Automated Image Acquisition Research Articles

Fast, high-precision autofocus on a motorised microscope: Automating blood sample imaging on the OpenFlexure Microscope.

The OpenFlexure Microscope is a 3D-printed, low-cost microscope capable of automated image acquisition through the use of a motorised translation stage and a Raspberry Pi imaging system. This automation has applications in research and healthcare, including in supporting the diagnosis of malaria in low-resource settings. The plasmodium parasites that cause malaria require high magnification imaging, which has a shallow depth of field, necessitating the development of an accurate and precise autofocus procedure. We present methods of identifying the focal plane of the microscope, and procedures for reliably acquiring a stack of focused images on a system affected by backlash and drift. We also present and assess a method to verify the success of autofocus during the scan. The speed, reliability and precision of each method are evaluated, and the limitations discussed in terms of the end users'requirements.

All-in-one: A spectral imaging laboratory system for standardised automated image acquisition and real-time spectral model deployment

Spectral imaging (SI) in analytical chemistry is widely used for the assessment of spatially distributed physicochemical properties of samples. Although massive development in instrument and chemometrics modelling has taken place in the recent years, the main challenge with SI is that available sensors require extensive system integration and calibration modelling before their use for routine analysis. Further, the models developed during one experiment are rarely useful once the system is reintegrated for a new experiment. To avoid system reintegration and reuse calibrated models, this study presents an intelligent All-In-One SI (ASI) laboratory system allowing standardised automated data acquisition and real-time spectral model deployment. The ASI system supplies a controlled standardised illumination environment, an in-built computing system, embedded software for automated image acquisition, and model deployment to predict the spatial distribution of sample properties in real-time. To show the capability of the ASI framework, exemplary cases of fruit property prediction in different fruits are presented. Furthermore, ASI is also benchmarked in performance against the current commercially available portable as well as high-end laboratory spectrometers.

Exploring a new paradigm for the fetal anomaly ultrasound scan: Artificial intelligence in real time.

Advances in artificial intelligence (AI) have demonstrated potential to improve medical diagnosis. We piloted the end-to-end automation of the mid-trimester screening ultrasound scan using AI-enabled tools. A prospective method comparison study was conducted. Participants had both standard and AI-assisted US scans performed. The AI tools automated image acquisition, biometric measurement, and report production. A feedback survey captured the sonographers' perceptions of scanning. Twenty-three subjects were studied. The average time saving per scan was 7.62min (34.7%) with the AI-assisted method (p<0.0001). There was no difference in reporting time. There were no clinically significant differences in biometric measurements between the two methods. The AI tools saved a satisfactory view in 93% of the cases (four core views only), and 73% for the full 13 views, compared to 98% for both using the manual scan. Survey responses suggest that the AI tools helped sonographers to concentrate on image interpretation by removing disruptive tasks. Separating freehand scanning from image capture and measurement resulted in a faster scan and altered workflow. Removing repetitive tasks may allow more attention to be directed identifying fetal malformation. Further workis required to improve the image plane detection algorithm for use in real time.

PSI-16 Estimation of pigs live body weight from digital images using reference objects

Abstract The live body weight (LBW) is an important parameter providing guidance for estimation of growth and feed conversion efficiency, body condition, presence of disease, and management of housing, nutrition and animal health in different life stages of livestock. This research study explores the possibility of developing a semi-automatic analytic system that estimates the LBW of pigs by applying machine learning methods that use approximated biometric measurements extracted from digital images acquired with consumer-level cameras in the presence of a reference object. Images corresponding to 12 pigs were sampled on two different dates 1 month apart and acquired using a consumer-level Motorola X4 mobile phone. The best 3 images for each pig were selected for each time point. Six measurements were extracted from each image using ImageJ. A total of 72 data points were analyzed using RStudio, and the generated correlation plot confirmed the positive correlations between the 6 predictors and LBW. Five machine learning (ML) methods were used to model the dependency between the measured parameters and LBW using WEKA. The Random Forest model outperformed all the other models and predicted LBW with the highest prediction accuracy (97%) and the lowest prediction error (MAE = 6.54), which could represent a good candidate for further studies. In conclusion, the semi-automatic image-based system is a promising approach combining machine learning, digital image analysis and manual scale extraction with the aid of reference objects of known size for accurate pig LBW estimation. The next stage of this study aims to integrate automatic image acquisition and image processing solutions in the current approach. This novel system will be cost- and time-efficient, and it can contribute to the development of intelligent solutions for scientific research and enhancing animal productivity in commercial pig farms.

IoT based Automated Plant Disease Classification using Support Vector Machine

Leaf - a significant part of the plant, produces food using the process called photosynthesis. Leaf disease can cause damage to the entire plant and eventually lowers crop production. Machine learning algorithm for classifying five types of diseases, such as Alternaria leaf diseases, Bacterial Blight, Gray Mildew, Leaf Curl and Myrothecium leaf diseases, is proposed in the proposed study. The classification of diseases needs front face of leafs. This paper proposes an automated image acquisition process using a USB camera interfaced with Raspberry PI SoC. The image is transmitted to host PC for classification of diseases using online web server. Pre-processing of the acquired image by host PC to obtain full leaf, and later classification model based on SVM is used to detect type diseases. Results were checked with a 97% accuracy for the collection of acquired images.

An Artificial Intelligence-Assisted Portable Low-Cost Device for the Rapid Detection of SARS-CoV-2

An artificial intelligence-assisted low-cost portable device for the rapid detection of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is presented here. This standalone temperature-controlled device houses tubes designed for conducting reverse transcription loop-mediated isothermal amplification (RT-LAMP) assays. Moreover, the device utilises tubes illuminated by LEDs, an in-built camera, and a small onboard computer with automated image acquisition and processing algorithms. This intelligent device significantly reduces the normal assay run time and removes the subjectivity associated with operator interpretation of colourimetric RT-LAMP results. To further improve this device’s usability, a mobile app has been integrated into the system to control the LAMP assay environment and to visually display the assay results by connecting the device to a smartphone via Bluetooth. This study was undertaken using ~5000 images produced from the ~200 LAMP amplification assays using the prototype device. Synthetic RNA and a small panel of positive and negative SARS-CoV-2 patient samples were assayed for this study. State-of-the-art image processing and artificial intelligence algorithms were applied to these images to analyse them and to select the most efficient algorithm. The template matching algorithm for image extraction and MobileNet CNN architecture for classification results provided 98.0% accuracy with an average run time of 20 min to confirm the endpoint result. Two working points were chosen based on the best compromise between sensitivity and specificity. The high sensitivity point has a sensitivity value of 99.12% and specificity value of 70.8%, while at the high specificity point, the sensitivity is 96.05% and specificity 93.59%. Furthermore, this device provides an efficient and cost-effective platform for non-health professionals to detect not only SARS-CoV-2 but also other pathogens in resource-limited laboratories, factories, airports, schools, universities, and homes.

User-friendly, High-throughput, and Fully Automated Data Acquisition Software for Single-particle Cryo-electron Microscopy

Single-particle cryo-electron microscopy demands a suitable software package and user-friendly pipeline for high-throughput automatic data acquisition. Here, we present the application of a fully automated image acquisition software package, Latitude-S, and a practical pipeline for data collection of vitrified biomolecules under low-dose conditions.

Review of Methods for Animal Videography Using Camera Systems that Automatically Move to Follow the Animal.

Digital photography and videography provide rich data for the study of animal behavior and are consequently widely used techniques. For fixed, unmoving cameras there is a resolution versus field-of-view tradeoff and motion blur smears the subject on the sensor during exposure. While these fundamental tradeoffs with stationary cameras can be sidestepped by employing multiple cameras and providing additional illumination, this may not always be desirable. An alternative that overcomes these issues of stationary cameras is to direct a high-magnification camera at an animal continually as it moves. Here, we review systems in which automatic tracking is used to maintain an animal in the working volume of a moving optical path. Such methods provide an opportunity to escape the tradeoff between resolution and field of view and also to reduce motion blur while still enabling automated image acquisition. We argue that further development will be useful and outline potential innovations that may improve the technology and lead to more widespread use.

New tropolone derivative to inhibit cell motility in colorectal cancer cells.

e15043 Background: Tropolones are small, naturally occurring organic compounds that have anti-cancer effects in a variety of cancer cell lines. Among proposed mechanisms of tropolone derivatives anti-cancer activity are caspase activation, matrix metalloproteinases and histone deacetylases inhibition and other actions. Since matrix metalloproteinases are the major proteins that contribute to tumor cell tissue invasion and migration their blockers are extensively studied as prospective anti metastatic agents. We assumed that tropolones may also have potential as metastases prevention drugs. The aim of the study was to investigate the influence of new tropolone derivative 2-(6,8-dimethyl-5-nitro-4-chloroquinolin-2-yl)-5,6,7-trichloro-1,3-tropolone on the cell migration capacity in colorectal cancer cells. Methods: Cell migration in permanent colorectal cancer cell lines HT29, Caco2 and HCT116 was studied using the standard wound healing assay with automated image acquisition and analysis on Lionheart FX (BioTek) cell imager. Cell cultures were maintained in DMEM medium with L-glutamine (1 μM) (Gibco) and 10% FBS (Gibco) at 37C0 and 5.0% CO2. The day before assay establishment cells were seeded at 250 000 cells per well on 24-well plates. After attachment of cells, a vertical wound in cell layer was made manually in each well, and then medium with studied tropolone derivative (5 μM) was added. Plates with cells were continuously incubated and photographed in cell imager at 37C0 and 5.0% CO2. The extent of cells migration was measured as the percent of wound area decrease after 48 hours of incubation in relation to starting time point. Data are given as: Mean ± 95% confidence interval. Results: The new tropolone derivative under study exhibited pronounced anti migratory activity in all studied colorectal cancer cell cultures. The most prominent effect was observed in HCT116 cell line: after 48 hours of incubation relative wound area decrease was 48.2%±5.7% in control and 7.2%±3.15% in experimental groups. In Caco2 cell line cell migration inhibition was the least with 65%± 6.9% relative wound area decrease in control and 51.5%± 7.7% in experimental groups. The moderate effect was observed in HT29 cell line (measured parameter was 63.6%±7.6% and 36.9%±11.6% in control and experimental groups accordingly). The obtained difference between control and experimental groups was significant in all tested cell cultures at 0.05 level (df = 30). Conclusions: Results of the present study prove tropolone derivative to be prospective candidates for anti-metastatic drug development. Further investigations are needed to reveal the exact mechanisms of tropolones anti-migratory activity.

Antimigratory effect of berberine in T98G, U87MG and primary glioma cell culture.

e15045 Background: Berberine is an alkaloid compound with a structure that is highly similar to that of intercalating agents. It affects numerous cell signaling pathways and is widely studied as potential anticancer drug. It is known that berberine affects cancer cells migration through metalloproteinase-2 inhibition, but this effect was never studied on glioma cells. Anti-migratory drugs are of special interest in brain cancer therapy since glioma's highly invasive nature makes total surgical removal of tumor practically impossible. The aim of the study was to evaluate berberine anti-migratory activity on glioma cells. Methods: Cell migration capacity of T98G and U87MG cell lines, as well as primary glioma cell culture established in our laboratory, was assessed via standard wound healing assay with automated image acquisition and analysis on Lionheart FX (BioTek) cell imager. Prior to assay setting up cell cultures were maintained in DMEM medium with L-glutamine (1 μM) (Gibco) and 10% FBS (Gibco) at 37C0 and 5.0% CO2. Cells were seeded at 250 000 cells per well on 24-well plates and incubated overnight in order to attach to plate bottom. After that a vertical wound was made manually in each well, and berberine was added to experimental wells to final concentration 50 mg/L. Plates with cells were continuously incubated and photographed in cell imager at 37C0 and 5.0% CO2. The extent of cells migration was measured as the percent of wound area decrease after 24 hours of incubation in relation to starting time point. Data are given as: Mean ± 95% confidence interval. Results: In our study we berberine exhibited anti-migratory activity in all cell cultures under study. In rather fast growing primary cell culture wound area decrease was 99.23%±0.62% in control sample and 91.75%±0.28% in experimental sample. The difference was small but significant at p &lt; 0.001 level (df = 30). Popular permanent glioma cell lines T98G and U87MG showed more prominent decrease in studied parameter with higher degree of variance at the same time. In T98G wound area decrease was 71.6%±12.3% in control and 48.8%± 7.6% in experimental samples after 24 hours of cultivation in presence of 50 mg/L berberine. While U87MG demonstrated 60.28%±5.13% and 37.5%± 8.34% wound area decrease accordingly. The obtained difference between control and experimental groups in permanent cell cultures was statistically significant at the 0.05 level (df = 30). Conclusions: Our preliminary research proved berberine to be potent anti-migratory agent in glioma treatment. Further investigations are needed to evaluate its ability to inhibit glioma cell expansion in vivo.

Electric-Field-Induced Neural Precursor Cell Differentiation in Microfluidic Devices.

Physiological electric fields (EF) play vital roles in cell migration, differentiation, division, and death. This paper describes a microfluidic cell culture system that was used for a long-term cell differentiation study using microscopy. The microfluidic system consists of the following major components: an optically transparent electrotactic chip, a transparent indium-tin-oxide (ITO) heater, a culture media-filling pump, an electrical power supply, a high-frequency power amplifier, an EF multiplexer, a programmable X-Y-Z motorized stage, and an inverted phase-contrast microscope equipped with a digital camera. The microfluidic system is beneficial in simplifying the overall experimental setup and, in turn, the reagent and sample consumption. This work involves the differentiation of neural stem and progenitor cells (NPCs) induced by direct current (DC) pulse stimulation. In the stem cell maintenance medium, the mouse NPCs (mNPCs) differentiated into neurons, astrocytes, and oligodendrocytes after the DC pulse stimulation. The results suggest that simple DC pulse treatment could control the fate of mNPCs and could be used to develop therapeutic strategies for nervous system disorders. The system can be used for cell culture in multiple channels, for long-term EF stimulation, for cell morphological observation, and for automatic time-lapse image acquisition. This microfluidic system not only shortens the required experimental time, but also increases the accuracy of control on the microenvironment.

Novel method for coke optical texture measurement and application on coke derived from different ranks of coking coal

AbstractThe optical texture of coke is undoubtedly one of the important factors that determined many properties of coke, but it still mostly depends on the skills and experience of manual analysis. In this paper, a novel automatic measurement method is proposed for coke optical texture. this method through two analyzer angle images(10° and 170°) under polarizing microscope to recognize isochromatic regions which occurs color changed between blue and red by image processing, then characterize the dimensional sizes of isochromatic regions and classify them into different anisotropic texture type. Two fitted ellipse methods are used to characterize the regions' dimensional sizes by the range of shape factorS, followed by classifying different types of anisotropic texture according to the long and short axes of fitted ellipse. An automatic image acquisition and analysis system is established based on the above method. Finally, eight coke samples derived from different ranks of coking coal were selected to test the proposed method. The result shows that it can effectively recognize different types of anisotropic texture and provide composition of different anisotropic texture type for each coke. Moreover, the anisotropic degree of coke shows a significant positive correlation with parent coal rank and a negative correlation with coke reactivity index (CRI).

A Data Augmentation Method for Deep Learning Based on Multi-Degree of Freedom (DOF) Automatic Image Acquisition

Deep learning technology is outstanding in visual inspection. However, in actual industrial production, the use of deep learning technology for visual inspection requires a large number of training data with different acquisition scenarios. At present, the acquisition of such datasets is very time-consuming and labor-intensive, which limits the further development of deep learning in industrial production. To solve the problem of image data acquisition difficulty in industrial production with deep learning, this paper proposes a data augmentation method for deep learning based on multi-degree of freedom (DOF) automatic image acquisition and designs a multi-DOF automatic image acquisition system for deep learning. By designing random acquisition angles and random illumination conditions, different acquisition scenes in actual production are simulated. By optimizing the image acquisition path, a large number of accurate data can be obtained in a short time. In order to verify the performance of the dataset collected by the system, the fabric is selected as the research object after the system is built, and the dataset comparison experiment is carried out. The dataset comparison experiment confirms that the dataset obtained by the system is rich and close to the real application environment, which solves the problem of dataset insufficient in the application process of deep learning to a certain extent.

Artificial Intelligence-Assisted Loop Mediated Isothermal Amplification (AI-LAMP) for Rapid Detection of SARS-CoV-2.

Until vaccines and effective therapeutics become available, the practical solution to transit safely out of the current coronavirus disease 19 (CoVID-19) lockdown may include the implementation of an effective testing, tracing and tracking system. However, this requires a reliable and clinically validated diagnostic platform for the sensitive and specific identification of SARS-CoV-2. Here, we report on the development of a de novo, high-resolution and comparative genomics guided reverse-transcribed loop-mediated isothermal amplification (LAMP) assay. To further enhance the assay performance and to remove any subjectivity associated with operator interpretation of results, we engineered a novel hand-held smart diagnostic device. The robust diagnostic device was further furnished with automated image acquisition and processing algorithms and the collated data was processed through artificial intelligence (AI) pipelines to further reduce the assay run time and the subjectivity of the colorimetric LAMP detection. This advanced AI algorithm-implemented LAMP (ai-LAMP) assay, targeting the RNA-dependent RNA polymerase gene, showed high analytical sensitivity and specificity for SARS-CoV-2. A total of ~200 coronavirus disease (CoVID-19)-suspected NHS patient samples were tested using the platform and it was shown to be reliable, highly specific and significantly more sensitive than the current gold standard qRT-PCR. Therefore, this system could provide an efficient and cost-effective platform to detect SARS-CoV-2 in resource-limited laboratories.

Machine Learning Algorithms for Automated NIF Capsule Mandrel Selection

Small shells, approximately 2 mm in diameter, made from Poly(α-methylstyrene) (PAMS) are used as mandrels in the production of glow discharge polymer capsules located at the center of inertial confinement fusion experiments. The visual inspection process of microscope images of these shell mandrels, including detection of micron-sized defects on the shell surface as well as the handling and sorting, is a very labor-intensive, repetitive, and highly subjective process that stands to benefit greatly from automation. As part of an effort to decrease the number of labor hours spent in capsule handling, inspection, and metrology, the development of robotic systems was presented in a paper by Carlson et al., “Automation in Target Fabrication” [Fusion Sci. Technol., Vol. 70, p. 274 (2016)]. The current work expands the automated image acquisition systems developed previously and adds the use of convolutional neural networks to select capsules best suited for use in the downstream production process. Through the use of these machine learning algorithms, the selection process becomes robust, repeatable, and operator independent. As an added benefit the system developed as part of this work is able to provide defect statistics on entire shell batches and feed this information upstream to the production team.

3084 – A MACHINE-LEARNING APPROACH TO SCORING ERYTHROID COLONIES IN SEMI-SOLID MEDIUM

The ability of erythroid progenitors to form colonies in semi-solid medium is the ‘gold-standard' approach to establishing their identity. However, colony formation assays are laborious, requiring multiple replicates and manual scoring by skilled observers. These drawbacks limit the scale of their use in the laboratory, in fundamental research and in the search for novel erythroid stimulating agents. Here we present a novel tool, ‘c-count', which combines high-resolution automated image acquisition with a convolutional neural network (CNN) machine-learning model to accurately count colony-forming-unit-erythroid (CFU-e) colonies. We performed colony assays in a methylcellulose medium supplemented with erythropoietin (Epo) and iron-saturated transferrin, in 35 mm plates. Colonies were stained for expression of hemoglobin with diaminobenzidine. Images were acquired using the Zeiss Axio Observer inverted microscope. Sixteen contiguous fields were scanned and stitched together in each of 4 identically placed regions totaling 1/16th of each plate's area. To generate data to train and test the CNN model, a Laplacian of Gaussian algorithm was used to identify ‘blobs' within the images, which were then labeled by 5 independent observers as either true or false CFU-e. The training data contained 1062 blobs with 420 labeled as CFU-e, while the validation data contained 188 blobs. The F1 score among the 5 independent observers on the same test data ranged from 0.84 to 0.89 (mean=0.87). The F1 score of the c-count classification against the validation data was 0.87. In an additional test experiment, the correlations between CFU-e counts generated by c-count and two independent observers were 0.96 and 0.97. We are currently extending evaluation to determine the model's performance under a variety of experimental conditions including tissue origin of CFU-e and culture conditions. We conclude that c-count is a promising tool for scoring CFU-e with the potential to speed up the study of erythroid progenitor biology.

Dielectric track detectors in fast neutron measurements and dosimetry

The main aim of the presented paper was to perform neutron measurements using Solid State Nuclear Track Detectors and to compare obtained results with those obtained on the basis of Monte Carlo simulations of the experimental setup. The CR-39 type track detectors coupled with various radiators were used to measure neutrons emitted from a DT neutron generator. The detectors were also shielded with various filters made of materials characterized by low cross sections for reactions with fast neutrons. The filters were used to block charged particles with energies lower than certain threshold. The track detectors after exposure to neutrons were etched in steps, and examined using an optical microscope equipped with automatic image acquisition system. The recorded images were analyzed using a specialized software. Based on the recorded track parameters, it is also possible to calculate linear energy transfer (LET) of the registered particles, and determine the radiation dose of the emitted neutrons.

Miniaturized Continuous-Flow Digital PCR for Clinical-Level Serum Sample Based on the 3D Microfluidics and CMOS Imaging Device.

In recent years, the development of polymerase chain reaction (PCR) technology has focused on digital PCR, which depends on the microfluidics. Based on continuous-flow microfluidic technology, this paper designed a miniaturized digital PCR amplification system, and greatly reduced the area required for microdroplet generation and reaction. The core rod. made of polydimethylsiloxane (PDMS), was combined with the Teflon tube to form 3D microfluidics, which requires only one heating source to form the temperature difference required for gene amplification. Only two 34 g needles can form and transmit micro-droplets in a 4-fold tapered Teflon tube, which is the simplest method to generate digital PCR droplets as far as we know, which allows the microdroplet generation device to be free from dependence on expensive chips. A complementary metal oxide semiconductor (CMOS) camera was used as a detection tool to obtain fluorescence video for the entire loop area or a specified loop area. In addition, we developed a homebrew for automatic image acquisition and processing to realize the function of digital PCR. This technique realizes the analysis of clinical serum samples of hepatitis B virus (HBV) and obtained the same results as real-time quantitative PCR. This system has greatly reduced the size and cost of the entire system, while maintaining a stable response.

Live circulating tumour cells selection on digitized self-assembled cell array (Digi-saca) chip by in-parallel/in-situ image analysis, cell capture, and cultivation

Cancer is one of the major and most deadly diseases of mankind. With present technologies, early detection and prevention of cancer disease is still a major bottleneck. Circulating Tumor Cells (CTC) is one of the cancer biomarkers for the diagnosis of metastasis. However, there is an urgent need to improve the turn around time of CTC detection from a small sample volume in order to promote its clinical implications. This paper highlights Digi-SACA, an automated workflow to capture CTCs from the whole blood on self-assembled cell array chips. The gravity force and lateral driving force based microfluidic chip drive the mononuclear blood cells from 4 ml of samples to form a monolayer without any external fluid control equipment. The subsequent immunostaining and automated image acquisition, image processing, CTC enumeration, and CTC harvest can be completed within 4 h. The sensitivity of Digi-SACA chip is 1 in ten million leukocytes which is very promising for early detection of the CTC. In a small pilot study series, the Digi-SACA chip is able to detect of the mean of 14.4 CTCs from 4 ml of blood in 10 clinical samples from patients with breast cancer, while mean of 17.6 CTCs were measured from the same sample set with IsoFlux™ platform from 4 ml of blood. We have also demonstrated the efficiency of our technique to pick CTC with glass micropipettes for standard cell culture growth up to 11–15 days. Overall, our study represents the liability of the Digi-SACA system in CTCs enumeration, detection, and isolation against the state-of-the-art.

Industrial cryo-EM facility setup and management.

Cryo-electron microscopy (cryo-EM) has rapidly expanded with the introduction of direct electron detectors, improved image-processing software and automated image acquisition. Its recent adoption by industry, particularly in structure-based drug design, creates new requirements in terms of reliability, reproducibility and throughput. In 2016, Thermo Fisher Scientific (then FEI) partnered with the Medical Research Council Laboratory of Molecular Biology, the University of Cambridge Nanoscience Centre and five pharmaceutical companies [Astex Pharmaceuticals, AstraZeneca, GSK, Sosei Heptares and Union Chimique Belge (UCB)] to form the Cambridge Pharmaceutical Cryo-EM Consortium to share the risks of exploring cryo-EM for early-stage drug discovery. The Consortium expanded with a second Themo Scientific Krios Cryo-EM at the University of Cambridge Department of Materials Science and Metallurgy. Several Consortium members have set up in-house facilities, and a full service cryo-EM facility with Krios and Glacios has been created with the Electron Bio-Imaging Centre for Industry (eBIC for Industry) at Diamond Light Source (DLS), UK. This paper will cover the lessons learned during the setting up of these facilities, including two Consortium Krios microscopes and preparation laboratories, several Glacios microscopes at Consortium member sites, and a Krios and Glacios at eBIC for Industry, regarding site evaluation and selection for high-resolution cryo-EM microscopes, the installation process, scheduling, the operation and maintenance of the microscopes and preparation laboratories, and image processing.