Input Data Research Articles

Evolution of publicly available large language models for complex decision-making in breast cancer care.

This study investigated the concordance of five different publicly available Large Language Models (LLM) with the recommendations of a multidisciplinary tumor board regarding treatment recommendations for complex breast cancer patient profiles. Five LLM, including three versions of ChatGPT (version 4 and 3.5, with data access until September 3021 and January 2022), Llama2, and Bard were prompted to produce treatment recommendations for 20 complex breast cancer patient profiles. LLM recommendations were compared to the recommendations of a multidisciplinary tumor board (gold standard), including surgical, endocrine and systemic treatment, radiotherapy, and genetic testing therapy options. GPT4 demonstrated the highest concordance (70.6%) for invasive breast cancer patient profiles, followed by GPT3.5 September 2021 (58.8%), GPT3.5 January 2022 (41.2%), Llama2 (35.3%) and Bard (23.5%). Including precancerous lesions of ductal carcinoma in situ, the identical ranking was reached with lower overall concordance for each LLM (GPT4 60.0%, GPT3.5 September 2021 50.0%, GPT3.5 January 2022 35.0%, Llama2 30.0%, Bard 20.0%). GPT4 achieved full concordance (100%) for radiotherapy. Lowest alignment was reached in recommending genetic testing, demonstrating a varying concordance (55.0% for GPT3.5 January 2022, Llama2 and Bard up to 85.0% for GPT4). This early feasibility study is the first to compare different LLM in breast cancer care with regard to changes in accuracy over time, i.e., with access to more data or through technological upgrades. Methodological advancement, i.e., the optimization of prompting techniques, and technological development, i.e., enabling data input control and secure data processing, are necessary in the preparation of large-scale and multicenter studies to provide evidence on their safe and reliable clinical application. At present, safe and evidenced use of LLM in clinical breast cancer care is not yet feasible.

A Survey of Adversarial Attacks: An Open Issue for Deep Learning Sentiment Analysis Models

In recent years, the use of deep learning models for deploying sentiment analysis systems has become a widespread topic due to their processing capacity and superior results on large volumes of information. However, after several years’ research, previous works have demonstrated that deep learning models are vulnerable to strategically modified inputs called adversarial examples. Adversarial examples are generated by performing perturbations on data input that are imperceptible to humans but that can fool deep learning models’ understanding of the inputs and lead to false predictions being generated. In this work, we collect, select, summarize, discuss, and comprehensively analyze research works to generate textual adversarial examples. There are already a number of reviews in the existing literature concerning attacks on deep learning models for text applications; in contrast to previous works, however, we review works mainly oriented to sentiment analysis tasks. Further, we cover the related information concerning generation of adversarial examples to make this work self-contained. Finally, we draw on the reviewed literature to discuss adversarial example design in the context of sentiment analysis tasks.

Meta learning based residual network for industrial production quality prediction with limited data

Due to the challenge of collecting a substantial amount of production-quality data in real-world industrial settings, the implementation of production quality prediction models based on deep learning is not effective. To achieve the goal of predicting production quality with limited data and address the issue of model degradation in the training process of deep learning networks, we propose Meta-Learning based on Residual Network (MLRN) models for production quality prediction with limited data. Firstly, the MLRN model is trained on a variety of learning tasks to acquire knowledge for predicting production quality. Furthermore, to obtain more features with limited data and avoid the issues of gradient disappearing or exploding in deep network training, the enhanced residual network with the effective channel attention (ECA) mechanism is chosen as the basic network structure of MLRN. Additionally, a multi-batch and multi-task data input approach is implemented to prevent overfitting. Finally, the availability of the MLRN model is demonstrated by comparing it with other models using both numerical and graphical datasets.

Tuberculosis Infection Control: Experiences and Considerations from a Web Based Tool Implementation.

Tuberculosis (TB) remains a significant global health challenge. Indeed, according to the World Health Organization (WHO), TB is classified as the second most common cause of death worldwide due to a single infectious agent in 2022, following COVID-19. To effectively manage tuberculosis patients, it is necessary to ensure accurate diagnosis, prompt treatment initiation, and vigilant monitoring of patients' progress. In 2017, the TB Ge network was implemented and launched in two primary hospitals within the Liguria Region in Italy, with the main purpose to manage tuberculosis infections. This system, organized as a web-based tool, simplifies the manual input of patient's data and therapies, while automating the integration of test results from hospitals' Laboratory Information Systems (LIS), without requiring human intervention. The goal of this paper is to highlight the outcomes achieved through the implementation of the TB Ge network in a period seriously affected by the COVID-19 pandemia and outline future directions. More specifically, the aim is to extend its adoption to all hospitals in the Liguria Region, thus improving the management of tuberculosis infections across healthcare facilities.

The innovation of AI-based software in oral diseases: clinical-histopathological correlation diagnostic accuracy primary study

BackgroundMachine learning (ML) through artificial intelligence (AI) could provide clinicians and oral pathologists to advance diagnostic problems in the field of potentially malignant lesions, oral cancer, periodontal diseases, salivary gland disease, oral infections, immune-mediated disease, and others. AI can detect micro-features beyond human eyes and provide solution in critical diagnostic cases.ObjectiveThe objective of this study was developing a software with all needed feeding data to act as AI-based program to diagnose oral diseases. So our research question was: Can we develop a Computer-Aided Software for accurate diagnosis of oral diseases based on clinical and histopathological data inputs?MethodThe study sample included clinical images, patient symptoms, radiographic images, histopathological images and texts for the oral diseases of interest in the current study (premalignant lesions, oral cancer, salivary gland neoplasms, immune mediated oral mucosal lesions, oral reactive lesions) total oral diseases enrolled in this study was 28 diseases retrieved from the archives of oral maxillofacial pathology department. Total 11,200 texts and 3000 images (2800 images were used for training data to the program and 100 images were used as test data to the program and 100 cases for calculating accuracy, sensitivity& specificity).ResultsThe correct diagnosis rates for group 1 (software users), group 2 (microscopic users) and group 3 (hybrid) were 87%, 90.6, 95% respectively. The reliability for inter-observer value was done by calculating Cronbach’s alpha and interclass correlation coefficient. The test revealed for group 1, 2 and 3 the following values respectively 0.934, 0.712 & 0.703. All groups showed acceptable reliability especially for Diagnosis Oral Diseases Software (DODS) that revealed higher reliability value than other groups. However, The accuracy, sensitivity & specificity of this software was lower than those of oral pathologists (master’s degree).ConclusionThe correct diagnosis rate of DODS was comparable to oral pathologists using standard microscopic examination. The DODS program could be utilized as diagnostic guidance tool with high reliability & accuracy.

Research on Input Schemes for Polarimetric SAR Classification Using Deep Learning

This study employs the reflection symmetry decomposition (RSD) method to extract polarization scattering features from ground object images, aiming to determine the optimal data input scheme for deep learning networks in polarimetric synthetic aperture radar classification. Eight distinct polarizing feature combinations were designed, and the classification accuracy of various approaches was evaluated using the classic convolutional neural networks (CNNs) AlexNet and VGG16. The findings reveal that the commonly employed six-parameter input scheme, favored by many researchers, lacks the comprehensive utilization of polarization information and warrants attention. Intriguingly, leveraging the complete nine-parameter input scheme based on the polarization coherence matrix results in improved classification accuracy. Furthermore, the input scheme incorporating all 21 parameters from the RSD and polarization coherence matrix notably enhances overall accuracy and the Kappa coefficient compared to the other seven schemes. This comprehensive approach maximizes the utilization of polarization scattering information from ground objects, emerging as the most effective CNN input data scheme in this study. Additionally, the classification performance using the second and third component total power values (P2 and P3) from the RSD surpasses the approach utilizing surface scattering power value (PS) and secondary scattering power value (PD) from the same decomposition.

Machine Learning-Based Hand Pose Generation Using a Haptic Controller

In this study, we present a novel approach to derive hand poses from data input via a haptic controller, leveraging machine learning techniques. The input values received from the haptic controller correspond to the movement of five fingers, each assigned a value between 0.0 and 1.0 based on the applied pressure. The wide array of possible finger movements requires a substantial amount of motion capture data, making manual data integration difficult. This challenge is primary due to the need to process and incorporate large volumes of diverse movement information. To tackle this challenge, our proposed method automates the process by utilizing machine learning algorithms to convert haptic controller inputs into hand poses. This involves training a machine learning model using supervised learning, where hand poses are matched with their corresponding input values, and subsequently utilizing this trained model to generate hand poses in response to user input. In our experiments, we assessed the accuracy of the generated hand poses by analyzing the angles and positions of finger joints. As the quantity of training data increased, the margin of error decreased, resulting in generated poses that closely emulated real-world hand movements.

Management of textile waste in Europe: An environmental and a socio-economic assessment of current and future scenarios

Textile production and consumption continues to grow, and yet textile waste management in Europe remains suboptimal. This study assesses and compares the life cycle impacts of managing selected textile waste fractions via selected technologies, under current and future framework conditions. The study combines environmental and socio-economic analyses with data input from stakeholder consultation within the European Union. Preparing for reuse is the most beneficial management pathway from both an environmental and a socio-economic perspective. Overall, mechanical recycling is the second-most preferred pathway for most environmental impact categories. However, for recycling pathways to be economically competitive against incineration, either CO2 prices will have to increase, or technology and processing costs will have to decrease. The uncertainty analysis identified critical parameters, such as substitution factors and yields of most desirable products from the recycling processes. The results offer a detailed basis for prioritising technology pathways for future textile waste management in Europe.

Unraveling the complexity: A taxonomy for characterizing and structuring smart energy services in the building sector

Achieving climate goals requires restructuring the current energy systems, which, among other things, involves decentralizing energy systems. This entails new challenges, e.g., a high degree of coordination and information exchange. Especially the building sector, being the largest energy consumer, offers great improvement potential to reduce emissions. Smart energy services promise various benefits in overcoming these challenges, e.g., through optimized energy generation, demand control, and more efficient energy control systems. Although smart energy services are crucial to transitioning to a more sustainable, reliable, and intelligent energy (eco)system, research lacks a holistic perspective of their essential characteristics and configuration options due to their novelty and heterogeneity. Nevertheless, this perspective is crucial as service providers, service innovators (e.g., business development and start-ups), and researchers face problems developing and deploying value-adding and sustaining smart energy services. Consequently, this work identifies the essential characteristics of building-related smart energy services by reviewing scientific and practitioner literature. It structures them by developing a taxonomy and organizing them in 15 dimensions and 54 characteristics, along smart energy service structure, value creation, delivery, and energy-related data input. The taxonomy and the derived theoretical and managerial implications support researchers and practitioners in designing and conceptualizing future smart energy service innovations.

An intelligent airflow perception model for metal mines based on CNN-LSTM architecture

In view of the harsh underground working environment and difficulties of airflow monitoring sensors placement, implementing an intelligent ventilation strategy is crucial for ensuring ventilation safety during mining process. The key issue lies in obtaining global real-time airflow parameters for ventilation safety and intelligent control. Thus, we propose an intelligent perception approach based on artificial intelligence (AI) method for acquiring airflow parameters, which operates by leveraging partial airflow data from specific monitoring points to predict airflow at undisclosed locations. Initially, this approach is facilitated by the training of an AI database through numerical simulations of ventilation networks. Subsequently, an intelligent airflow perception model is constructed, incorporating convolution neural network (CNN), long short-term memory (LSTM), and hybrid CNN-LSTM architectures. Through iterative updates and enhancements, these models demonstrate average deviations between predicted and actual airflow parameters of less than 5% in both simulated scenarios and empirical applications. Furthermore, in the case study, the CNN-LSTM architecture model exhibits superior performance for intelligent airflow perception. This architecture combing with airflow monitoring system, and utilizing partial real-time data inputs to obtain perception point outputs, can effectively enhance employee productivity, reduce energy consumption, and prevent resource wastage.

Irrigation with Artificial Intelligence: Problems, Premises, Promises

AbstractProtagonists allege that artificial intelligence (AI) is revolutionising contemporaneous mindscapes. Here, we authoritatively review the status quo of AI and machine learning application in irrigated agriculture, evaluating the potential of, and challenges associated with, a wide range of existential AI approaches. We contend that aspiring developers of AI irrigation systems may benefit from human-centred AI, a nascent algorithm that captures diverse end-user views, behaviours and actions, potentially facilitating refinement of proposed systems through iterative stakeholder feedback. AI-guided human–machine collaboration can streamline integration of user needs, allowing customisation towards situational farm management adaptation. Presentation of big data in intuitive, legible and actionable forms for specialists and laypeople also urgently requires attention: here, AI-explainable interpretability may help harness human expertise, enabling end-users to contribute their experience within an AI pipeline for bespoke outputs. Transfer learning holds promise in contextualising place-based AI to agroecological regions, production systems or enterprise mixes, even with limited data inputs. We find that the rate of AI scientific and software development in recent times has outpaced the evolution of adequate legal and institutional regulations, and often social, moral and ethical license to operate, revealing consumer issues associated with data ownership, legitimacy and trust. We opine that AI has great potential to elicit sustainable outcomes in food security, social innovation and environmental stewardship, albeit such potential is more likely to be realised through concurrent development of appropriate ethical, moral and legal dimensions.

Deriving a point of departure for assessing the skin sensitization risk of wearable device constituents with in vitro methods

Wearable devices are in contact with the skin for extended periods. As such, the device constituents should be evaluated for their skin sensitization potential, and a Point of Departure (PoD) should be derived to conduct a proper risk assessment. Without historical in vivo data, the PoD must be derived with New Approach Methods (NAMs). To accomplish this, regression models trained on LLNA data that use data inputs from OECD-validated in vitro tests were used to derive a predicted EC3 value, the LLNA value used to classify skin sensitization potency, for three adhesive monomers (Isobornyl acrylate (IBOA), N, N- Dimethylacrylamide (NNDMA), and Acryloylmorpholine (ACMO) and one dye (Solvent Orange 60 (SO60)). These chemicals can be used as constituents of wearable devices and have been associated with causing allergic contact dermatitis (ACD). Using kinetic DPRA and KeratinoSens™ data, the PoDs obtained with the regression model were 180, 215, 1535, and 8325 μg/cm2 for IBOA, SO60, ACMO, and NNDMA, respectively. The PoDs derived with the regression model using NAMs data will enable a proper skin sensitization risk assessment without using animals.

Investigating a hybrid extreme learning machine coupled with Dingo Optimization Algorithm for modeling liquefaction triggering in sand-silt mixtures

Liquefaction is a devastating consequence of earthquakes that occurs in loose, saturated soil deposits, resulting in catastrophic ground failure. Accurate prediction of such geotechnical parameter is crucial for mitigating hazards, assessing risks, and advancing geotechnical engineering. This study introduces a novel predictive model that combines Extreme Learning Machine (ELM) with Dingo Optimization Algorithm (DOA) to estimate strain energy-based liquefaction resistance. The hybrid model (ELM-DOA) is compared with the classical ELM, Adaptive Neuro-Fuzzy Inference System with Fuzzy C-Means (ANFIS-FCM model), and Sub-clustering (ANFIS-Sub model). Also, two data pre-processing scenarios are employed, namely traditional linear and non-linear normalization. The results demonstrate that non-linear normalization significantly enhances the prediction performance of all models by approximately 25% compared to linear normalization. Furthermore, the ELM-DOA model achieves the most accurate predictions, exhibiting the lowest root mean square error (484.286 J/m3), mean absolute percentage error (24.900%), mean absolute error (404.416 J/m3), and the highest correlation of determination (0.935). Additionally, a Graphical User Interface (GUI) has been developed, specifically tailored for the ELM-DOA model, to assist engineers and researchers in maximizing the utilization of this predictive model. The GUI provides a user-friendly platform for easy input of data and accessing the model's predictions, enhancing its practical applicability. Overall, the results strongly support the proposed hybrid model with GUI serving as an effective tool for assessing soil liquefaction resistance in geotechnical engineering, aiding in predicting and mitigating liquefaction hazards.

FireXplainNet: Optimizing Convolution Block Architecture for Enhanced Wildfire Detection and Interpretability

The early detection of wildfires is a crucial challenge in environmental monitoring, pivotal for effective disaster management and ecological conservation. Traditional detection methods often fail to detect fires accurately and in a timely manner, resulting in significant adverse consequences. This paper presents FireXplainNet, a Convolutional Neural Network (CNN) base model, designed specifically to address these limitations through enhanced efficiency and precision in wildfire detection. We optimized data input via specialized preprocessing techniques, significantly improving detection accuracy on both the Wildfire Image and FLAME datasets. A distinctive feature of our approach is the integration of Local Interpretable Model-agnostic Explanations (LIME), which facilitates a deeper understanding of and trust in the model’s predictive capabilities. Additionally, we have delved into optimizing pretrained models through transfer learning, enriching our analysis and offering insights into the comparative effectiveness of FireXplainNet. The model achieved an accuracy of 87.32% on the FLAME dataset and 98.70% on the Wildfire Image dataset, with inference times of 0.221 and 0.168 milliseconds, respectively. These performance metrics are critical for the application of real-time fire detection systems, underscoring the potential of FireXplainNet in environmental monitoring and disaster management strategies.

Gaushala Safety – Automation of Cattle Livestock

Automation of Cattle Livestock is a ground-breaking initiative aimed at transforming cattle care and management. In a nation where cattle play a pivotal role in rural economies, the absence of accessible veterinary care, delayed vaccinations, and disease outbreaks pose significant threats to cattle health and the livelihoods of dependent communities. This project offers a comprehensive solution by harnessing cutting-edge technology. It includes a robust cattle information management system, utilizing speech recognition for data input to accommodate diverse user literacy levels..The core feature is an intelligent vaccination scheduler, creating personalized vaccination plans based on individual cattle profiles, and sending timely reminders to owners. Additionally, it employs advanced image processing to detect early signs of disease. The project ensures accessibility through voice-enabled interaction and simplifies the registration of new cattle and along with that it includes the cattle tracker and shed cleaning. By enhancing cattle health, increasing accessibility, and promoting sustainable cattle rearing, Automation of Cattle Livestock seeks to empower farmers, protect livestock, and drive rural community development. This project represents a pioneering step toward a brighter future for both cattle and their caretakers in India.

A multi-source change detection algorithm supporting user customization and near real-time deforestation detections

The abundance of free and accessible satellite data has revolutionized our ability to study deforestation with remote sensing. Recent advances have enabled us to monitor deforestation in near real-time (NRT), and a number of operational NRT alert systems using both optical and synthetic aperture radar (SAR) data have been developed. Yet despite their success, there are three primary issues with current systems. First, alerts often require multiple observations to confirm deforestation, which does not help enforcement if deforestation events are short-lived. Second, the methods can be computationally intensive and often focus on presenting binary detection alerts. Third, results are often reported as static accuracies or temporal lags, with relatively little discussion about how either could be improved for different use-cases. We therefore propose a novel, multisource NRT monitoring algorithm that addresses these issues by quickly achieving high-accuracy detections, using simple data harmonization to combine data sources while yielding probabilities of deforestation, and can be customized to meet management goals where users can explicitly change the algorithm to prioritize latency or accuracy in their detections based on their chosen data inputs. To demonstrate the algorithm's potential, we used combined Landsat-8, Sentinel-2, and Sentinel-1 SAR data in northern Myanmar to monitor areas with contrasting forest types. When testing all possible parameterizations, we achieved comparable results to previous studies, with median false positive rates and false negative rates under 10% and 25%, respectively, and median F1 scores consistently above 75% across the sensor combinations. Median post-data acquisition detection times were faster than other studies and ranged from 2 to 8 days, though the majority were between 2 and 5 days. In attempting to optimize the algorithm for our demonstrated region, we discovered there was no optimal parameterization that simultaneously provided the best accuracy and detection. In addition, we found that including Sentinel-1 data did not improve latency or accuracy, most likely due to relatively high levels of noise in our monsoonal study system that experienced high seasonal precipitation variation. We finally chose an example parameterization to create daily landscape maps of deforestation at 10 m resolution in two regions within the study area (∼60,000 ha). We assert this new, customizable method with the quantification of accuracy and latency trade-offs represents a clear advance for NRT change detection and can improve current operational systems.

DEPLOYMENT OF ENHANCED DEEP LEARNING MODEL WITH THE BEST ESTIMATORS ON OPTIMIZERS AND ACTIVATION FUNCTIONS FOR HEALTHCARE IN WEB APPLICATION

Breast cancer and brain tumor stand as leading global causes of mortality. Brain tumor uses Magnetic Resonance Imaging (MRI) which offers superior clarity in visualizing brain structures compared to other imaging modalities, while Breast cancer uses ultrasonography (US) serves as a common tool for detecting breast cancer despite its inherent limitations in image quality. Motion artifacts frequently hinder MRI scans, necessitating skilled radiologists for accurate interpretation. Computer-aided diagnosis (CAD) systems driven by artificial intelligence, present a promising solution by consistently assisting radiologists in analyzing US images. Convolutional neural networks (CNNs) leverage various optimizers like Adam and Stochastic Gradient Descent (SGD), RMSprop, Adagrad, and Adadelta as well as activation functions including PReLu, LeakyReLu, Elu, and ReLu for their construction and training. The comparative analysis highlights the importance of optimizers and activation functions in deep learning algorithms for predicting brain tumors and breast cancer. The Adam optimizer combined with the ReLU activation function achieved an accuracy of 85% for breast cancer prediction, while RMSprop combined with ReLU activation function achieved a higher accuracy of 93% for brain tumor classification. From this research, considerable deep learning configurations are identified for both breast cancer and brain tumor prediction, facilitating more precise and efficient diagnoses. The comparative analysis provides valuable insights for those involved in medical imaging applications. Furthermore, The CNN model is deployed in web interface using flask framework to streamline the integration of these models into healthcare systems. This interface simplifies the input of medical data including image data and provides real-time predictions. KEYWORDS—Convolutional neural networks, Comparative Analysis, Activation functions, Optimizers, Ultrasound images, MRI images, Breast cancer, Brain tumor, medical imaging applications, FLASK.

AI-Driven Dim-Light Adaptive Camera (DimCam) for Lunar Robots

Abstract. The past decade has been a boom in lunar exploration. China, India, Japan and other countries have successfully landed landers or rovers on the lunar surface (Wu et al., 2014, 2018, 2020; Prasad et al., 2023). Future missions to explore the Moon are focusing on the lunar south pole (Peña-Asensio et al., 2024). The solar altitude angle at the lunar south pole is extremely low, resulting in low solar irradiance and large areas often in dim light or shadows. The permanently shadowed regions (PSRs) at the lunar south pole are also likely containing substantial amounts of water ice (Li et al., 2018). Future lunar robots exploring the lunar south pole will need to operate in low light or shadowed regions, making camera sensors sensitive to the dim-light environments necessary for these robots. Common night vision sensors usually use near-infrared cameras. However, sensors based on passive infrared technology have image resolution limited by several factors, including the intensity of infrared radiation emitted by the object, the sensitivity of the camera, and the performance of the optical system. For instance, thermal imagers typically have a resolution of 388*284 pixels only. We have developed a dim-light adaptive camera (DimCam) that is ultra-sensitive to the varying illumination conditions driven by AI to achieve high-definition imaging of 1080P or above, for future lunar robots operating in shadows or dim-light regions. The DimCam integrates two starlight-level ultra-sensitive imaging sensors connected by a rigid base to provide stereo vision in low illumination environment. An AI edge computing unit is embedded inside the DimCam to adaptively denoise and enhance image quality. The AI module uses an end-to-end image denoising network to identify and remove noises in the images more accurately by utilizing depth information from the stereo sensors. Compared with traditional monocular denoising algorithms, the denoising network based on stereo vision can significantly improve denoising effects and efficiency by enhancing the signal-to-noise ratio of the data input in the front end. The superposition of overlapping scenes can be regarded as a delayed exposure. Concurrently, the residual analysis of the aligned images aids in noise identification. In addition, for pixels obscured by noise, more accurate pixel values can be restored through interpolation or replacement using depth information obtained from the stereo sensors. Subsequently, a pre-trained lightweight deep network modified from Zero-DCE (Guo et al., 2020) is used for image quality enhancement in terms of brightness and contrast, providing high-quality images even in low-light environments for subsequent applications, such as positioning and navigation of robots, 3D mapping of the surrounding environment, and autonomous driven. We have tested the DimCam in a simulated environment in the laboratory, and the results show that the DimCam has promising performances and great potential for various applications.

A Deep Learning Classification Scheme for PolSAR Image Based on Polarimetric Features

Polarimetric features extracted from polarimetric synthetic aperture radar (PolSAR) images contain abundant back-scattering information about objects. Utilizing this information for PolSAR image classification can improve accuracy and enhance object monitoring. In this paper, a deep learning classification method based on polarimetric channel power features for PolSAR is proposed. The distinctive characteristic of this method is that the polarimetric features input into the deep learning network are the power values of polarimetric channels and contain complete polarimetric information. The other two input data schemes are designed to compare the proposed method. The neural network can utilize the extracted polarimetric features to classify images, and the classification accuracy analysis is employed to compare the strengths and weaknesses of the power-based scheme. It is worth mentioning that the polarized characteristics of the data input scheme mentioned in this article have been derived through rigorous mathematical deduction, and each polarimetric feature has a clear physical meaning. By testing different data input schemes on the Gaofen-3 (GF-3) PolSAR image, the experimental results show that the method proposed in this article outperforms existing methods and can improve the accuracy of classification to a certain extent, validating the effectiveness of this method in large-scale area classification.

Predicting Progression of Intracranial Hemorrhage in the Prehospital TXA for TBI Trial.

Progression of intracranial hemorrhage is a common, potentially devastating complication after moderate/severe traumatic brain injury (TBI). Clinicians have few tools to predict which patients with traumatic intracranial hemorrhage on their initial head computed tomography (hCT) scan will progress. The objective of this investigation was to identify clinical, imaging, and/or protein biomarkers associated with progression of intracranial hemorrhage (PICH) after moderate/severe TBI and to create an accurate predictive model of PICH based on clinical features available at presentation. We analyzed a subset of subjects from the phase II double-blind, multi-center, randomized "Prehospital Tranexamic Acid Use for TBI" trial. This subset was limited to the placebo arm of the parent trial with evidence of hemorrhage on the initial hCT and a follow-up hCT 6 h after. PICH was defined as an increase in hemorrhage size by 30% or more, or the development of new hemorrhage in the intra- and extra-axial intracranial vault between the initial and the follow-up hCT. Two independent radiologists evaluated each hCT, and conflicts were adjudicated by a third. Clinical and radiographic characteristics were collected, along with plasma protein biomarkers at admission. Principal component analysis (PCA) was performed, and each principal component (PC) was interrogated for its association with PICH. Finally, expert opinion and recursive feature extraction (RFE) were used to select input features for the construction of several supervised classification models. Their ability to predict PICH was quantified and compared. In this subset of subjects (n = 104), 46% (n = 48) demonstrated PICH. Univariate analyses showed no association between PICH and age, sex, admission Glasgow Coma Scale (GCS), GCS motor subscore, presence of midline shift, admission platelet count or admission INR. Radiographic severity scores (Marshall score [p = 0.007], Rotterdam score [p = 0.004]), and initial hematoma volume [p = 0.005] were associated with PICH. Higher levels of admission glial fibrillary acidic protein (p < 0.001) and MAP (p = 0.011) were also associated with PICH. Of the PCs, PC1 was significantly associated with PICH (p = 0.0125). Using multimodal data input, machine learning classifiers successfully discriminated patients with or without PICH. Models composed of machine-selected features performed better than models composed of expert-selected variables (reaching an average of 77% accuracy, AUC = 0.78 versus AUC = 0.68 for the expert-selected variables). Predictive models utilizing variables measured at admission can accurately predict PICH, confirmed by the 6-hour follow-up hCT. Our best-performing models must now be externally validated in a separate cohort of TBI patients with low GCS and initial hCT positive for hemorrhage.