Accurate Extraction Research Articles

Depth extraction of 3D defects on curved surfaces with multi-line lasers

Abstract The adoption of computer vision technology has significantly impacted surface defect inspection by providing a non-contact, cost-effective solution that has been widely accepted. Among the various techniques available, 3D defect inspection using multi-line lasers is notable for its simplicity, high detection speed, and extensive coverage. The accuracy of this method is significantly constrained by the precision of laser stripe extraction. In industrial environments, achieving accurate extraction is hindered by the intricate surface geometries of objects and the challenge of maintaining uniform brightness in multi-line laser stripes. To address these challenges, we propose a novel approach to extract the depth of 3D defects on surfaces using multi-line lasers. Our method combines guided filtering and the Frankle-McCann Retinex (FMR) algorithms to improve the quality of captured images. We have refined the laser stripe extraction process and proposed an advanced adaptive threshold segmentation technique that utilizes the OTSU method to determine threshold coefficients, followed by secondary segmentation based on a neighborhood search. The extracted laser strips are then processed using the quadratic weighted gray gravity method. Additionally, we proposed an innovative region-growth segmentation method based on neighborhood search that effectively segments individual laser strips. We also design a strategy for determining 3D defect depths in situations where precise camera calibration is challenging. The efficacy of our proposed method was rigorously tested on a hot-rolled seamless steel tube with a diameter of 145 mm. The resulting 3D defect depth exhibited an error of less than 0.5 mm, meeting the stringent standards required for practical applications.

P-773 The absence and heterogeneity of pregnancy outcome definitions in randomized controlled trials of infertility: an overview of the literature

Abstract Study question What proportion of infertility trials reported pregnancy and live birth, how often were these outcomes defined, and what is the consistency in their definitions? Summary answer Almost 2/3 of infertility trials did not report live birth. One in four did not define clinical pregnancy. There was considerable heterogeneity in outcome definitions. What is known already The absent or inconsistent outcome definitions in randomized controlled trials (RCTs) complicate the interpretation of the trial results and undermine evidence synthesis. While reporting live births in infertility trials has been a long-running recommendation, the extent to which this is adhered to is unclear. Furthermore, it is unknown if the use of core outcome sets in infertility has improved the standardization of outcome reporting and what is the uptake of its recommendation on reporting live birth. Study design, size, duration We studied all RCTs in reproductive medicine published between 2012 and 2023. We aim to ascertain i) whether clinical pregnancy, ongoing pregnancy and live birth were reported; and ii) whether and how these pregnancy outcomes were defined. Participants/materials, setting, methods We systematically searched Embase, Medline and CENTRAL for RCTs in reproductive medicine from 2012 to 2023. RCTs involving infertile women that reported either biochemical pregnancy, clinical pregnancy, ongoing pregnancy, or live birth were eligible, while conference abstracts, secondary analyses or non-English written RCTs were not. Two authors independently screened articles. We used text-mining to extract pregnancy definitions in R. The accuracy of automatic data extraction was verified with a random sample of 20 articles. Main results and the role of chance After screening 5708 articles, 1,311 were included, of which 21 have been retracted. Manual checks of 20 papers confirmed the accuracy of text-mining. Among 1311 RCTs, 61% of articles reported pregnancy without data on live birth, while only 14% of all RCTs reported on clinical pregnancy, ongoing pregnancy and live birth; 13% of RCTs reported pregnancy without specifying a definition (biochemical, clinical or ongoing). Definitions for clinical pregnancy, ongoing pregnancy and live birth were not provided in 27% (270/1010), 39% (153/389) and 68% (373/549) of articles reporting data on respectively clinical pregnancy, ongoing pregnancy and live birth. Among 740 clinical pregnancy definitions, 46% included gestational age (weeks), ranging from 3 to 16 weeks, typically at 6 to 7 weeks in 40% of definitions. Gestational sac and fetal heartbeat appeared respectively in 64% and 55% of the definitions. Of 236 ongoing pregnancy definitions, 94% included gestational age (weeks), stretching from 6 to 28 weeks, typically at 12 weeks in 52% of definitions. Gestational sac and fetal heartbeat appeared in 9% and 63% of the definitions, respectively. Among 176 live birth definitions, 64% used a gestational age (weeks) threshold, varying from 20 to 34 weeks, typically at week 24 in 30% of definitions. Limitations, reasons for caution While we optimized the text-mining algorithm attempting to identify all types of outcome definitions and manually curated all extracted definitions, there were definitions that we incorrectly omitted. Wider implications of the findings Despite recommendations to report live births in core outcome sets, only a third of RCTs did so, highlighting a disconnection between the advocated outcome and what researchers are reporting. The significant lack and variability of pregnancy definitions underscore the imperative to increase the dissemination and uptake of standardized pregnancy outcomes. Trial registration number Not applicable

Advancements and Challenges in Hand Gesture Recognition: A Comprehensive Review

Hand gesture recognition is a quickly developing field with many uses in human-computer interaction, sign language recognition, virtual reality, gaming, and robotics. This paper reviews different ways to model hands, such as vision-based, sensor-based, and data glove-based techniques. It emphasizes the importance of accurate hand modeling and feature extraction for capturing and analyzing gestures. Key features like motion, depth, color, shape, and pixel values and their relevance in gesture recognition are discussed. Challenges faced in hand gesture recognition include lighting variations, complex backgrounds, noise, and real-time performance. Machine learning algorithms are used to classify and recognize gestures based on extracted features. The paper emphasizes the need for further research and advancements to improve hand gesture recognition systems' robustness, accuracy, and usability. This review offers valuable insights into the current state of hand gesture recognition, its applications, and its potential to revolutionize human-computer interaction and enable natural and intuitive interactions between humans and machines. In simpler terms, hand gesture recognition is a way for computers to understand what people are saying with their hands. It has many potential applications, such as allowing people to control computers without touching them or helping people with disabilities communicate. The paper reviews different ways to develop hand gesture recognition systems and discusses the challenges and opportunities in this area.

Application of machine reading comprehension techniques for named entity recognition in materials science

Materials science is an interdisciplinary field that studies the properties, structures, and behaviors of different materials. A large amount of scientific literature contains rich knowledge in the field of materials science, but manually analyzing these papers to find material-related data is a daunting task. In information processing, named entity recognition (NER) plays a crucial role as it can automatically extract entities in the field of materials science, which have significant value in tasks such as building knowledge graphs. The typically used sequence labeling methods for traditional named entity recognition in material science (MatNER) tasks often fail to fully utilize the semantic information in the dataset and cannot effectively extract nested entities. Herein, we proposed to convert the sequence labeling task into a machine reading comprehension (MRC) task. MRC method effectively can solve the challenge of extracting multiple overlapping entities by transforming it into the form of answering multiple independent questions. Moreover, the MRC framework allows for a more comprehensive understanding of the contextual information and semantic relationships within materials science literature, by integrating prior knowledge from queries. State-of-the-art (SOTA) performance was achieved on the Matscholar, BC4CHEMD, NLMChem, SOFC, and SOFC-Slot datasets, with F1-scores of 89.64%, 94.30%, 85.89%, 85.95%, and 71.73%, respectively in MRC approach. By effectively utilizing semantic information and extracting nested entities, this approach holds great significance for knowledge extraction and data analysis in the field of materials science, and thus accelerating the development of material science.Scientific contributionWe have developed an innovative NER method that enhances the efficiency and accuracy of automatic entity extraction in the field of materials science by transforming the sequence labeling task into a MRC task, this approach provides robust support for constructing knowledge graphs and other data analysis tasks.

BcRflow: a Nextflow pipeline for characterizing B cell receptor repertoires from non-targeted transcriptomic data.

B cells play a critical role in the adaptive recognition of foreign antigens through diverse receptor generation. While targeted immune sequencing methods are commonly used to profile B cell receptors (BCRs), they have limitations in cost and tissue availability. Analyzing B cell receptor profiling from non-targeted transcriptomics data is a promising alternative, but a systematic pipeline integrating tools for accurate immune repertoire extraction is lacking. Here, we present bcRflow, a Nextflow pipeline designed to characterize BCR repertoires from non-targeted transcriptomics data, with functional modules for alignment, processing, and visualization. bcRflow is a comprehensive, reproducible, and scalable pipeline that can run on high-performance computing clusters, cloud-based computing resources like Amazon Web Services (AWS), the Open OnDemand framework, or even local desktops. bcRflow utilizes institutional configurations provided by nf-core to ensure maximum portability and accessibility. To demonstrate the functionality of the bcRflow pipeline, we analyzed a public dataset of bulk transcriptomic samples from COVID-19 patients and healthy controls. We have shown that bcRflow streamlines the analysis of BCR repertoires from non-targeted transcriptomics data, providing valuable insights into the B cell immune response for biological and clinical research. bcRflow is available at https://github.com/Bioinformatics-Core-at-Childrens/bcRflow.

Analysis of the Spatiotemporal Evolution and Driving Mechanisms of Impervious Surfaces along the Jiaozhou Bay (China) Coast over the Past Four Decades

Impervious surfaces serve as critical indicators for monitoring urbanization processes and assessing urban ecological conditions. The precise extraction and analysis of the spatiotemporal variations in impervious surfaces are essential for informing urban planning strategies. The unique location advantage of Jiaozhou Bay makes it an important urban gathering area. Based on Landsat remote sensing image data, the extraction effect and accuracy of urban built-up area index, biophysical index, and random forest classification were compared and analyzed. Then, the optimal random forest method was used to extract impervious water information from 8 Landsat satellite images of the coastal area of Jiaozhou Bay from 1986 to 2022. Over the past four decades, the impervious surface area in the Jiaozhou Bay coastal region has expanded dramatically from 71.53 km2 in 1986 to 1049.16 km2 in 2022, with the most significant increase, nearly doubling, occurring between 2011 and 2017. Spatially, the distribution of impervious surfaces has expanded progressively from coastal to inland areas and from central to peripheral zones, particularly toward the southwest in Huangdao District and Jiaozhou City. The distribution of impervious surfaces in the Jiaozhou Bay coastal area is primarily confined to flat and gently sloping nearshore regions due to natural constraints like terrain slope. Concurrently, policy initiatives, along with population and economic growth, have catalyzed the rapid expansion of these surfaces. These insights are invaluable for comprehending the urban spatiotemporal dynamics and patterns along the Jiaozhou Bay coast and offer fresh perspectives for research into urban transformations and the sustainable development of ecological environments in other coastal regions.

A High-Throughput Method for Accurate Extraction of Intact Rice Panicle Traits.

Rice panicle traits serve as critical indicators of both yield potential and germplasm resource quality. However, traditional manual measurements of these traits, which typically involve threshing, are not only laborious and time-consuming but also prone to introducing measurement errors. This study introduces a high-throughput and nondestructive method, termed extraction of panicle traits (EOPT), along with the software Panicle Analyzer, which is designed to assess unshaped intact rice panicle traits, including the panicle grain number, grain length, grain width, and panicle length. To address the challenge of grain occlusion within an intact panicle, we define a panicle morphology index to quantify the occlusion levels among the rice grains within the panicle. By calibrating the grain number obtained directly from rice panicle images based on the panicle morphology index, we substantially improve the grain number detection accuracy. For measuring grain length and width, the EOPT selects rice grains using an intersection over union threshold of 0.8 and a confidence threshold of 0.7 during the grain detection process. The mean values of these grains were calculated to represent all the panicle grain lengths and widths. In addition, EOPT extracted the main path of the skeleton of the rice panicle using the Astar algorithm to determine panicle lengths. Validation on a dataset of 1,554 panicle images demonstrated the effectiveness of the proposed method, achieving 93.57% accuracy in panicle grain counting with a mean absolute percentage error of 6.62%. High accuracy rates were also recorded for grain length (96.83%) and panicle length (97.13%). Moreover, the utility of EOPT was confirmed across different years and scenes, both indoors and outdoors. A genome-wide association study was conducted, leveraging the phenotypic traits obtained via EOPT and genotypic data. This study identified single-nucleotide polymorphisms associated with grain length, width, number per panicle, and panicle length, further emphasizing the utility and potential of this method in advancing rice breeding.

TGIN: Translation-Based Graph Inference Network for Few-Shot Relational Triplet Extraction.

Extracting relational triplets aims at detecting entity pairs and their semantic relations. Compared with pipeline models, joint models can reduce error propagation and achieve better performance. However, all of these models require large amounts of training data, therefore performing poorly on many long-tail relations in reality with insufficient data. In this article, we propose a novel end-to-end model, called TGIN, for few-shot triplet extraction. The core of TGIN is a multilayer heterogeneous graph with two types of nodes (entity node and relation node) and three types of edges (relation-entity edge, entity-entity edge, and relation-relation edge). On the one hand, this heterogeneous graph with entities and relations as nodes can intuitively extract relational triplets jointly, thereby reducing error propagation. On the other hand, it enables the triplet information of limited labeled data to interact better, thus maximizing the advantage of this information for few-shot triplet extraction. Moreover, we devise a graph aggregation and update method that utilizes translation algebraic operations to mine semantic features while retaining structure features between entities and relations, thereby improving the robustness of the TGIN in a few-shot setting. After updating the node and edge features through layers, TGIN propagates the label information from a few labeled examples to unlabeled examples, thus inferring triplets from these unlabeled examples. Extensive experiments on three reconstructed datasets demonstrate that TGIN can significantly improve the accuracy of triplet extraction by 2.34%~10.74% compared with the state-of-the-art baselines. To the best of our knowledge, we are the first to introduce a heterogeneous graph for few-shot relational triplet extraction.

The SRG/eROSITA All-Sky Survey

Context. Number counts of galaxy clusters across redshift are a powerful cosmological probe if a precise and accurate reconstruction of the underlying mass distribution is performed – a challenge called mass calibration. With the advent of wide and deep photometric surveys, weak gravitational lensing (WL) by clusters has become the method of choice for this measurement. Aims. We measured and validated the WL signature in the shape of galaxies observed in the first three years of the Dark Energy Survey (DES Y3) caused by galaxy clusters and groups selected in the first all-sky survey performed by SRG (Spectrum Roentgen Gamma)/eROSITA (eRASS1). These data were then used to determine the scaling between the X-ray photon count rate of the clusters and their halo mass and redshift. Methods. We empirically determined the degree of cluster member contamination in our background source sample. The individual cluster shear profiles were then analyzed with a Bayesian population model that self-consistently accounts for the lens sample selection and contamination and includes marginalization over a host of instrumental and astrophysical systematics. To quantify the accuracy of the mass extraction of that model, we performed mass measurements on mock cluster catalogs with realistic synthetic shear profiles. This allowed us to establish that hydrodynamical modeling uncertainties at low lens redshifts (z < 0.6) are the dominant systematic limitation. At high lens redshift, the uncertainties of the sources’ photometric redshift calibration dominate. Results. With regard to the X-ray count rate to halo mass relation, we determined its amplitude, its mass trend, the redshift evolution of the mass trend, the deviation from self-similar redshift evolution, and the intrinsic scatter around this relation. Conclusions. The mass calibration analysis performed here sets the stage for a joint analysis with the number counts of eRASS1 clusters to constrain a host of cosmological parameters. We demonstrate that WL mass calibration of galaxy clusters can be performed successfully with source galaxies whose calibration was performed primarily for cosmic shear experiments, opening the way for the cluster cosmological exploitation of future optical and NIR surveys like Euclid and LSST.

A GAN-based fast focusing method for circular SAR images

Circular Synthetic Aperture Radar(CSAR) imaging is vulnerable to perturbations in the atmosphere and various other elements that can lead to position offset errors in the antenna's phase center as well as induce motion errors. Traditional phase compensation methods that operate in the time domain, such as Auto-regressive Back-projection (ARBP), typically require computation on a direction-by-direction basis, which can result in the considerable expenditure of time and memory resources. To address these challenges, thispaperintroduces a novel approach for focusing on CSAR images. This method leverages the training of a Generative Adversarial Network (GAN) to directly achieve focus on CSAR sub-aperture images. Additionally, to counteract the network's tendency towards low-frequency preferences, the Auto-focus Frequency Loss (AFFL) is introduced. Moreover, to enhance the accuracy of focus position extraction, the Focus Position Feature Attention (FPFA) is proposed. These innovations, along with a new fusion strategy for the sub-aperture images post-focusing, have been experimentally validated, demonstrating significant improvements in the efficiency and accuracy of CSAR image focusing.

Online commodity recommendation model for interaction between user ratings and intensity-weighted hierarchical sentiment: A case study of LYCOM

The online commodity recommendation (OCR) model mines users’ historical behavior characteristics and recommends products that may be of interest according to user preferences. Online reviews are among the most important information sources for OCR. However, the explicit and implicit emotional words in online review texts have different structures in the expression of multi-attribute emotions. To fully utilize review information and improve the recommendation accuracy, we propose an OCR model that considers the interaction of multiple attributes and hierarchical emotions and calculates a score weighted by emotion intensity. First, to balance the efficiency and accuracy of information extraction while considering the coexistence of explicit and implicit expressions in online review text, a multi-attribute hierarchical emotion lexicon construction method is proposed. Second, based on the advantage of intuitionistic fuzzy sets in terms of information expression superiority, multi-attribute review text information expression of the affective polarity and intensity of online review text is realized. Then, combined with the weighted singular value decomposition and factorization machine method, we propose an OCR model for interactions between multi-attribute emotions and scores through fusion and recombination of the eigenvectors of users and products. Finally, tourism products on the LYCOM website are used as an example to verify the effectiveness of the proposed method.

A dataset of spatial distribution of tea plantations at 10 m resolution in Fujian Province from 2000 to 2020

Being a major tea producing province in China, Fujian Province is badly in need of quick and accurate spatial distribution of tea plantation for decision-making in both the agricultural economic development and the ecological environment construction for the province. This study retrieved and processed Sentinel-1 (S1) radar data and Sentinel-2 (S2) multispectral data on the GEE cloud platform; and it extracted 98 features, such as spectral features, texture features, and terrain features from the terrain data. The recursive elimination support vector machine algorithm (SVM RFE) is used to screen the characteristic variables, resulting in the creation of 4 feature combination schemes. Using a support vector machine (SVM) classifier to extract the distribution data of tea plantations and assessing the accuracy of the 4 feature combination schemes, we obtained the spatial distribution data of tea plantations at 10 m resolution in Fujian Province in 2020. On this basis, we used the GEE cloud platform to access the vegetation disturbance information in Fujian Province from 2000 to 2020. We finally obtained a dataset of spatial distribution of tea plantations at 10 m resolution in Fujian Province from 2000 to 2020 by excluding the non-tea plantation areas from the images between 2000 and 2015 with the mask generated from the 2020 tea garden extraction results. This dataset has been manually validated using sample points from key tea-producing counties and townships. The results indicate an extraction accuracy of over 92% for tea plantations in 2020. The extraction accuracy of tea plantations obtained using interference data removal method in 2000, 2005, 2010, and 2015 was all above 80%. The dataset with a high accuracy in tea plantation extraction can provide support for relevant departments in tea plantation management.

A dataset of surface water area with 12-day resolution in Lijiang River Basin from 2015 to 2022

Widely distributed in Guangxi, Karst landforms are characterized by heavy rainfall during the flood season, rapid hydrological dynamics, and notable spatiotemporal variations. A water is an important ecological foundation and natural resource, the timely and accurate acquisition of spatiotemporal changes in surface water bodies holds practical significance for the protection and management of water resources, rapid assessment of drought and flood disasters, and the realization of sustainable development across various economic industries. This dataset takes the Lijiang River Basin in Guilin, Guangxi as the research area. Addressing prevalent challenges such as infrequenct surface water monitoring, limited water body extraction methods due to sparse sample information, and low extraction accuracy in complex terrain areas, we combined the data from Sentinel-1 radar and Sentinel-2 in the study. Considering the different imaging principles of radar and optical sensors, we developed a vector classification method based on empirical thresholds for optical and other multi-source remote sensing data. The method can effectively eliminate interference of most ground object shadows, allowing for the construction of dataset of surface water area with 12-day resolution in Lijiang River Basin from 2015 to 2022. The dataset comprises a total of 196 periods of surface water body vector data. According to the results after verifying the spatial distribution and quantitative accuracy of the dataset, the dataset demonstrates its capability to accurately extract inlets and outlets of rivers, lakes and reservoirs, as well as small water bodies in complex terrain areas. The overall accuracy reaches 92.73%,with a Kappa coefficient of 0.85. This dataset can be used to support water resources protection and sustainable management of the ecological environment in the Lijiang River Basin. Additionally, it can provide reliable data for decision-making in areas such as emergency management, disaster prevention, water conservancy development, and economic development.

A Novel Two-Stage Approach for Automatic Extraction and Multi-View Generation of Litchis

Obtaining consistent multi-view images of litchis is crucial for various litchi-related studies, such as data augmentation and 3D reconstruction. This paper proposes a two-stage model that integrates the Mask2Former semantic segmentation network with the Wonder3D multi-view generation network. This integration aims to accurately segment and extract litchis from complex backgrounds and generate consistent multi-view images of previously unseen litchis. In the first stage, the Mask2Former model is utilized to predict litchi masks, enabling the extraction of litchis from complex backgrounds. To further enhance the accuracy of litchi branch extraction, we propose a novel method that combines the predicted masks with morphological operations and the HSV color space. This approach ensures accurate extraction of litchi branches even when the semantic segmentation model’s prediction accuracy is not high. In the second stage, the segmented and extracted litchi images are passed as input into the Wonder3D network to generate multi-view of the litchis. After comparing different semantic segmentation and multi-view synthesis networks, the Mask2Former and Wonder3D networks demonstrated the best performance. The Mask2Former network achieved a mean Intersection over Union (mIoU) of 79.79% and a mean pixel accuracy (mPA) of 85.82%. The Wonder3D network achieved a peak signal-to-noise ratio (PSNR) of 18.89 dB, a structural similarity index (SSIM) of 0.8199, and a learned perceptual image patch similarity (LPIPS) of 0.114. Combining the Mask2Former model with the Wonder3D network resulted in an increase in PSNR and SSIM scores by 0.21 dB and 0.0121, respectively, and a decrease in LPIPS by 0.064 compared to using the Wonder3D model alone. Therefore, the proposed two-stage model effectively achieves automatic extraction and multi-view generation of litchis with high accuracy.

EUNet: Edge-UNet for Accurate Building Extraction and Edge Emphasis in Gaofen-7 Images

Deep learning is currently the mainstream approach for building extraction tasks in remote-sensing imagery, capable of automatically learning features of buildings in imagery and yielding satisfactory extraction results. However, due to the diverse sizes, irregular layouts, and complex spatial relationships of buildings, extracted buildings often suffer from incompleteness and boundary issues. Gaofen-7 (GF-7), as a high-resolution stereo mapping satellite, provides well-rectified images from its rear-view imagery, which helps mitigate occlusions in highly varied terrain, thereby offering rich information for building extraction. To improve the integrity of the edges of the building extraction results, this paper proposes a dual-task network (Edge-UNet, EUnet) based on UNet, incorporating an edge extraction branch to emphasize edge information while predicting building targets. We evaluate this method using a self-made GF-7 Building Dataset, the Wuhan University (WHU) Building Dataset, and the Massachusetts Buildings Dataset. Comparative analysis with other mainstream semantic segmentation networks reveals significantly higher F1 scores for the extraction results of our method. Our method exhibits superior completeness and accuracy in building edge extraction compared to unmodified algorithms, demonstrating robust performance.

ResTrans‐Unet: A Residual‐Aware Transformer‐Based Approach to Medical Image Segmentation

ABSTRACTThe convolutional neural network has significantly enhanced the efficacy of medical image segmentation. However, challenges persist in the deep learning‐based method for medical image segmentation, necessitating the resolution of the following issues: (1) Medical images, characterized by a vast spatial scale and complex structure, pose difficulties in accurate edge information extraction; (2) In the decoding process, the assumption of equal importance among different channels contradicts the reality of their varying significance. This study addresses challenges observed in earlier medical image segmentation networks, particularly focusing on the precise extraction of edge information and the inadequate consideration of inter‐channel importance during decoding. To address these challenges, we introduce ResTrans‐Unet (residual transformer medical image segmentation network), an automatic segmentation model based on Residual‐aware transformer. The Transformer is enhanced through the incorporation of ResMLP, resulting in enhanced edge information capture in images and improved network convergence speed. Additionally, Squeeze‐and‐Excitation Networks, which emphasize channel relationships, are integrated into the decoder to precisely highlight important features and suppress irrelevant ones. Experimental validations on two public datasets were carried out to assess the proposed model, comparing its performance with that of advanced models. The experimental results unequivocally demonstrate the superior performance of ResTrans‐Unet in medical image segmentation tasks.

Outcome Prediction Using Multi-Modal Information: Integrating Large Language Model-Extracted Clinical Information and Image Analysis.

Survival prediction post-cystectomy is essential for the follow-up care of bladder cancer patients. This study aimed to evaluate artificial intelligence (AI)-large language models (LLMs) for extracting clinical information and improving image analysis, with an initial application involving predicting five-year survival rates of patients after radical cystectomy for bladder cancer. Data were retrospectively collected from medical records and CT urograms (CTUs) of bladder cancer patients between 2001 and 2020. Of 781 patients, 163 underwent chemotherapy, had pre- and post-chemotherapy CTUs, underwent radical cystectomy, and had an available post-surgery five-year survival follow-up. Five AI-LLMs (Dolly-v2, Vicuna-13b, Llama-2.0-13b, GPT-3.5, and GPT-4.0) were used to extract clinical descriptors from each patient's medical records. As a reference standard, clinical descriptors were also extracted manually. Radiomics and deep learning descriptors were extracted from CTU images. The developed multi-modal predictive model, CRD, was based on the clinical (C), radiomics (R), and deep learning (D) descriptors. The LLM retrieval accuracy was assessed. The performances of the survival predictive models were evaluated using AUC and Kaplan-Meier analysis. For the 163 patients (mean age 64 ± 9 years; M:F 131:32), the LLMs achieved extraction accuracies of 74%~87% (Dolly), 76%~83% (Vicuna), 82%~93% (Llama), 85%~91% (GPT-3.5), and 94%~97% (GPT-4.0). For a test dataset of 64 patients, the CRD model achieved AUCs of 0.89 ± 0.04 (manually extracted information), 0.87 ± 0.05 (Dolly), 0.83 ± 0.06~0.84 ± 0.05 (Vicuna), 0.81 ± 0.06~0.86 ± 0.05 (Llama), 0.85 ± 0.05~0.88 ± 0.05 (GPT-3.5), and 0.87 ± 0.05~0.88 ± 0.05 (GPT-4.0). This study demonstrates the use of LLM model-extracted clinical information, in conjunction with imaging analysis, to improve the prediction of clinical outcomes, with bladder cancer as an initial example.

Multi-Scale Earthquake Damaged Building Feature Set

Earthquake disasters are marked by their unpredictability and potential for extreme destructiveness. Accurate information on building damage, captured in post-earthquake remote sensing images, is critical for an effective post-disaster emergency response. The foundational features within these images are essential for the accurate extraction of building damage data following seismic events. Presently, the availability of publicly accessible datasets tailored specifically to earthquake-damaged buildings is limited, and existing collections of post-earthquake building damage characteristics are insufficient. To address this gap and foster research advancement in this domain, this paper introduces a new, large-scale, publicly available dataset named the Major Earthquake Damage Building Feature Set (MEDBFS). This dataset comprises image data sourced from five significant global earthquakes and captured by various optical remote sensing satellites, featuring diverse scale characteristics and multiple spatial resolutions. It includes over 7000 images of buildings pre- and post-disaster, each subjected to stringent quality control and expert validation. The images are categorized into three primary groups: intact/slightly damaged, severely damaged, and completely collapsed. This paper develops a comprehensive feature set encompassing five dimensions: spectral, texture, edge detection, building index, and temporal sequencing, resulting in 16 distinct classes of feature images. This dataset is poised to significantly enhance the capabilities for data-driven identification and analysis of earthquake-induced building damage, thereby supporting the advancement of scientific and technological efforts for emergency earthquake response.

Dominant Color Detection For Online Fashion Retrievals

This paper introduces a novel approach aimed at efficiently extracting dominant colors from online fashion images. The method addresses challenges related to detecting overlapping objects and computationally expensive methods by combining K-means clustering and graph-cut techniques into a framework. This framework incorporates an adaptive weighting strategy to enhance color extraction accuracy. Additionally, it introduces a two-phase fashion apparel detection method called YOLOv4, which utilizes U-Net architecture for clothing segmentation to precisely separate clothing items from the background or other elements. Experimental results show that K-means with YOLOv4 outperforms K-means with the U-Net model. These findings suggest that the U-Net architecture and YOLOv4 models can be effective methods for complex image segmentation tasks in online fashion retrieval and image processing, particularly in the rapidly evolving e-commerce environment.

Synchronous Measurements of Extracellular Action Potentials and Neurochemical Activity with Carbon Fiber Electrodes in Nonhuman Primates.

Measuring the dynamic relationship between neuromodulators, such as dopamine, and neuronal action potentials is imperative to understand how these fundamental modes of neural signaling interact to mediate behavior. We developed methods to measure concurrently dopamine and extracellular action potentials (i.e., spikes) in monkeys. Standard fast-scan cyclic voltammetric (FSCV) electrochemical (EChem) and electrophysiological (EPhys) recording systems are combined and used to collect spike and dopamine signals, respectively, from an array of carbon fiber (CF) sensors implanted in the monkey striatum. FSCV requires the application of small voltages at the implanted sensors to measure redox currents generated from target molecules, such as dopamine. These applied voltages create artifacts at neighboring EPhys measurement sensors which may lead to misclassification of these signals as physiological spikes. Therefore, simple automated temporal interpolation algorithms were designed to remove these artifacts and enable accurate spike extraction. We validated these methods using simulated artifacts and demonstrated an average spike recovery rate of 84.5%. We identified and discriminated cell type-specific units in the monkey striatum that were shown to correlate to specific behavioral task parameters related to reward size and eye movement direction. Synchronously recorded spike and dopamine signals displayed contrasting relations to the task variables, suggesting a complex relationship between these two modes of neural signaling. Future application of our methods will help advance our understanding of the interactions between neuromodulator signaling and neuronal activity, to elucidate more detailed mechanisms of neural circuitry and plasticity mediating behaviors in health and in disease.