Biology subjects

Duffy, S. P.

Publications and source records attributed to Duffy, S. P..

11 recordsLinked to original sources

bioRxiv2025

Regularized Single-cell Imaging Enables Generalizable AI models for Stain-free Cell Viability Screening

Cell viability assays are essential tools in biomedical research and drug development. Artificial intelligence (AI) offers the potential to simplify these assays by predicting cell viability directly from brightfield microscopy images, but current models lack generalizability across diverse cell types and treatments. Here, we introduce a strategy called "regularized imaging", where single cells are isolated in nanowells to generate standardized image patches that simplify segmentation and improve training data quality. We trained our model using example images of live and dead cells from a single cell line exposed to four cytotoxic conditions (ethanol, andrographolide, daunorubicin, and serum starvation). Despite this narrow training dataset, the resulting model accurately identified live and dead cells after treatment by previously unseen compounds, successfully replicating dose-response curves comparable to fluorescence live/dead assays. Importantly, this model effectively generalized across diverse cell types, including both adherent and suspension cells. Additionally, microscopy-based cell viability analysis is non-destructive, enabling repeated measurements to perform kinetic studies to distinguish between fast- and slow-acting compounds. Our findings highlight how regularized single cell imaging enables the training of broadly generalizable AI models to recognize biologically relevant cell features for label-free cell screening workflows. One-sentence SummaryRegularized single-cell imaging in nanowells enables training of generalizable AI models for accurate, stain-free viability screening across previously unseen cell types and contexts.

bioengineering↗

bioRxiv2025

High-throughput Single Cell Motility Analysis using Nanowell-in-Microwells

Cell motility is important to many biological processes including cancer, immune response, and tissue repair. Conventional assays measure bulk cell motility, potentially overlooking important heterogeneity and missing important high motility subpopulations. Here, we introduce a high-throughput single-cell motility assay using nanowell-in-microwell plates to precisely track single cell position and analyze their migratory trajectories. By physically confining individual cells in nanowells, we eliminate cell-cell interactions and simplify cell segmentation and tracking. Using this platform, we characterized the motility of single cells across different culture conditions to identify distinct motility phenotypes. Single-cell trajectory analysis revealed pronounced directional persistence, with cells predominantly maintaining their direction of travel and trajectory along nanowell boundaries. Additionally, our approach facilitates the generation of labeled image datasets suitable for AI models to rapidly identify cell motility phenotypes from single-cell images. Together, our platform provides a robust, scalable method to analyze cell motility phenotypes and migration behavior at single-cell resolution. Significance StatementA novel single-cell motility assay that uses confinement in nanowell-in-microwell to enable high-throughput profiling of single cell motility phenotypes.

bioengineering↗

bioRxiv2022

Amplified cortical neural responses as animals learn to use novel activity patterns

Cerebral cortex supports representations of the world in patterns of neural activity, used by the brain to make decisions and guide behavior. Past work has found diverse, or limited, changes in the primary sensory cortex in response to learning, suggesting the key computations might occur in downstream regions. Alternatively, sensory cortical changes may be central to learning. We studied cortical learning by using controlled inputs we insert: we trained mice to recognize entirely novel, non-sensory patterns of cortical activity in the primary visual cortex (V1) created by optogenetic stimulation. As animals learned to use these novel patterns, we found their detection abilities improved by an order of magnitude or more. The behavioral change was accompanied by large increases in V1 neural responses to fixed optogenetic input. Neural response amplification to novel optogenetic inputs had little effect on existing visual sensory responses. A recurrent cortical model shows that this amplification can be achieved by a small mean shift in recurrent network synaptic strength. Amplification would seem to be desirable to improve decision-making in a detection task, and therefore these results suggest that adult recurrent cortical plasticity plays a significant role in improving behavioral performance during learning.

neuroscience↗

bioRxiv2021

Deep Learning-based Automated Rare Sperm Identification from Testes Biopsies

Non-obstructive azoospermia (NOA), the most severe form of male infertility, is currently treated using microsurgical sperm extraction (microTESE) to retrieve sperm cells for in vitro fertilization via intracytoplasmic sperm injection (IVF-ICSI). The success rate of this procedure for NOA patients is currently limited by the ability of andrologists to identify a few rare sperm cells among millions of background testis cells. To improve this success rate, we developed a convolution neural network (CNN) to detect rare sperm from low-resolution microscopy images of microTESE samples. Our CNN uses the U-Net architecture to perform pixel-based classification on image patches from brightfield microscopy, which is followed by morphological analysis to detect individual sperm instances. This CNN is trained using microscopy images of fluorescently labeled sperm, which is fixed to eliminate their motility, and doped into testis biopsies obtained from NOA patients. We initially tested this algorithm using purified sperm samples at different imaging magnifications in order to determine the upper bounds of performance. We then tested this algorithm by doping rare sperm cells into testis biopsy samples from NOA patients and found a sperm detection F1 score of 85.2%. These results demonstrate the potential to use automated microscopy to dramatically increase the amount of testis biopsy tissue that could be comprehensively examined, which greatly increases the chance of finding rare viable sperm, and thereby increases the success rates of IVF-ICSI for couples with NOA.

bioengineering↗

bioRxiv2021

Deep Learning Image Classification of Red Blood Cell Deformability

Red blood cells (RBCs) must be highly deformable to transit through the microvasculature to deliver oxygen to tissues. The loss of RBC deformability resulting from pathology, natural aging, or storage in blood bags can impede the proper function of these cells. A variety of methods have been developed to measure RBC deformability, but these methods require specialized equipment, long measurement time, and highly skilled personnel. To address this challenge, we investigated whether a machine learning approach could be applied to determine donor RBC deformability using single cell microscope images. We used the microfluidic ratchet device to sort RBCs based on deformability. Sorted cells are then imaged and used to train a deep learning model to classify RBCs based on deformability. This model correctly predicted deformability of individual RBCs with 84 {+/-} 11% accuracy averaged across ten donors. Using this model to score the deformability of RBC samples were accurate to within 4.4 {+/-} 2.5% of the value obtained using the microfluidic ratchet device. While machine learning methods are frequently developed to automate human image analysis, our study is remarkable in showing that deep learning of single cell microscopy images could be used to measure RBC deformability, a property not normally measurable by imaging. Measuring RBC deformability by imaging is also desirable because it can be performed rapidly using a standard microscopy system, potentially enabling RBC deformability studies to be performed as part of routine clinical assessments.

bioengineering↗

bioRxiv2021

Degradation of Red Blood Cell Deformability during Cold Storage in Blood Bags

Red blood cells (RBCs) stored in blood bags develop a storage lesion that include structural, metabolic, and morphologic transformations resulting in a progressive loss of RBC deformability. The speed of RBC deformability loss is donor-dependent, which if properly characterized, could be used as a biomarker to select high-quality RBC units for sensitive recipients or to provide customized storage timelines depending on the donor. We used the microfluidic ratchet device to measure the deformability of red blood cells stored in blood bags every 14 days over a span of 56 days. We observed that storage in blood bags generally prevented RBC deformability loss over the current standard 42-day storage window. However, between 42 and 56 days, the deformability loss profile varied dramatically between donors. In particular, we observed accelerated RBC deformability loss for a majority of male donors, but for none of the female donors. Together, our results suggest that RBC deformability loss could be used to screen for donors who can provide stable RBCs for sensitive transfusion recipients or to identify donors capable of providing RBCs that could be stored for longer than the current 42-day expiration window.

bioengineering↗

bioRxiv2021

Blood Unit Segments Accurately Represent the Biophysical Properties of Red Blood Cells in Blood Bags but Not Hemolysis

BACKGROUNDThe biophysical properties of red blood cells (RBCs) provide potential biomarkers for the quality of donated blood. Blood unit segments provide a simple and nondestructive way to sample RBCs in clinical studies of transfusion efficacy, but it is not known whether RBCs sampled from segments accurately represent the biophysical properties of RBCs in blood bags. STUDY DESIGN AND METHODSRBCs were sampled from blood bags and segments every two weeks during 8 weeks of storage at 4 {degrees}C. RBC deformability was measured by deformabilitybased sorting using the microfluidic ratchet device in order to derive a rigidity score. Standard hematological parameters, including mean corpuscular volume (MCV), red cell distribution width (RDW), mean cell hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), and hemolysis were measured at the same time points. RESULTSDeformability of RBCs stored in blood bags was retained over 4 weeks storage but a progressive loss of deformability was observed at weeks 6 and 8. This trend was mirrored in blood unit segments with a strong correlation to the blood bag data. Strong correlations were also observed between blood bag and segment for MCV, MCHC and MCH, but not for hemolysis. CONCLUSIONRBCs sampled from blood unit segments accurately represents the biophysical properties of RBCs in blood bags, but not hemolysis. Blood unit segments provide a simple and non-destructive sample for measuring RBC biophysical properties in clinical studies.

bioengineering↗

bioRxiv2020

Selective Cell Propagation via Micropatterning of Thermal-activated Hydrogel

The ability to selectively propagate specific cells is fundamentally important to the development of clonal cell populations. Current methods rely on techniques such as limiting dilution, colony picking, and flow cytometry to transfer single cells into single wells, resulting in workflows that are low-throughput, slowed by propagation kinetics, and susceptible to contamination. Here, we developed a method, called selective laser gelation (SLG), to micropattern hydrogels in cell culture media in order to encapsulate specific cells to selectively arrest their growth. This process relies on the inverse gelation of methylcellulose, which forms a hydrogel when heated rather than cooled. Local heating using an infrared laser enables hydrogel micropatterning, while phase transition hysteresis retains the hydrogel after laser excitation. As a demonstration, we used this approach to selectively propagate transgenic CHO cells with increased antibody productivity. More generally, hydrogel micropatterning provides a simple and non-contact method to selective propagation of cells based on features identified by imaging. One Sentence SummaryInverse gelation of methylcellulose enables hydrogel micropatterning to selectively propagate cells identified by microscopy.

bioengineering↗

bioRxiv2020

Lossless Immunocytochemistry using Photo-polymerized Hydrogel Thin-films

Immunocytochemistry (ICC), or immunofluorescence microscopy, is an essential biological technique for phenotyping cells in both research and diagnostic applications. Standard ICC methods often do not work well when the cell sample contains a small number of cells (<10,000) because of the significant cell loss that occurs during washing, staining, and centrifugation steps. Cell loss is particularly relevant when working with rare cells, such as circulating tumor cells, where such losses could significantly bias experimental outcomes. In order to eliminate cell loss in ICC protocols, we present a method to encapsulate the cell sample in a photo-polymerized hydrogel thin-film. The hydrogel thin-film is permeable to antibodies and other ICC reagents, thereby allowing the use of standard ICC protocols without modification. The cell sample is physically constrained by the hydrogel at the bottom surface of a standard (unmodified) imaging microtiter plate, thereby enabling the acquisition of high-quality micrographs regardless of the properties of the cell sample or staining reagents. Furthermore, while standard ICC requires several centrifugation steps during staining and washing, our hydrogel encapsulation method requires only a single centrifugation step. This property greatly reduces the time required to perform ICC protocols and is more compatible with robotic platforms. In this study, we show that standard ICC and Cytospin protocols are extremely lossy (>70% loss) when the sample contains less than 10,000 cells, while encapsulating the cells using a permeable hydrogel thin-film results in a lossless ICC process.

bioengineering↗

bioRxiv2019

Deformability based sorting of stored red blood cells reveals donor-dependent aging curves

A fundamental challenge in the transfusion of red blood cells (RBCs) is that a subset of donated RBC units may not provide optimal benefit to transfusion recipients. This variability stems from the inherent ability of donor RBCs to withstand the physical and chemical insults of cold storage, which ultimately dictate their survival in circulation. The loss of RBC deformability during cold storage is well-established and has been identified as a potential biomarker for the quality of donated RBCs. While RBC deformability has traditionally been indirectly inferred from rheological characteristics of the bulk suspension, there has been considerable interest in directly measuring the deformation of RBCs. Microfluidic technologies have enabled single cell measurement of RBC deformation but have not been able to consistently distinguish differences between RBCs between healthy donors. Using the microfluidic ratchet mechanism, we developed a method to sensitively and consistently analyze RBC deformability. We found that the aging curve of RBC deformability varies significantly across donors, but is consistent for each donor over multiple donations. Specifically, certain donors seem capable of providing RBCs that maintain their deformability during two weeks of cold storage in standard test tubes. The ability to distinguish between RBC units with different storage potential could provide a valuable opportunity to identify donors capable of providing RBCs that maintain their integrity, in order to reserve these units for sensitive transfusion recipients.

bioengineering↗

bioRxiv2019

Image-based Cell Phenotyping Using Deep Learning

The ability to phenotype cells is fundamentally important in biological research and medicine. Current methods rely primarily on fluorescence labeling of specific markers. However, there are many situations where this approach is unavailable or undesirable. Machine learning has been used for image cytometry but has been limited by cell agglomeration and it is unclear if this approach can reliably phenotype cells indistinguishable to the human eye. Here, we show disaggregated single cells can be phenotyped with a high degree of accuracy using low-resolution bright-field and non-specific fluorescence images of the nucleus, cytoplasm, and cytoskeleton. Specifically, we trained a convolutional neural network using automatically segmented images of cells from eight standard cancer cell-lines. These cells could be identified with an average classification accuracy of 94.6%, tested using separately acquired images. Our results demonstrate the potential to develop an \"electronic eye\" to phenotype cells directly from microscopy images indistinguishable to the human eye.

bioinformatics↗