Generation Step Research Articles

High-Accuracy Filtering of Forest Scenes Based on Full-Waveform LiDAR Data and Hyperspectral Images

Airborne light detection and ranging (LiDAR) technology has been widely utilized for collecting three-dimensional (3D) point cloud data on forest scenes, enabling the generation of high-accuracy digital elevation models (DEMs) for the efficient investigation and management of forest resources. Point cloud filtering serves as the crucial initial step in DEM generation, directly influencing the accuracy of the resulting DEM. However, forest filtering presents challenges in dealing with sparse point clouds and selecting appropriate initial ground points. The introduction of full-waveform LiDAR data offers a potential solution to the problem of sparse point clouds. Additionally, advancements in multi-source data integration and machine learning algorithms have created new avenues that can address the issue of initial ground point selection. To tackle these challenges, this paper proposes a novel filtering method for forest scenes utilizing full-waveform LiDAR data and hyperspectral image data. The proposed method consists of two main steps. Firstly, we employ the improved dynamic graph convolutional neural network (IDGCNN) to extract initial ground points. In this step, we utilize three types of low-correlation features: LiDAR features, waveform features, and spectral features. To enhance its accuracy and adaptability, a self-attention module was incorporated into the DGCNN algorithm. Comparative experiments were conducted to evaluate the effectiveness of the algorithm, demonstrating that the IDGCNN algorithm achieves the highest classification accuracy with an overall accuracy (OA) value of 99.38% and a kappa coefficient of 95.95%. The second-best performer was the RandLA-net algorithm, achieving an OA value of 98.73% and a kappa coefficient of 91.68%. The second step involves refining the initial ground points using the cloth simulation filter (CSF) algorithm. By employing the CSF algorithm, non-ground points present in the initial ground points are effectively filtered out. To validate the efficacy of the proposed filtering method, we generated a DEM with a resolution of 0.5 using the ground points extracted in the first step, the refined ground points obtained with the combination of the first and second steps, and the ground points obtained directly using the CSF algorithm. A comparative analysis with 23 reference control points revealed the effectiveness of our proposed method, as evidenced by the median error of 0.41 m, maximum error of 0.75 m, and average error of 0.33 m.

In-situ production of lactate driving the biotransformation of waste activated sludge to medium-chain fatty acid

Medium-chain fatty acids (MCFAs) have drawn great attention due to their high energy density and superior hydrophobicity. Waste activated sludge (WAS) has been documented as a renewable feedstock for MCFAs production via anaerobic fermentation. However, MCFAs production from WAS depends on exogenous addition of electron donor (ED, e.g., lactate) for chain elongation (CE) bioprocess, which results in increased economic cost and limited practical application. In this study, a novel biotechnology was proposed to produce MCFAs from WAS with in-situ self-formed lactate by inoculating Yoghurt starter powder containing with Lactobacillales cultures. The batch experimental results revealed that the lactate was in-situ generated from WAS and the maximum production of MCFAs increased from 1.17 to 3.99 g COD/L with the increased addition of Lactobacillales cultures from 6✕107 to 2.3✕108 CFU/mL WAS. In continuous long-term test over 97 days, average MCFA production reached up to 3.94 g COD/L with a caproate yield of 82.74% at sludge retention time (SRT) 12 days, and the average MCFA production increased to 5.87 g COD/L with 69.28% caproate and 25.18% caprylate at SRT 15 days. A comprehensive analysis of the metagenome and metatranscriptome demonstrated that the genus of Lactobacillus and Streptococcus were capable of producing lactate from WAS and upgrading to MCFAs. Moreover, another genus, i.e., Candidatus Promineofilum, was firstly revealed that it might be responsible for lactate and MCFAs production. Further investigation of related microbial pathways and enzyme expression suggested that D-lactate dehydrogenase and pyruvate ferredoxin oxidoreductase contributed to lactate and acetyl-CoA production, which were the crucial steps for MCFAs generation and were most actively expressed. This study provides a conceptual framework of MCFAs from WAS with endogenous ED, potentially enhancing the energy recovery from WAS treatment.

Evidence-based modalities in the management of Psoriasis and Psoriatic arthritis.

Psoriasis is a waxing and waning skin disorder, often associated with a plethora of co-morbidities, including psoriatic arthritis (PsA), a severe form of chronic inflammatory arthritis. All forms of psoriasis and PsA are immune-mediated diseases where the patient's immune system is overactive in the production of certain factors that stimulate and activate the function of certain immune cells. Recent evidence has uncovered an important role for cell-mediated immunity in the aetiology and course of psoriasis and PsA, with a critical role played by the pro-inflammatory IL-23/TH17 axis. Taken together, these new lines of evidence suggest new and improved therapeutic interventions for patients with psoriasis and PsA. The hypothesis-driven process of inquiry of the best available evidence and its implication, application and evaluation in the context of clinical practice pertains to the meta-science of evidence-based health care (EBHC). EBHC consists in the initial step of research synthesis and generation of the systematic review of the best available evidence, estimated both qualitatively and quantitatively (i.e., meta-analysis). Evidence-based decision-making, a process driven and controlled by the expertise of the clinician and by the clinical needs and personal wants of the patient, is the principal, most timely and critical aspect of evidence-based practice. Recent and systematic reviews for the treatment of psoriasis and PsA consistently updated for emerging new and revised data (i.e., living systematic reviews) confirm the efficacy and the effectiveness of methotrexate (MTX) in containing and controlling psoriasis. The outcomes of MTX intervention for PsA remain mixed and inconclusive.

Design for Additive Manufacturing: Recent Innovations and Future Directions

Design for additive manufacturing (DfAM) provides a necessary framework for using novel additive manufacturing (AM) technologies for engineering innovations. Recent AM advances include shaping nickel-based superalloys for lightweight aerospace applications, reducing environmental impacts with large-scale concrete printing, and personalizing food and medical devices for improved health. Although many new capabilities are enabled by AM, design advances are necessary to ensure the technology reaches its full potential. Here, DfAM research is reviewed in the context of Fabrication, Generation, and Assessment phases that bridge the gap between AM capabilities and design innovations. Materials, processes, and constraints are considered during fabrication steps to understand AM capabilities for building systems with specified properties and functions. Design generation steps include conceptualization, configuration, and optimization to drive the creation of high-performance AM designs. Assessment steps are necessary for validating, testing, and modeling systems for future iterations and improvements. These phases provide context for discussing innovations in aerospace, automotives, construction, food, medicine, and robotics while highlighting future opportunities for design services, bio-inspired design, fabrication robots, and machine learning. Overall, DfAM has positively impacted diverse engineering applications, and further research has great potential for driving new developments in design innovation.

DyRRen: A Dynamic Retriever-Reranker-Generator Model for Numerical Reasoning over Tabular and Textual Data

Numerical reasoning over hybrid data containing tables and long texts has recently received research attention from the AI community. To generate an executable reasoning program consisting of math and table operations to answer a question, state-of-the-art methods use a retriever-generator pipeline. However, their retrieval results are static, while different generation steps may rely on different sentences. To attend to the retrieved information that is relevant to each generation step, in this paper, we propose DyRRen, an extended retriever-reranker-generator framework where each generation step is enhanced by a dynamic reranking of retrieved sentences. It outperforms existing baselines on the FinQA dataset.

MIDMs: Matching Interleaved Diffusion Models for Exemplar-Based Image Translation

We present a novel method for exemplar-based image translation, called matching interleaved diffusion models (MIDMs). Most existing methods for this task were formulated as GAN-based matching-then-generation framework. However, in this framework, matching errors induced by the difficulty of semantic matching across cross-domain, e.g., sketch and photo, can be easily propagated to the generation step, which in turn leads to the degenerated results. Motivated by the recent success of diffusion models, overcoming the shortcomings of GANs, we incorporate the diffusion models to overcome these limitations. Specifically, we formulate a diffusion-based matching-and-generation framework that interleaves cross-domain matching and diffusion steps in the latent space by iteratively feeding the intermediate warp into the noising process and denoising it to generate a translated image. In addition, to improve the reliability of diffusion process, we design confidence-aware process using cycle-consistency to consider only confident regions during translation. Experimental results show that our MIDMs generate more plausible images than state-of-the-art methods.

Predictive evaluation of solar energy variables for a large-scale solar power plant based on triple deep learning forecast models

The advanced development of large-scale solar power plants (LSSPs) has made it necessary to improve accurate forecasting models for the output of solar energy. Solar energy is still hampered by the lack of predictability in its output, which remains a major hurdle in the solar industry. This paper focuses on triple deep learning (DL) techniques such as Artificial Neural Network (ANN), Recurrent Neural Network (RNN) and Convolutional Neural Network- Long-Short Term Memory CNN-LSTM to address this problem. These techniques are utilized in solar energy variables (SEVs) such as power generation (MWh), soiling loss (%), and performance ratio (PR %) to determine the optimal forecast model. The novelty of this research is that it is the first time that important solar system parameters such as PR and soiling loss have been studied to predict a feasible forecast model using a different DL scheme. The SEVs real-time dataset is procured from the largest solar plant in Pakistan, titled “Quaid-e-Azam Solar Park” (QASP). The main significance of the study is that ANN, RNN, and CNN-LSTM-based models were developed in the DL process through feature generation, data scaling, training, and testing steps to predict the optimal model. The prediction values were compared with the solar plant's actual values over the last 7 years, and then a comparison was made to predict the future forecast trend over the next 20 years. The aim and goal is to develop three models to investigate the accurate results of time-series forecasting on the SEVs dataset, as well as to evaluate the performance measure errors to determine the appropriate model. Based on the forecasting/prediction graphic and error results, it was demonstrated that the CNN-LSTM hybrid model is a more capable forecasting model for the output of power generation and PR values and a more accurate predictor of the future trend over the ANN and RNN models. However, the ANN model is slightly better performed in predicting of soling loss value than CNN-SLTM and RNN. Thus, the CNN-LSTM hybrid model is an optimal model for SEVs, which can guarantee a variety of LSSPs of similar nature in the following time-series forecasting investigations which shows key findings and its importance for the solar industrial forecast issue.

CuZK: Accelerating Zero-Knowledge Proof with A Faster Parallel Multi-Scalar Multiplication Algorithm on GPUs

Zero-knowledge proof is a critical cryptographic primitive. Its most practical type, called zero-knowledge Succinct Non-interactive ARgument of Knowledge (zkSNARK), has been deployed in various privacy-preserving applications such as cryptocurrencies and verifiable machine learning. Unfortunately, zkSNARK like Groth16 has a high overhead on its proof generation step, which consists of several time-consuming operations, including large-scale matrix-vector multiplication (MUL), number-theoretic transform (NTT), and multi-scalar multiplication (MSM). Therefore, this paper presents cuZK, an efficient GPU implementation of zkSNARK with the following three techniques to achieve high performance. First, we propose a new parallel MSM algorithm. This MSM algorithm achieves nearly perfect linear speedup over the Pippenger algorithm, a well-known serial MSM algorithm. Second, we parallelize the MUL operation. Along with our self-designed MSM scheme and well-studied NTT scheme, cuZK achieves the parallelization of all operations in the proof generation step. Third, cuZK reduces the latency overhead caused by CPU-GPU data transfer by 1) reducing redundant data transfer and 2) overlapping data transfer and device computation. The evaluation results show that our MSM module provides over 2.08x (up to 2.94x) speedup versus the state-of-the-art GPU implementation. cuZK achieves over 2.65x (up to 4.86x) speedup on standard benchmarks and 2.18× speedup on a GPU-accelerated cryptocurrency application, Filecoin.

CARVING-DETC: A network scaling and NMS ensemble for Balinese carving motif detection method

Balinese carvings are cultural objects that adorn sacred buildings. The carvings consist of several motifs, each representing the values adopted by the Balinese people. Detection of Balinese carving motifs is challenging due to the unavailability of a Balinese carving dataset for detection tasks, high variance, and tiny-size carving motifs. This research aims to improve carving motif detection performance on challenging Balinese carving motifs detection task through a modification of YOLOv5 to support a digital carving conservation system. We proposed CARVING-DETC, a deep learning-based Balinese carving detection method consisting of three steps. First, the data generation step performs data augmentation and annotation on Balinese carving images. Second, we proposed a network scaling strategy on the YOLOv5 model and performed non-maximum suppression (NMS) on the model ensemble to generate the most optimal predictions. The ensemble model utilizes NMS to produce higher performance by optimizing the detection results based on the highest confidence score and suppressing other overlap predictions with a lower confidence score. Third, performance evaluation on scaled-YOLOv5 versions and NMS ensemble models. The research findings are beneficial in conserving the cultural heritage and as a reference for other researchers. In addition, this study proposed a novel Balinese carving dataset through data collection, augmentation, and annotation. To our knowledge, it is the first Balinese carving dataset for the object detection task. Based on experimental results, CARVING-DETC achieved a detection performance of 98%, which outperforms the baseline model.

Pyruvate enhances stallion sperm function increasing glycolysis efficiency through NAD+

Pyruvate entry into the mitochondrial matrix is a critical step for the efficient generation of reducing equivalents and ATP (Yiew and Finck, 2022). Improvements in sperm function using metabolites readily metabolizable through the tricarboxylic acid cycle (TCA), have been attributed to their ability to readily feed the electron transport chain (ETC) in the mitochondria; 8 NADH and 2 FADH2 are produced in the TCA cycle. However, noncanonical mechanisms may explain improvements in sperm functions. For example, the direct use of lactate in the TCA (Martinez-Reyes and Chandel, 2017) or a more efficient glycolysis (Hereng et al., 2011); in both mechanisms lactate dehydrogenase (LDH) plays a major role. If pyruvate improves the efficiency of glycolysis, the detrimental effects of higher glucose concentrations could be counteracted by increasing pyruvate concentration in the media. To test this hypothesis stallion spermatozoa were incubated in presence of high glucose concentration (67mM) in presence of 1- or 10-mM pyruvate. To investigate the role of LDH, we used the specific inhibitor oxamate. If our hypothesis is correct improved functionality will be observed in high pyruvate media. Flow cytometry, CASA, and metabolomics were used to disclose the potential mechanisms behind this. Stallion spermatozoa (n=4, ANOVA and Dunnett´s test) incubated for 3 h at 37°C in 67mM Tyrode´s media supplemented with 10 mM pyruvate showed increased total motility in comparison with aliquots incubated in Tyrode's 5mM glucose and Tyrode´s67mM glucose, from 57.1 ± 3.5 and 58.1 ± 1.9 to 73.0± 1.1 % (P<0.01). Spermatozoa incubated in Tyrode's with 67 mM glucose and 10mM pyruvate maintained viability along the incubation period (64.03 ± 15.4% vs 61.3 ± 10.2%), while spermatozoa incubated in 67 mM Glucose-Tyrode's showed a decrease in viability (38.01 ± 11.2, P<0.01). The presence of 40 mM oxamate (LDH inhibitor) reduced sperm viability (P<0.05, from 76± 5 % in the media containing 67 mM Glucose and 10 mM pyruvate to 68.0 ± 4.3 % in presence of oxamate, P<0.05). Apoptotic markers also increased when the LDH inhibitor oxamate was present (P<0.01). The incorporation of glucose was reduced in the presence of 10mM pyruvate, suggesting that the stallion spermatozoa have mechanisms to regulate the incorporation of this hexose. Furthermore, UHPLC/MS/MS showed that in presence of 10 mM pyruvate, the relative amounts of lactate, ATP, NAD+, and NADH increased. Our findings indicate that the mechanisms that explain the improvement of sperm function in presence of 10 mM pyruvate may involve the conversion of pyruvate to lactate producing NAD+ that improves the efficiency of glycolysis. Acknowledgements: Ministerio de Ciencia-European Fund for Regional Development (EFRD) Spain, grants PID2021-122351OB-I00 and PRE2018-083354 Junta de Extremadura-EFDR (IB 20008).

Visual Paraphrase Generation with Key Information Retained

Visual paraphrase generation task aims to rewrite a given image-related original sentence into a new paraphrase, where the paraphrase needs to have the same expressed meaning as the original sentence but have a difference in expression form. Existing studies mainly extract two semantic vectors to represent the entire image and the entire original sentence, respectively, for paraphrase generation. However, these semantic vectors for an image or a sentence may lead to the model failing to focus on some key objects in the original sentence, which may generate semantically inconsistent sentences by changing key object information. In this article, we propose an object-level paraphrase generation model, which generates paraphrases by adjusting the permutation of key objects and modifying their associated descriptions. To adjust the permutation of key objects, an object-sorting module aims to obtain new object sequences based on the key object information and original sentences. Then, a sequence generation module sequentially generates paraphrases based on the permutation of the newly object sequences. Each generation step focuses on different image features associated with different key objects to generate descriptions with differences. Furthermore, we use a semantic discriminator module to promote the generated paraphrase to be semantically close to the original sentence. Specifically, the loss function of the discriminator penalizes the excessive distance between the paraphrase and the original sentence. Extensive experiments on the MS COCO dataset show that the proposed model outperforms the baselines.

Most Expected Winner: An Interpretation of Winners over Uncertain Voter Preferences

It remains an open question how to determine the winner of an election when voter preferences are incomplete or uncertain. One option is to assume some probability space over the voting profile and select the Most Probable Winner (MPW) -- the candidate or candidates with the best chance of winning. In this paper, we propose an alternative winner interpretation, selecting the Most Expected Winner (MEW) according to the expected performance of the candidates. We separate the uncertainty in voter preferences into the generation step and the observation step, which gives rise to a unified voting profile combining both incomplete and probabilistic voting profiles. We use this framework to establish the theoretical hardness of MEW over incomplete voter preferences, and then identify a collection of tractable cases for a variety of voting profiles, including those based on the popular Repeated Insertion Model (RIM) and its special case, the Mallows model. We develop solvers customized for various voter preference types to quantify the candidate performance for the individual voters, and propose a pruning strategy that optimizes computation. The performance of the proposed solvers and pruning strategy is evaluated extensively on real and synthetic benchmarks, showing that our methods are practical.

Developing a Targeting Approach for Syngas Generation from Natural Gas

The conversion of natural and unconventional gas into syngas is a crucial intermediate step in the production of various important chemicals and liquid fuels. The syngas generation step usually requires the largest capital investment of the process and may also be very energy intensive. Therefore, determining the most efficient method of converting feedstock into syngas of the correct H2:CO ratio is of significant importance. The aim of this work was to set design and performance targets for different H2:CO ratios (depending on the downstream requirements) in terms of the carbon efficiency (including CO2 utilization or emissions), water usage, and energy requirements. It was shown that the overall process for natural gas tri-reforming is limited by the enthalpy change (ΔH = 0) and this process was able to produce work. It was further shown that high syngas ratios not only require significant amounts of natural gas and oxygen but also emit CO2.

Hierarchical structure regulation for sequential steps in solar vapor generation

AbstractIncluding the transpiration of leaves, water evaporation from solar irradiation is a universal phenomenon in nature. Currently, solar vapor generation allows clean water to be obtained from various waterbodies. Since water is transported through porous structures and evaporates on their surfaces, the properties of the nano‐micro structure, especially the surfaces are significantly important. For instance, the surface energy, determined by the localized atomic arrangement, can modify the interactions between water and the substrate. Moreover, the construction of a three‐dimensional hierarchical structure can efficiently enlarge the surface area, and the provided channels play a vital role in mass transfer. In this review, we summarize recent research on the structural regulation in tuning the sequential steps in photo‐vapor generation. We hope this review can provide a rational and systemic basis for the development of advanced solar vapor generating materials, especially from the view of surface engineering.image

Transfer from Lunar Gateway to Sun-Earth Halo Orbits Using Solar Sails

To extend the usability of solar sails in the sun–Earth–moon system, we analyze the transfer trajectories from the 9:2 Earth–moon near-rectilinear halo orbit (NRHO) to halo orbits around the sun–Earth L1 and L2 points under the assumption of a future mission for a solar sail spacecraft equipped with a solar electric propulsion (SEP) system deployed from the Lunar Orbital Platform-Gateway. The dynamics are modeled using the bicircular restricted four-body problem, where the gravitational forces from the sun, Earth, and moon as well as solar radiation pressure (SRP) are considered. We propose a trajectory design method that utilizes both SRP and SEP. The method consists of initial guess generation and optimization steps. The initial guess generation comprises the forward propagation of the escape trajectory from the NRHO, the backward propagation of the stable manifold of the target halo orbits, and their apoapsis patching process. Optimization is conducted to minimize propellant consumption by effectively controlling SRP. We perform optimizations with various parameters, namely, the sail area-to-mass ratio (), specifications of SEP, target sun–Earth halo orbit, and departure direction. The results validate the proposed trajectory design method and verify that solar sail acceleration can reduce the necessary amount of propellant, which indicates that such missions can be realized by small CubeSats.

Expression, purification and micelle reconstruction of the transmembrane domain of the human amyloid precursor protein for NMR studies

The multiple-step cleavage of amyloid precursor protein (APP) generates amyloid-β peptides (Aβ), highly toxic molecules causing Alzheimer's disease (AD). The nonspecific cleavage between the transmembrane region of APP (APPTM) and γ-secretase is the key step of Aβ generation. Reconstituting APPTM under physiologically-relevant conditions is crucial to investigate how it interacts with γ-secretase and for future AD drug discovery. Although producing recombinant APPTM was reported before, the large scale purification was hindered by the use of biological protease in the presence of membrane protein. Here, we expressed recombinant APPTM in Escherichia coli using the pMM-LR6 vector and recovered the fusion protein from inclusion bodies. By combining Ni-NTA chromatography, cyanogen bromide cleavage, and reverse phase high performance liquid chromatography (RP-HPLC), isotopically-labeled APPTM was obtained in high yield and high purity. The reconstitution of APPTM into dodecylphosphocholine (DPC) micelle generated mono dispersed 2D 15N-1H HSQC spectra in high quality. We successfully established an efficient and reliable method for the expression, purification and reconstruction of APPTM, which may facilitate future investigation of APPTM and its complex in more native like membrane mimetics such as bicelle and nanodiscs.

A new deep learning approach for hyperspectral image classification based on multifeature local kernel descriptors

During the last few years, many deep learning (DL) methods have been proposed for the classification of hyperspectral images (HSI), and the final results show that these models have great potential. However, most existing deep models only consider the original spectral HSI datacube, excluding other sources of information such as textural and geometrical features. In addition to their computational complexity, most DL approaches cannot effectively consider local nonlinear relationships between features. In order to address these issues, this paper develops a new DL approach called MLKD-CNN, which uses multifeature local kernel descriptors (MLKDs) and convolutional neural networks (CNN) to classify HSIs. This method produces local kernel descriptors (LKDs) of spectral, extended morphological profiles, and Gabor features in the feature generation step. A simple CNN is then used to classify MLKD generated from spatial and spectral features. The proposed MLKD considers local nonlinear relationships of both spectral and spectral features, making them capable of achieving high accuracy levels while reducing the computational burden of deep methods. The experiments on four benchmark HSIs show that the proposed MLKD-CNN outperforms some recent HSI classification methods in terms of accuracy and computation time.

Experimental investigation of thermochemical syngas production in a scrap iron-based oxidizer reactor for industrial decarbonisation

The syngas production performance of scrap iron reacting with carbon dioxide and water steam was assessed under different operating conditions in a fixed bed oxidizer reactor. The syngas generation step is part of a novel process scheme encompassing the reutilization of iron scrap from steelmaking and the combined splitting of industrially captured carbon dioxide and steam into a syngas. At 1050 °C, a maximum volume percentage of 37 % carbon monoxide was detected in the product gas with the injection of 1 NL/min carbon dioxide. The carbon dioxide conversion was confirmed to be promoted by temperature. Subsequently, combined tests with carbon dioxide and water steam were carried out to assess the production and quality of the syngas (H2 and CO) by varying the reactants total flow rate, the iron bed mass and the reactants molar ratio. By decreasing the total reactants flow rate, the reactants splitting process was promoted and below a certain flow rate carbon dioxide splitting prevailed on that of water steam. By increasing the H2Ov/CO2 molar ratio, the splitting was enhanced for both species. In particular, for the tested flow rate the water splitting increased by 10% compared to 3.5% of the CO2 splitting. This indicated that a high H2Ov/CO2 ratio optimizes syngas production in the designed system. Finally, with H2Ov/CO2 = 6 and the optimal thermochemical syngas composition was achieved, including 41 % H2 and 12.1 % CO, being the remaining part constituted by CO2 when computed on a dry basis.

Reference‐based Screentone Transfer via Pattern Correspondence and Regularization

AbstractAdding screentone to initial line drawings is a crucial step for manga generation, but is a tedious and human‐laborious task. In this work, we propose a novel data‐driven method aiming to transfer the screentone pattern from a reference manga image. This not only ensures the quality, but also adds controllability to the generated manga results. The reference‐based screentone translation task imposes several unique challenges. Since manga image often contains multiple screentone patterns interweaved with line drawing, as an abstract art, this makes it even more difficult to extract disentangled style code from the reference. Also, finding correspondence for mapping between the reference and the input line drawing without any screentone is hard. As screentone contains many subtle details, how to guarantee the style consistency to the reference remains challenging. To suit our purpose and resolve the above difficulties, we propose a novel Reference‐based Screentone Transfer Network (RSTN). We encode the screentone style through a 1D stylegram. A patch correspondence loss is designed to build a similarity mapping function for guiding the translation. To mitigate the generated artefacts, a pattern regularization loss is introduced in the patch‐level. Through extensive experiments and a user study, we have demonstrated the effectiveness of our proposed model.

Abstract 1117: Rapid label-free imaging-based evaluation of cancer dependencies in zero-passage primary cells

Abstract To successfully map cancer dependencies, it is essential to conduct genetic and pharmacological screenings in a diversity of cell models. However, existing model development approaches require long periods of culture time and it is difficult to create long-term models of many cancers, greatly limiting the share of patient samples that can be studied. To enable high-throughput perturbation screens in primary cells without an intermediate model generation step, we are developing a label-free imaging-based platform for early living tissue perturbation. Here, we present our ex vivo system for the preservation and morphological characterization of malignant ascites from patients with gastroesophageal cancer, whose prognosis remains poor and there is an urgent need for rapid evidence-based therapeutic discovery. First, we established a workflow to acquire and perturb cells within 1 day of sample collection. We found that mixing ascites fluid with organoid media improves the preservation of cellular composition and viability of the samples. Next, we hypothesized that label-free microscopy can be a potential alternative for fluorescence-based biomarkers of which signal fades over time in live-cell imaging. To test this, we designed a systematic approach for data generation to assess the reproducibility of measures, and we built a dataset consisting of over 1.0M cells from 14 samples (10 unique patients). For training input, we extracted morphological features of each identified cell by generating several projections from z-stacks of label-free bright-field microscopy images. For training labels, we used fluorescence labels to annotate cell type and viability during the imaging screens. Then, using solely bright-field morphological data as input, we trained models to infer cell identity and viability, we found that the accuracy of predictions was 92% and 82%, respectively. Strong correlations were found between tumor fractions determined by flow cytometry and the prediction of tumor cell fraction from label-free morphology only. Based on single-cell RNA sequencing data, we designed a candidate panel of 28 compounds that are anticipated to exhibit antitumor activity via different mechanisms that are of relevance to our study cohort. We observed that label-free inference of compound activity showed a strong correlation (R2 &amp;gt; 0.8) with fluorescent-based predictions. We are now expanding the scale of our rapid screens by finding the minimal number of necessary z-stacks of bright-field and fluorescent channels maintaining the prediction accuracy of cell identity and viability. By finding the appropriate minimum setting for imaging setup, the throughput of the system can be increased both in the compound and sample number. Our approach couples the timing of the perturbation with the preservation of subcellular heterogeneity, it will serve as a strong foundation for preclinical studies. Citation Format: Mushriq Al-Jazrawe, Csaba Molnar, Elisabeth Abeyta, Steven Blum, Niklas Rindtorff, Kathryn Cebula, Sean Misek, Maria Alimova, William Colgan, Carmen Rios, Moony Tseng, James McFarland, Aviv Regev, Beth Cimini, Anne Carpenter, Adam Bass, Samuel Klempner, Jesse Boehm. Rapid label-free imaging-based evaluation of cancer dependencies in zero-passage primary cells [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1117.