow cytometry for single-cell intrinsic characterization

Lab on a Chip, Journal Year: 2021, Volume and Issue: 22(2), P. 240 - 249

Published: Nov. 22, 2021

Single-cell impedance flow cytometry (IFC) is emerging as a label-free and non-invasive method for characterizing the electrical properties revealing sample heterogeneity. At present, most IFC studies utilize phenomenological parameters (e.g., amplitude, phase opacity) to characterize single cells instead of intrinsic biophysical metrics radius r, cytoplasm conductivity σi specific membrane capacitance Csm). Intrinsic are normally calculated off-line by time-consuming model-fitting methods. Here, we propose employ neural network (NN)-enhanced achieve both real-time single-cell characterization parameter-based cell classification at high throughput. Three (r, Csm) can be obtained online in via trained NN 0.3 ms per event, achieving significant improvement calculation speed. Experiments involving four cancer one lymphocyte demonstrated 91.5% accuracy type test group 9751 samples. By performing viability assay, provide evidence that se would not substantially affect property. We envision NN-enhanced will new platform high-throughput, characterization.

Language: Английский

---

Automated biophysical classification of apoptotic pancreatic cancer cell subpopulations by using machine learning approaches with impedance cytometry

Lab on a Chip, Journal Year: 2022, Volume and Issue: 22(19), P. 3708 - 3720

Published: Jan. 1, 2022

Unrestricted cell death can lead to an immunosuppressive tumor microenvironment, with dysregulated apoptotic signaling that causes resistance of pancreatic cancer cells cytotoxic therapies. Hence, modulating by distinguishing the progression subpopulations under drug treatment from viable towards early apoptotic, late and necrotic states is interest. While flow cytometry after fluorescent staining monitor apoptosis single-cell sensitivity, background non-viable within non-immortalized tumors xenografts confound distinction intensity each state. Based on impedance drug-treated are obtained differing levels gemcitabine we identify biophysical metrics distinguish quantify cellular at versus states, using machine learning methods train for recognition phenotype. supervised has previously been used classification datasets known classes, our advancement utilization optimal positive controls class, so clustering unsupervised occur unknown datasets, without human interference or manual gating. In this manner, automated be follow in heterogeneous sample, developing treatments modulate advance longitudinal analysis discern emergence resistant phenotypes.

Language: Английский

---

Computer vision meets microfluidics: a label-free method for high-throughput cell analysis

Microsystems & Nanoengineering, Journal Year: 2023, Volume and Issue: 9(1)

Published: Sept. 21, 2023

In this paper, we review the integration of microfluidic chips and computer vision, which has great potential to advance research in life sciences biology, particularly analysis cell imaging data. Microfluidic enable generation large amounts visual data at single-cell level, while vision techniques can rapidly process analyze these extract valuable information about cellular health function. One key advantages integrative approach is that it allows for noninvasive low-damage characterization, important studying delicate or fragile microbial cells. The use provides a highly controlled environment growth manipulation, minimizes experimental variability improves accuracy analysis. Computer be used recognize target species within heterogeneous populations, understanding physiological status cells complex biological systems. As hardware artificial intelligence algorithms continue improve, expected become an increasingly powerful tool situ microelectromechanical devices combination with could development label-free, automatic, low-cost, fast recognition high-throughput responses different compounds, broad applications fields such as drug discovery, diagnostics, personalized medicine.

Language: Английский

---

Impedance‐Based Multimodal Electrical‐Mechanical Intrinsic Flow Cytometry

Small, Journal Year: 2023, Volume and Issue: 19(45)

Published: July 12, 2023

Abstract Reflecting various physiological states and phenotypes of single cells, intrinsic biophysical characteristics (e.g., mechanical electrical properties) are reliable important, label‐free biomarkers for characterizing cells. However, single‐modal or properties alone not specific enough to characterize cells accurately, it has been long challenging couple the conventionally image‐based characterization impedance‐based characterization. In this work, spatial‐temporal impedance sensing signal leveraged, an multimodal electrical‐mechanical flow cytometry framework on‐the‐fly high‐dimensional measurement is proposed, that is, Young's modulus E , fluidity β radius r cytoplasm conductivity σ i membrane capacitance C sm With characterization, can better reveal difference in cell types, demonstrated by experimental results with three types cancer (HepG2, MCF‐7, MDA‐MB‐468) 93.4% classification accuracy pharmacological perturbations cytoskeleton (fixed Cytochalasin B treated cells) 95.1% accuracy. It envisioned provides a new perspective accurate single‐cell

Language: Английский

---

Modified Red Blood Cells as Multimodal Standards for Benchmarking Single-Cell Cytometry and Separation Based on Electrical Physiology

Analytical Chemistry, Journal Year: 2022, Volume and Issue: 94(6), P. 2865 - 2872

Published: Feb. 2, 2022

Biophysical cellular information at single-cell sensitivity is becoming increasingly important within analytical and separation platforms that associate the cell phenotype with markers of disease, infection, immunity. Frequency-modulated electrically driven microfluidic measurement systems offer ability to sensitively identify single cells based on biophysical information, such as their size shape, well subcellular membrane morphology cytoplasmic organization. However, there a lack reliable reproducible model particles well-tuned electrical phenotypes can be used standards benchmark physiology unknown types or dielectrophoretic metrics novel device strategies. Herein, application red blood (RBCs) multimodal standard systematically modulated electrophysiology associated fluorescence level presented. Using glutaraldehyde fixation vary capacitance by resealing after electrolyte penetration interior conductivity in correlated manner, each modified RBC type identified phenomenological impedance fitted dielectric models compute information. In this data from mapped versus these for facile determination conditions, without need time-consuming algorithms often require fitting parameters. Such internal cytometry advance in-line phenotypic recognition

Language: Английский

---

Morphological profiling by high-throughput single-cell biophysical fractometry

Communications Biology, Journal Year: 2023, Volume and Issue: 6(1)

Published: April 24, 2023

Abstract Complex and irregular cell architecture is known to statistically exhibit fractal geometry, i.e., a pattern resembles smaller part of itself. Although variations in cells are proven be closely associated with the disease-related phenotypes that otherwise obscured standard cell-based assays, analysis single-cell precision remains largely unexplored. To close this gap, here we develop an image-based approach quantifies multitude biophysical fractal-related properties at subcellular resolution. Taking together its high-throughput imaging performance (~10,000 cells/sec), technique, termed fractometry, offers sufficient statistical power for delineating cellular heterogeneity, context lung-cancer subtype classification, drug response assays cell-cycle progression tracking. Further correlative shows fractometry can enrich morphological profiling depth spearhead systematic how morphology encodes health pathological conditions.

Language: Английский

---

Electro-Optical Classification of Pollen Grains via Microfluidics and Machine Learning

IEEE Transactions on Biomedical Engineering, Journal Year: 2021, Volume and Issue: 69(2), P. 921 - 931

Published: Sept. 3, 2021

Objective: In aerobiological monitoring and agriculture there is a pressing need for accurate, label-free automated analysis of pollen grains, in order to reduce the cost, workload possible errors associated traditional approaches. xmlns:xlink="http://www.w3.org/1999/xlink">Methods: We propose new multimodal approach that combines electrical sensing optical imaging classify grains flowing microfluidic chip at throughput 150 per second. Electrical signals synchronized images are processed by two independent machine learning-based classifiers, whose predictions then combined provide final classification outcome. xmlns:xlink="http://www.w3.org/1999/xlink">Results: The applicability method demonstrated proof-of-concept experiment involving eight classes from different taxa. average balanced accuracy 78.7% classifier, 76.7% classifier 84.2% classifier. 82.8% 84.1% 88.3% xmlns:xlink="http://www.w3.org/1999/xlink">Conclusion: provides better results with respect based on or features alone. xmlns:xlink="http://www.w3.org/1999/xlink">Significance: proposed methodology paves way palynology. Moreover, it can be extended other fields, such as diagnostics cell therapy, where could used identification populations heterogeneous samples.

Language: Английский

---

Typing of acute leukemia by intelligent optical time-stretch imaging flow cytometry on a chip

Lab on a Chip, Journal Year: 2023, Volume and Issue: 23(6), P. 1703 - 1712

Published: Jan. 1, 2023

Intelligent optical time-stretch imaging flow cytometry on a chip is developed for high-throughput, high-accuracy and label-free acute leukemia typing.

Language: Английский

---