Virtual Staining Research Articles

Whole-Slide Images and Patches of Clear Cell Renal Cell Carcinoma Tissue Sections Counterstained with Hoechst 33342, CD3, and CD8 Using Multiple Immunofluorescence

In recent years, there has been an increased effort to digitise whole-slide images of cancer tissue. This effort has opened up a range of new avenues for the application of deep learning in oncology. One such avenue is virtual staining, where a deep learning model is tasked with reproducing the appearance of stained tissue sections, conditioned on a different, often times less expensive, input stain. However, data to train such models in a supervised manner where the input and output stains are aligned on the same tissue sections are scarce. In this work, we introduce a dataset of ten whole-slide images of clear cell renal cell carcinoma tissue sections counterstained with Hoechst 33342, CD3, and CD8 using multiple immunofluorescence. We also provide a set of over 600,000 patches of size 256 × 256 pixels extracted from these images together with cell segmentation masks in a format amenable to training deep learning models. It is our hope that this dataset will be used to further the development of deep learning methods for digital pathology by serving as a dataset for comparing and benchmarking virtual staining models.

Unstained Tissue Imaging and Virtual Hematoxylin and Eosin Staining of Histologic Whole Slide Images

Tissue structures, phenotypes, and pathology are routinely investigated based on histology. This includes chemically staining the transparent tissue sections to make them visible to the human eye. Although chemical staining is fast and routine, it permanently alters the tissue and often consumes hazardous reagents. On the other hand, on using adjacent tissue sections for combined measurements, the cell-wise resolution is lost owing to sections representing different parts of the tissue. Hence, techniques providing visual information of the basic tissue structure enabling additional measurements from the exact same tissue section are required. Here we tested unstained tissue imaging for the development of computational hematoxylin and eosin (HE) staining. We used unsupervised deep learning (CycleGAN) and whole slide images of prostate tissue sections to compare the performance of imaging tissue in paraffin, as deparaffinized in air, and as deparaffinized in mounting medium with section thicknesses varying between 3 and 20 μm. We showed that although thicker sections increase the information content of tissue structures in the images, thinner sections generally perform better in providing information that can be reproduced in virtual staining. According to our results, tissue imaged in paraffin and as deparaffinized provides a good overall representation of the tissue for virtually HE-stained images. Further, using a pix2pix model, we showed that the reproduction of overall tissue histology can be clearly improved with image-to-image translation using supervised learning and pixel-wise ground truth. We also showed that virtual HE staining can be used for various tissues and used with both 20× and 40× imaging magnifications. Although the performance and methods of virtual staining need further development, our study provides evidence of the feasibility of whole slide unstained microscopy as a fast, cheap, and feasible approach to producing virtual staining of tissue histology while sparing the exact same tissue section ready for subsequent utilization with follow-up methods at single-cell resolution.

Label-free virtual staining of neutrophil extracellular traps (NETs) in microfluidics.

Neutrophils are the most abundant circulating white blood cells and one of their critical functions to eliminate pathogenic threats includes the release of extracellular DNA, also known as neutrophil extracellular traps (NETs), which is dysregulated in many diseases including cancer, type 2 diabetes mellitus and infectious diseases. Currently, conventional methods to quantify the NET formation (NETosis) rely on fluorescence antibody-based NET labelling or circulating NET-associated protein detection by ELISA, which are expensive, laborious, and time-consuming. In this work, we employed a novel "virtual staining" using deep convolutional neural networks (CNNs) to facilitate label-free quantification of NETs trapped in a micropillar array in a microfluidic device. Virtual staining is constructed to establish relations between morphological features in phase contrast images and fluorescence features in Sytox-green (DNA dye) images. We first investigated the effect of different learning rates on model training and optimized the learning rate to achieve the best model which can provide outputs close to Sytox green staining based on various reconstruction metrics (e.g., structural similarity (SSIM) and pixel-wise error (MAE, MSE)). The virtual staining of different NET concentrations was investigated which showed a linear correlation with fluorescent staining. As a proof of concept for clinical testing, the model was used to characterize purified neutrophils treated with NETosis inducers, including lipopolysaccharide (LPS), phorbol 12-myristate 13-acetate (PMA), and calcium ionophore (CaI), and successfully detected different NET profiles for different treatments. Collectively, these results demonstrated the potential of using deep learning for enhanced label-free image analysis of NETs for clinical research, drug discovery and point-of-care testing of diseases.

CellVisioner: A Generalizable Cell Virtual Staining Toolbox based on Few-Shot Transfer Learning for Mechanobiological Analysis.

Visualizing cellular structures especially the cytoskeleton and the nucleus is crucial for understanding mechanobiology, but traditional fluorescence staining has inherent limitations such as phototoxicity and photobleaching. Virtual staining techniques provide an alternative approach to addressing these issues but often require substantial amount of user training data. In this study, we develop a generalizable cell virtual staining toolbox (termed CellVisioner) based on few-shot transfer learning that requires substantially reduced user training data. CellVisioner can virtually stain F-actin and nuclei for various types of cells and extract single-cell parameters relevant to mechanobiology research. Taking the label-free single-cell images as input, CellVisioner can predict cell mechanobiological status (e.g., Yes-associated protein nuclear/cytoplasmic ratio) and perform long-term monitoring for living cells. We envision that CellVisioner would be a powerful tool to facilitate on-site mechanobiological research.

VSGD-Net: Virtual Staining Guided Melanocyte Detection on Histopathological Images.

Detection of melanocytes serves as a critical prerequisite in assessing melanocytic growth patterns when diagnosing melanoma and its precursor lesions on skin biopsy specimens. However, this detection is challenging due to the visual similarity of melanocytes to other cells in routine Hematoxylin and Eosin (H&E) stained images, leading to the failure of current nuclei detection methods. Stains such as Sox10 can mark melanocytes, but they require an additional step and expense and thus are not regularly used in clinical practice. To address these limitations, we introduce VSGD-Net, a novel detection network that learns melanocyte identification through virtual staining from H&E to Sox10. The method takes only routine H&E images during inference, resulting in a promising approach to support pathologists in the diagnosis of melanoma. To the best of our knowledge, this is the first study that investigates the detection problem using image synthesis features between two distinct pathology stainings. Extensive experimental results show that our proposed model outperforms state-of-the-art nuclei detection methods for melanocyte detection. The source code and pre-trained model are available at: https://github.com/kechunl/VSGD-Net.

Rapid and label-free histological imaging of unprocessed surgical tissues via dark-field reflectance ultraviolet microscopy

Routine examination for intraoperative histopathologic assessment is lengthy and laborious. Here, we present the dark-field reflectance ultraviolet microscopy (DRUM) that enables label-free imaging of unprocessed and thick tissues with subcellular resolution and a high signal-to-background ratio. To the best of our knowledge, DRUM provides image results for pathological assessment with the shortest turnaround time (2-3min in total from sample preparation to tissue imaging). We also proposed a virtual staining process to convert DRUM images into pseudo-colorized images and enhance the image familiarity of pathologists. By imaging various tissues, we found DRUM can resolve cell nuclei and some extranuclear features, which are comparable to standard H&E images. Furthermore, the essential diagnostic features of intraoperatively excised tumor tissues also can be revealed by DRUM, demonstrating its potential as an additional aid for intraoperative histopathology.

Virtual Stain Transfer in Histology via Cascaded Deep Neural Networks

Pathological diagnosis relies on the visual inspection of histologically stained thin tissue specimens, where different types of stains are applied to bring contrast to and highlight various desired histological features. However, the destructive histochemical staining procedures are usually irreversible, making it very difficult to obtain multiple stains on the same tissue section. Here, we demonstrate a virtual stain transfer framework via a cascaded deep neural network (C-DNN) to digitally transform hematoxylin and eosin (H&E) stained tissue images into other types of histological stains. Unlike a single neural network structure that only takes one stain type as input to digitally output images of another stain type, C-DNN first uses virtual staining to transform autofluorescence microscopy images into H&E and then performs a stain transfer from H&E to the domain of the other stain in a cascaded manner. This cascaded structure in the training phase allows the model to directly exploit histochemically stained image data on both H&E and the target special stain of interest. This advantage alleviates the challenge of paired data acquisition and improves the image quality and color accuracy of the virtual stain transfer from H&E to another stain. We validated the superior performance of this C-DNN approach using kidney needle core biopsy tissue sections and successfully transferred the H&E stained tissue images into a virtual PAS (periodic acid-Schiff) stain. This method provides high-quality virtual images of special stains using existing, histochemically stained slides and creates new opportunities in digital pathology by performing highly accurate stain-to-stain transformations.

MVFStain: Multiple virtual functional stain histopathology images generation based on specific domain mapping.

To the best of our knowledge, artificial intelligence stain generation is an urgent requirement for histopathology images. Pathological examinations usually only utilize hematoxylin and eosin (H&E) regular staining to show histomorphological characteristics, but to accurately diagnose the disease, functional staining (such as oil red O and Ki67) are also required to provide important auxiliary information. However, the same tissue sectionis usually stained with one stain, and additional functional staining is not only time-consuming but also causes inevitable morphological misalignment due to manual manipulation. This brings difficulties to the development of artificial intelligence pathological image analysis tools. In this work, we propose a histopathology staining transfer framework to generate virtual functional staining images from H&E regular staining images. Compared with the framework that emphasizes generation diversity in the natural image field, we use KL loss and histo loss to align and separate style feature spaces in different domains to obtain domain-variant style features. The proposed multiple virtual functional stain (MVFStain) abstracts staining conversion to domain mapping and comprehensively utilizes multiple staining information. We evaluated the proposed method on four datasets (lung lesion, lung lobes, breast, and atherosclerotic lesion). The experiment involves the translation of H&E to nine other functional stains: CC10, Ki67, proSPC, HER2, PR, ER, oil red O, α-SMA, and macrophages. The major quantitative results are divided into image quality and positive signal prediction. MVFStain is close to or even surpasses one-to-one image translation on psnr and HTI image quality metrics. The best psnr reaches 26.1919, and HIT reaches 0.9430. We used mIOD to evaluate the optical density of positive signals, and CNR and gCNR to evaluate the lesion detectability. The results show that the mIOD of positive signals of virtual staining was slightly lower than the ground truth and close the lesion detectability of artificial staining. These results prove that the potential exists to develop a successful clinical alternative to artificial functional stains.

Modeling and simulation of diffusion and reaction processes during the staining of tissue sections on slides

Histological slides are an important tool in the diagnosis of tumors as well as of other diseases that affect cell shapes and distributions. Until now, the research concerning an optimal staining time has been mainly done empirically. In experimental investigations, it is often not possible to stain an already-stained slide with another stain to receive further information. To overcome these challenges, in the present paper a continuum-based model was developed for conducting a virtual (re-)staining of a scanned histological slide. This model is capable of simulating the staining of cell nuclei with the dye hematoxylin (C.I. 75,290). The transport and binding of the dye are modeled (i) along with the resulting RGB intensities (ii). For (i), a coupled diffusion–reaction equation is used and for (ii) Beer–Lambert’s law. For the spatial discretization an approach based on the finite element method (FEM) is used and for the time discretization a finite difference method (FDM). For the validation of the proposed model, frozen sections from human liver biopsies stained with hemalum were used. The staining times were varied so that the development of the staining intensity could be observed over time. The results show that the model is capable of predicting the staining process. The model can therefore be used to perform a virtual (re-)staining of a histological sample. This allows a change of the staining parameters without the need of acquiring an additional sample. The virtual standardization of the staining is the first step towards universal cross-site comparability of histological slides.

In vivo simultaneous nonlinear absorption Raman and fluorescence (SNARF) imaging of mouse brain cortical structures

Label-free multiphoton microscopy is a powerful platform for biomedical imaging. Recent advancements have demonstrated the capabilities of transient absorption microscopy (TAM) for label-free quantification of hemoglobin and stimulated Raman scattering (SRS) microscopy for pathological assessment of label-free virtual histochemical staining. We propose the combination of TAM and SRS with two-photon excited fluorescence (TPEF) to characterize, quantify, and compare hemodynamics, vessel structure, cell density, and cell identity in vivo between age groups. In this study, we construct a simultaneous nonlinear absorption, Raman, and fluorescence (SNARF) microscope with the highest reported in vivo imaging depth for SRS and TAM at 250–280 μm to enable these multimodal measurements. Using machine learning, we predict capillary-lining cell identities with 90% accuracy based on nuclear morphology and capillary relationship. The microscope and methodology outlined herein provides an exciting route to study several research topics, including neurovascular coupling, blood-brain barrier, and neurodegenerative diseases.

The potential of artificial intelligence-based applications in kidney pathology.

The field of pathology is currently undergoing a significant transformation from traditional glass slides to a digital format dependent on whole slide imaging. Transitioning from glass to digital has opened the field to development and application of image analysis technology, commonly deep learning methods (artificial intelligence [AI]) to assist pathologists with tissue examination. Nephropathology is poised to leverage this technology to improve precision, accuracy, and efficiency in clinical practice. Through a multidisciplinary approach, nephropathologists, and computer scientists have made significant recent advances in developing AI technology to identify histological structures within whole slide images (segmentation), quantification of histologic structures, prediction of clinical outcomes, and classifying disease. Virtual staining of tissue and automation of electron microscopy imaging are emerging applications with particular significance for nephropathology. AI applied to image analysis in nephropathology has potential to transform the field by improving diagnostic accuracy and reproducibility, efficiency, and prognostic power. Reimbursement, demonstration of clinical utility, and seamless workflow integration are essential to widespread adoption.

Virtual Staining of Defocused Autofluorescence Images of Unlabeled Tissue Using Deep Neural Networks

Deep learning-based virtual staining was developed to introduce image contrast to label-free tissue sections, digitally matching the histological staining, which is time-consuming, labor-intensive, and destructive to tissue. Standard virtual staining requires high autofocusing precision during the whole slide imaging of label-free tissue, which consumes a significant portion of the total imaging time and can lead to tissue photodamage. Here, we introduce a fast virtual staining framework that can stain defocused autofluorescence images of unlabeled tissue, achieving equivalent performance to virtual staining of in-focus label-free images, also saving significant imaging time by lowering the microscope’s autofocusing precision. This framework incorporates a virtual autofocusing neural network to digitally refocus the defocused images and then transforms the refocused images into virtually stained images using a successive network. These cascaded networks form a collaborative inference scheme: the virtual staining model regularizes the virtual autofocusing network through a style loss during the training. To demonstrate the efficacy of this framework, we trained and blindly tested these networks using human lung tissue. Using 4× fewer focus points with 2× lower focusing precision, we successfully transformed the coarsely-focused autofluorescence images into high-quality virtually stained H&E images, matching the standard virtual staining framework that used finely-focused autofluorescence input images. Without sacrificing the staining quality, this framework decreases the total image acquisition time needed for virtual staining of a label-free whole-slide image (WSI) by ~32%, together with a ~89% decrease in the autofocusing time, and has the potential to eliminate the laborious and costly histochemical staining process in pathology.

Virtual Staining, Segmentation, and Classification of Blood Smears for Label-Free Hematology Analysis.

Objective and Impact Statement. We present a fully automated hematological analysis framework based on single-channel (single-wavelength), label-free deep-ultraviolet (UV) microscopy that serves as a fast, cost-effective alternative to conventional hematology analyzers. Introduction. Hematological analysis is essential for the diagnosis and monitoring of several diseases but requires complex systems operated by trained personnel, costly chemical reagents, and lengthy protocols. Label-free techniques eliminate the need for staining or additional preprocessing and can lead to faster analysis and a simpler workflow. In this work, we leverage the unique capabilities of deep-UV microscopy as a label-free, molecular imaging technique to develop a deep learning-based pipeline that enables virtual staining, segmentation, classification, and counting of white blood cells (WBCs) in single-channel images of peripheral blood smears. Methods. We train independent deep networks to virtually stain and segment grayscale images of smears. The segmented images are then used to train a classifier to yield a quantitative five-part WBC differential. Results. Our virtual staining scheme accurately recapitulates the appearance of cells under conventional Giemsa staining, the gold standard in hematology. The trained cellular and nuclear segmentation networks achieve high accuracy, and the classifier can achieve a quantitative five-part differential on unseen test data. Conclusion. This proposed automated hematology analysis framework could greatly simplify and improve current complete blood count and blood smear analysis and lead to the development of a simple, fast, and low-cost, point-of-care hematology analyzer.

Resection-inspired histopathological diagnosis of cerebral cavernous malformations using quantitative multiphoton microscopy.

Rationale: Cerebral cavernous malformation (CCM) is prone to recurring microhemorrhage, which can lead to drug-resistant epilepsy. Surgical resection is the first choice to control seizures for CCM-associated epilepsy. At present, removal of resection-related tissue only depends on cautious visual identification of CCM lesions and perilesional yellowish hemosiderin rim by the neurosurgeon. Inspired by the resection requirements, we proposed quantitative multiphoton microscopy (qMPM) for a histopathology-level diagnostic paradigm to assist clinicians in precisely complete resection.Methods: A total of 35 sections specimens collected from 12 patients with the CCM-related epilepsy were included in this study. First, qMPM utilized a label-free multi-channel selective detection to image the histopathological features based on the spectral characteristics of CCM tissues. Then, qMPM developed three customized algorithms to provide quantitative information, a vascular patterns classifier based on linear support vector machine, visualization of microhemorrhage regions based on hemosiderin-related parameters, and the CCM-oriented virtual staining generative adversarial network (CCM-stainGAN) was constructed to generate two types of virtual staining.Results: Focused on CCM lesion and perilesional regions, qMPM imaged malformed vascular patterns and micron-scale hemosiderin-related products. Four vascular patterns were automatically identified by the classifier with an area under the receiver operating characteristic curve of 0.97. Moreover, qMPM mapped different degrees of hemorrhage regions onto fresh tissue while providing three quantitative hemosiderin-related indicators. Besides, qMPM realized virtual staining by the CCM-stainGAN with 98.8% diagnostic accuracy of CCM histopathological features in blind analysis. Finally, we provided pathologists and surgeons with the qMPM-based CCM histopathological diagnostic guidelines for a more definitive intraoperative or postoperative diagnosis.Conclusions: qMPM can provide decision-making supports for histopathological diagnosis, and resection guidance of CCM from the perspectives of high-resolution precision detection and automated quantitative assessment. Our work will promote the development of MPM diagnostic instruments and enable more optical diagnostic applications for epilepsy.

Novel Pixelwise Co-Registered Hematoxylin-Eosin and Multiphoton Microscopy Image Dataset for Human Colon Lesion Diagnosis

Colorectal cancer presents one of the most elevated incidences of cancer worldwide. Colonoscopy relies on histopathology analysis of hematoxylin-eosin (H&E) images of the removed tissue. Novel techniques such as multi-photon microscopy (MPM) show promising results for performing real-time optical biopsies. However, clinicians are not used to this imaging modality and correlation between MPM and H&E information is not clear. The objective of this paper is to describe and make publicly available an extensive dataset of fully co-registered H&E and MPM images that allows the research community to analyze the relationship between MPM and H&E histopathological images and the effect of the semantic gap that prevents clinicians from correctly diagnosing MPM images. The dataset provides a fully scanned tissue images at 10x optical resolution (0.5 µm/px) from 50 samples of lesions obtained by colonoscopies and colectomies. Diagnostics capabilities of TPF and H&E images were compared. Additionally, TPF tiles were virtually stained into H&E images by means of a deep-learning model. A panel of 5 expert pathologists evaluated the different modalities into three classes (healthy, adenoma/hyperplastic, and adenocarcinoma). Results showed that the performance of the pathologists over MPM images was 65% of the H&E performance while the virtual staining method achieved 90%. MPM imaging can provide appropriate information for diagnosing colorectal cancer without the need for H&E staining. However, the existing semantic gap among modalities needs to be corrected.

Simultaneous Morphology, Motility, and Fragmentation Analysis of Live Individual Sperm Cells for Male Fertility Evaluation

A new technique for sperm analysis is presented, measuring DNA fragmentation, morphology with virtual staining, and motility, all three criteria on the same individual unstained live cell. The method relies on quantitative stain‐free interferometric imaging, providing unique topographic structural and content maps of the cell, becoming available for the first time for clinical use, together with deep‐learning frameworks and least‐squares linear approximation. In the common clinical practice, only motility evaluation can be carried out on live human cells, while full morphological evaluation and DNA fragmentation assays require different staining protocols, and therefore cannot be performed on the same cell, resulting in inconsistencies in fertility evaluation. A clinic‐ready interferometric module is used to acquire dynamic sperm cells without chemical staining, together with deep learning to evaluate all three scores per cell with accuracy of 93.1%, 88%, and 90% for morphology, motility, and DNA fragmentation, respectively. It is shown that the expected number of cells that pass all three criteria based on the current evaluations performed separately does not correspond with the number of cells that pass all criteria, demonstrating the importance of the suggested method. The proposed stain‐free evaluation method is expected to decrease uncertainty in infertility diagnosis, increasing treatment success rates.

Biopsy-free in vivo virtual histology of skin using deep learning

An invasive biopsy followed by histological staining is the benchmark for pathological diagnosis of skin tumors. The process is cumbersome and time-consuming, often leading to unnecessary biopsies and scars. Emerging noninvasive optical technologies such as reflectance confocal microscopy (RCM) can provide label-free, cellular-level resolution, in vivo images of skin without performing a biopsy. Although RCM is a useful diagnostic tool, it requires specialized training because the acquired images are grayscale, lack nuclear features, and are difficult to correlate with tissue pathology. Here, we present a deep learning-based framework that uses a convolutional neural network to rapidly transform in vivo RCM images of unstained skin into virtually-stained hematoxylin and eosin-like images with microscopic resolution, enabling visualization of the epidermis, dermal-epidermal junction, and superficial dermis layers. The network was trained under an adversarial learning scheme, which takes ex vivo RCM images of excised unstained/label-free tissue as inputs and uses the microscopic images of the same tissue labeled with acetic acid nuclear contrast staining as the ground truth. We show that this trained neural network can be used to rapidly perform virtual histology of in vivo, label-free RCM images of normal skin structure, basal cell carcinoma, and melanocytic nevi with pigmented melanocytes, demonstrating similar histological features to traditional histology from the same excised tissue. This application of deep learning-based virtual staining to noninvasive imaging technologies may permit more rapid diagnoses of malignant skin neoplasms and reduce invasive skin biopsies.

Deep learning-enabled whole slide imaging (DeepWSI): oil-immersion quality using dry objectives, longer depth of field, higher system throughput, and better functionality.

Whole slide imaging (WSI) has moved the traditional manual slide inspection process to the era of digital pathology. A typical WSI system translates the sample to different positions and captures images using a high numerical aperture (NA) objective lens. Performing oil-immersion microscopy is a major obstacle for WSI as it requires careful liquid handling during the scanning process. Switching between dry objective and oil-immersion lens is often impossible as it disrupts the acquisition process. For a high-NA objective lens, the sub-micron depth of field also poses a challenge to acquiring in-focus images of samples with uneven topography. Additionally, it implies a small field of view for each tile, thus limiting the system throughput and resulting in a long acquisition time. Here we report a deep learning-enabled WSI platform, termed DeepWSI, to substantially improve the system performance and imaging throughput. With this platform, we show that images captured with a regular dry objective lens can be transformed into images comparable to that of a 1.4-NA oil immersion lens. Blurred images with defocus distance from -5 µm to +5 µm can be virtually refocused to the in-focus plane post measurement. We demonstrate an equivalent data throughput of >2 gigapixels per second, the highest among existing WSI systems. Using the same deep neural network, we also report a high-resolution virtual staining strategy and demonstrate it for Fourier ptychographic WSI. The DeepWSI platform may provide a turnkey solution for developing high-performance diagnostic tools for digital pathology.

Self-Attention Based Virtual Staining for Bright-field Images of Label-free Human Carotid Atherosclerotic Plaque Tissue Section.

Histological analysis of carotid atherosclerotic plaque tissue specimens is a widely used method for studying the diagnosis of ischemic heart disease and stroke. Understanding the physiological and pathological mechanisms of carotid atherosclerotic plaque is of great significance for the effective prevention and treatment of plaque formation and rupture. In this work, we adapted a self-attention generative adversarial model to virtually stain label-free human carotid atherosclerotic plaque tissue sections into corresponding H&E stained sections. The self-attention mechanism and multi-layer structure are introduced into the residual steps of the generator and in the discriminator. Our method achieved the best performance (SSIM, PSNR, and LPIPS of 0.53, 20.29, and 0.30, respectively) in comparison with other state-of-the-art methods.Clinical Relevance - The proposed approach allows for the virtual staining of unlabeled human carotid plaque tissue images. It identifies the histopathological features of atherosclerotic plaques in the same tissue sample which could facilitate the development of personalized prevention and other interventional treatments for carotid atherosclerosis.

Learning to see colours: Biologically relevant virtual staining for adipocyte cell images.

Fluorescence microscopy, which visualizes cellular components with fluorescent stains, is an invaluable method in image cytometry. From these images various cellular features can be extracted. Together these features form phenotypes that can be used to determine effective drug therapies, such as those based on nanomedicines. Unfortunately, fluorescence microscopy is time-consuming, expensive, labour intensive, and toxic to the cells. Bright-field images lack these downsides but also lack the clear contrast of the cellular components and hence are difficult to use for downstream analysis. Generating the fluorescence images directly from bright-field images using virtual staining (also known as "label-free prediction" and "in-silico labeling") can get the best of both worlds, but can be very challenging to do for poorly visible cellular structures in the bright-field images. To tackle this problem deep learning models were explored to learn the mapping between bright-field and fluorescence images for adipocyte cell images. The models were tailored for each imaging channel, paying particular attention to the various challenges in each case, and those with the highest fidelity in extracted cell-level features were selected. The solutions included utilizing privileged information for the nuclear channel, and using image gradient information and adversarial training for the lipids channel. The former resulted in better morphological and count features and the latter resulted in more faithfully captured defects in the lipids, which are key features required for downstream analysis of these channels.