Multiple Evaluation Metrics Research Articles

Deep learning-based multi-criteria recommender system for technology-enhanced learning

Multi-Criteria Recommender Systems (MCRSs) improve personalization by incorporating multiple user preferences. However, their application in Technology-Enhanced Learning (TEL) remains limited due to challenges such as data sparsity, over-specialization, and cold-start problems. Traditional techniques, such as Singular Value Decomposition (SVD) and SVD + + , struggle to effectively model the complex interactions within multi-criteria rating data, leading to suboptimal recommendations. This paper introduces a hybrid DeepFM-SVD + + model, which integrates deep learning and factorization-based techniques to improve multi-criteria recommendations. The model captures both low-order feature interactions using factorization machines and high-order dependencies through deep neural networks, enabling more adaptive recommendations. To evaluate its performance, the model is tested on two multi-criteria datasets: ITM-Rec (TEL domain) and Yahoo Movies (non-TEL domain). The experimental results show that DeepFM-SVD + + consistently outperforms the traditional techniques across multiple evaluation metrics. The findings highlight significant improvements in accuracy, demonstrating the model’s effectiveness in sparse datasets and its generalization across domains. By addressing the limitations of existing MCRS techniques, this study contributes to advancing personalized learning recommendations in TEL and expands the applicability of deep learning-based MCRS beyond educational contexts.

A CNN-transformer-based hybrid U-shape model with long-range relay for esophagus 3D CT image gross tumor volume segmentation.

Accurate and reliable segmentation of esophageal gross tumor volume (GTV) in computed tomography (CT) is beneficial for diagnosing and treating. However, this remains a challenging task because the esophagus has a variable shape and extensive vertical range, resulting in tumors potentially appearing at any position within it. This study introduces a novel CNN-transformer-based U-shape model (LRRM-U-TransNet) designed to enhance the segmentation accuracy of esophageal GTV. By leveraging advanced deep learning techniques, we aim to address the challenges posed by the variable shape and extensive range of the esophagus, ultimately improving diagnostic and treatmentoutcomes. Specifically, we propose a long-range relay mechanism to converge all layer feature information by progressively passing adjacent layer feature maps in the pixel and semantic pathways. Moreover, we propose two ready-to-use blocks to implement this mechanism concretely. The Dual FastViT block interacts with feature maps from two paths to enhance feature representation capabilities. The Dual AxialViT block acts as a secondary auxiliary bottleneck to acquire global information for more precise feature mapreconstruction. We build a new esophageal tumor dataset with 1665 real-world patient CT samples annotated by five expert radiologists and employ multiple evaluation metrics to validate our model. Results of a five-fold cross-validation on this dataset show that LRRM-U-TransNet achieves a Dice coefficient of 0.834, a Jaccard coefficient of 0.730, a Precision of 0.840, a HD95 of 3.234 mm, and a Volume Similarity of 0.143. We propose a CNN-Transformer hybrid deep learning network to improve the segmentation effect of esophageal tumors. We utilize the local and global information between shallower and deeper layers to prevent early information loss and enhance the cross-layer communication. To validate our model, we collect a dataset composed of 1665 CT images of esophageal tumors from Sichuan Tumor Hospital. The results show that our model outperforms the state-of-the-art models. It is of great significance to improve the accuracy and clinical application of esophageal tumorsegmentation.

One-Shot Reference-based Structure-Aware Image to Sketch Synthesis

Generating sketches that accurately reflect the content of reference images presents numerous challenges. Current methods either require paired training data or fail to accommodate a wider range and diversity of sketch styles. While pre-trained diffusion models have shown strong text-based control capabilities for reference-based content sketch generation, state-of-the-art methods still struggle with reference-based sketch generation for given content. The main difficulties lie in (1) balancing content preservation with style enhancement, and (2) representing content image textures at varying levels of abstraction to approximate the reference sketch style. In this paper, we propose a method (Ref2Sketch-SA) that transforms a given content image into a sketch based on a reference sketch. The core strategies include (1) using DDIM Inversion to enhance structural consistency in the sketch generation of content images; (2) injecting noise into the input image during the denoising process to produce a sketch that retains content attributes while aligning with, yet differing in texture from, the reference. Our model demonstrates superior performance across multiple evaluation metrics, including user style preference.

Graphic Design with Large Multimodal Model

In the field of graphic design, automating the integration of design elements into a cohesive multi-layered artwork not only boosts productivity but also paves the way for the democratization of graphic design. One existing practice is Graphic Layout Generation (GLG), which aims to layout sequential design elements. It has been constrained by the necessity for a predefined correct sequence of layers, thus limiting creative potential and increasing user workload. In this paper, we present Hierarchical Layout Generation (HLG) as a more flexible and pragmatic setup, which creates graphic composition from any-ordered sets of design elements. To tackle the HLG task, we introduce Graphist, the first layout generation model based on large multimodal models. Graphist efficiently reframes the HLG as a sequence generation problem, utilizing RGB-A images as input, outputs a JSON draft protocol, indicating the coordinates, size, and order of each element. We develop multiple evaluation metrics for HLG. Graphist outperforms prior arts and establishes a strong baseline for this field.

Machine Learning-Based COVID-19 Prediction from Chest X-Ray Images

The rapid spread of COVID-19 has underscored the need for efficient and accurate diagnostic methods. While Reverse Transcription Polymerase Chain Reaction (RT-PCR) remains the gold standard for COVID-19 diagnosis, it is often time-consuming and resource-intensive. In this paper, we propose an automated approach for COVID-19 detection using chest X-ray (CXR) images, leveraging deep learning techniques, specifically Convolutional Neural Networks (CNNs). Our model is designed to classify CXR images into three categories: COVID-19, pneumonia, and normal cases. We utilized publicly available datasets and incorporated data augmentation techniques to improve model generalization and robustness. The proposed system achieves high performance across multiple evaluation metrics, including precision, recall, and F1-score, demonstrating its potential as a fast, cost-effective, and non-invasive diagnostic tool. This work highlights the promise of deep learning for COVID-19 screening, particularly in resource-limited settings where access to RT-PCR may be constrained Keywords COVID-19 detection, chest X-ray, deep learning, convolutional neural networks (CNNs), data augmentation, automated diagnosis, precision, recall, F1-score, resource-limited settings.

An explainable machine-learning model for predicting persistent sepsis associated acute kidney injury: development, validation, and comparison with CCL14.

Persistent sepsis-associated acute kidney injury (SA-AKI) portends worse clinical outcomes and remains a therapeutic challenge for clinicians. Early identification and prediction of persistent SA-AKI is crucial. The aim of this study was to develop and validate an interpretable machine learning (ML) model that predicts persistent SA-AKI, and to compare its diagnostic performance with CCL14 in a prospective cohort. Four retrospective cohorts and one prospective cohort were used for model derivation and validation. The derivation cohort utilized the MIMIC-IV database, randomly split into 80% for model construction and 20% for internal validation. External validation is conducted using subsets of the MIMIC-III dataset, the e-ICU dataset, and retrospective cohorts from the ICU of a Northern Jiangsu people's hospital. Prospective data from the same ICU were used for validation and compared with urinary CCL14 biomarker measurements. AKI was defined based on serum creatinine and urine output, using the kidney disease: Improving Global Outcomes (KDIGO) criteria. Routine clinical data within the first 24 hours of ICU admission were collected, and eight ML algorithms were utilized to construct the prediction model. Multiple evaluation metrics, including the area under the receiver operating characteristic curve (AUC), were employed to compare predictive performance. Feature importance was ranked using SHAP, and the final model was explained accordingly. In addition, the model is developed into a web-based application using the Streamlit framework to facilitate its clinical application. In this study, a total of 46,097 sepsis patients from multiple cohorts were enrolled for analysis. Among the 17,928 sepsis patients in the derivation cohort, 8,081 cases (45.1%) developed into persistent SA-AKI. Among eight ML models, the Gradient Boosting Machine (GBM) model demonstrated superior discriminative ability. Following feature importance ranking, a final interpretable GBM model comprising twelve features (AKI stage, Δcreatinine, urine output, furosemide dose, BMI, SOFA score, KRT, mechanical ventilation, lactate, Bun, PT and age) was established. The final model accurately predicted the occurrence of persistent SA-AKI in both internal (AUC = 0.870) and external validation cohorts (MIMIC-III subset: AUC = 0.891, e-ICU dataset: AUC = 0.932, North Jiangsu people's Hospital retrospective cohort: AUC = 0.983). In the prospective cohort, the GBM model outperformed urinary CCL14 in predicting persistent SA-AKI (GBM AUC = 0.852 vs. CCL14 AUC = 0.821). Additionally, the model has been transformed into an online clinical tool to facilitate its application in clinical settings. The interpretable GBM model has been shown to successfully and accurately predict the occurrence of persistent SA-AKI, demonstrating good predictive ability in both internal and external validation cohorts. Furthermore, the model has been demonstrated to outperform the biomarker CCL14 in prospective cohort validation.

Rapid Response Antimicrobial Peptide Design Strategy Driven by Meta-Learning for Emerging Drug-Resistant Pathogens.

Antimicrobial resistance (AMR) presents a critical global health threat requiring urgent intervention. In order to swiftly respond to and control the spread of emerging drug-resistant bacteria at the onset of their proliferation, our aim is to develop a Rapid Response Antimicrobial Peptide (AMP) design strategy (RR-ADS). This framework addresses the challenge of limited pathogen-specific data by achieving robust generalization from minimal samples by meta-learning and reinforcement learning, optimizing both biocompatibility and efficacy against drug-resistant pathogens. Our model has achieved satisfactory results across multiple evaluation metrics, demonstrating the capability to accurately identify and generate AMPs targeted against drug-resistant bacteria with minimal sample sizes. Within 2 weeks, we successfully designed and experimentally verified AMPs against multidrug-resistant Acinetobacter baumannii, achieving a 93.3% positive rate. RR-ADS has effectively demonstrated the potential of meta-learning in tasks involving bioactive peptides and holds promise as an effective alternative measure to address infectious disease public health emergencies.

Bootstrapping BI-RADS classification using large language models and transformers in breast magnetic resonance imaging reports

Breast cancer is one of the most common malignancies among women globally. Magnetic resonance imaging (MRI), as the final non-invasive diagnostic tool before biopsy, provides detailed free-text reports that support clinical decision-making. Therefore, the effective utilization of the information in MRI reports to make reliable decisions is crucial for patient care. This study proposes a novel method for BI-RADS classification using breast MRI reports. Large language models are employed to transform free-text reports into structured reports. Specifically, missing category information (MCI) that is absent in the free-text reports is supplemented by assigning default values to the missing categories in the structured reports. To ensure data privacy, a locally deployed Qwen-Chat model is employed. Furthermore, to enhance the domain-specific adaptability, a knowledge-driven prompt is designed. The Qwen-7B-Chat model is fine-tuned specifically for structuring breast MRI reports. To prevent information loss and enable comprehensive learning of all report details, a fusion strategy is introduced, combining free-text and structured reports to train the classification model. Experimental results show that the proposed BI-RADS classification method outperforms existing report classification methods across multiple evaluation metrics. Furthermore, an external test set from a different hospital is used to validate the robustness of the proposed approach. The proposed structured method surpasses GPT-4o in terms of performance. Ablation experiments confirm that the knowledge-driven prompt, MCI, and the fusion strategy are crucial to the model’s performance.

CRISPR-MFH: A Lightweight Hybrid Deep Learning Framework with Multi-Feature Encoding for Improved CRISPR-Cas9 Off-Target Prediction

Background: The CRISPR-Cas9 system has emerged as one of the most promising gene-editing technologies in biology. However, off-target effects remain a significant challenge. While recent advances in deep learning have led to the development of models for off-target prediction, these models often fail to fully leverage sequence pair information. Furthermore, as the models’ parameter sizes increase, so do their complexities, limiting their practical applicability. Methods: In this study, we introduce a novel multi-feature independent encoding method, which encodes the gRNA-DNA sequence pair into three distinct feature matrices to minimize information loss. Additionally, we propose a lightweight hybrid deep learning framework, CRISPR-MFH, that integrates multi-scale separable convolutions and hybrid attention mechanisms for efficient and accurate off-target prediction. Results: Extensive experiments across multiple benchmark datasets demonstrate that the proposed encoding method effectively captures critical features and that CRISPR-MFH outperforms or matches state-of-the-art models with significantly fewer parameters across multiple evaluation metrics. Conclusions: This study offers a novel perspective for advancing deep learning technology in the realm of CRISPR-Cas9 off-target detection.

Dynamic Hierarchical Convolutional Attention Network for Recognizing Motor Imagery Intention.

The neural activity patterns of localized brain regions are crucial for recognizing brain intentions. However, existing electroencephalogram (EEG) decoding models, especially those based on deep learning, predominantly focus on global spatial features, neglecting valuable local information, potentially leading to suboptimal performance. Therefore, this study proposed a dynamic hierarchical convolutional attention network (DH-CAN) that comprehensively learned discriminative information from both global and local spatial domains, as well as from time-frequency domains in EEG signals. Specifically, a multiscale convolutional block was designed to dynamically capture time-frequency information. The channels of EEG signals were mapped to different brain regions based on motor imagery neural activity patterns. The spatial features, both global and local, were then hierarchically extracted to fully exploit the discriminative information. Furthermore, regional connectivity was established using a graph attention network, incorporating it into the local spatial features. Particularly, this study shared network parameters between symmetrical brain regions to better capture asymmetrical motor imagery patterns. Finally, the learned multilevel features were integrated through a high-level fusion layer. Extensive experimental results on two datasets demonstrated that the proposed model performed excellently across multiple evaluation metrics, exceeding existing benchmark methods. These findings suggested that the proposed model offered a novel perspective for EEG decoding research.

REU-Net: A Remote Sensing Image Building Segmentation Network Based on Residual Structure and the Edge Enhancement Attention Module

Building segmentation from high-resolution remote sensing images plays a crucial role in cadastral measurement, ecological monitoring, urban planning, and other applications. To address the current challenges in building segmentation from high-resolution remote sensing images, this paper proposes an improved deep learning-based network—REU-Net(2EEAM). The network replaces traditional convolutional blocks in U-Net with Residual Structures, deepening the network and alleviating the issue of vanishing gradients. Additionally, it substitutes the direct skip connections with two Edge Enhancement Attention Modules (EEAMs), enhancing the network’s ability to extract building edge information. Furthermore, a hybrid loss function combining edge consistency loss and binary cross-entropy loss is used to train the network, aiming to improve segmentation accuracy. Experimental results show that REU-Net(2EEAM) achieves optimal performance across multiple evaluation metrics (such as P, MPA, MIoU, and FWIoU), particularly excelling in the accurate recognition of building edges, significantly outperforming other network models. This method provides a reliable foundation for the further optimization of building segmentation algorithms for remote sensing images.

A Deep Learning Model Based on Bidirectional Temporal Convolutional Network (Bi-TCN) for Predicting Employee Attrition

Employee attrition, which causes a significant loss for an organization, is the term used to describe the natural decline in the number of employees in an organization as a result of numerous unavoidable events. If a company can predict the likelihood of an employee leaving, it can take proactive steps to address the issue. In this study, we introduce a deep learning framework based on a Bidirectional Temporal Convolutional Network (Bi-TCN) to predict employee attrition. We conduct extensive experiments on two publicly available datasets, including IBM and Kaggle, comparing our model’s performance against classical machine learning, deep learning models, and state-of-the-art approaches across multiple evaluation metrics. The proposed model yields promising results in predicting employee attrition, achieving accuracy rates of 89.65% on the IBM dataset and 97.83% on the Kaggle dataset. We also apply a fully connected GAN-based data augmentation technique and three oversampling methods to augment and balance the IBM dataset. The results show that our proposed model, combined with the GAN-based approach, improves accuracy to 92.17%. We also applied the SHAP method to identify the key features that most significantly influence employee attrition. These findings demonstrate the efficacy of our model, showcasing its potential for use in various industries and organizations.

A Semi-Supervised Diffusion-Based Framework for Weed Detection in Precision Agricultural Scenarios Using a Generative Attention Mechanism

The development of smart agriculture has created an urgent demand for efficient and accurate weed recognition and detection technologies. However, the diverse and complex morphology of weeds, coupled with the scarcity of labeled data in agricultural scenarios, poses significant challenges to traditional supervised learning methods. To address these issues, a weed detection model based on a semi-supervised diffusion generative network is proposed. This model integrates a generative attention mechanism and semi-diffusion loss to enable the efficient utilization of both labeled and unlabeled data. Experimental results demonstrate that the proposed method outperforms existing approaches across multiple evaluation metrics, achieving a precision of 0.94, recall of 0.90, accuracy of 0.92, and mAP@50 and mAP@75 of 0.92 and 0.91, respectively. Compared to traditional methods such as DETR, precision and recall are improved by approximately 10% and 8%, respectively. Additionally, compared to the enhanced YOLOv10, mAP@50 and mAP@75 are increased by 1% and 2%, respectively. The proposed semi-supervised diffusion weed detection model provides an efficient and reliable solution for weed recognition and introduces new research perspectives for the application of semi-supervised learning in smart agriculture. This framework establishes both theoretical and practical foundations for addressing complex target detection challenges in the agricultural domain.

House Price Prediction

House price prediction is a crucial aspect of real estate analytics, impacting buyers, sellers, investors, and policymakers. The real estate industry is a dynamic sector where property values fluctuate based on numerous factors such as economic conditions, location, infrastructure, demand-supply balance, and regulatory policies. Traditional valuation methods often rely on expert opinions and manual assessments, which may introduce biases and inconsistencies. In contrast, data-driven approaches, particularly machine learning (ML) techniques, offer a more accurate and systematic way to estimate property values. This study explores the effectiveness of machine learning models in predicting house prices by analyzing historical real estate data. Various predictive models, including linear regression, decision trees, random forests, gradient boosting machines (XGBoost), and deep learning algorithms, are evaluated in terms of their performance. The research aims to determine which model provides the highest accuracy while ensuring interpretability and efficiency. Data pre-processing plays a vital role in enhancing the accuracy of house price predictions. The dataset used in this study includes essential property attributes such as square footage, the number of bedrooms and bathrooms, locality, age of the property, and market trends. Before applying predictive models, pre-processing techniques such as handling missing values, encoding categorical variables, scaling numerical features, and feature selection are implemented to optimize the dataset. The study also examines the impact of feature engineering, where derived attributes such as neighbourhood crime rates, proximity to schools, and public transportation accessibility are incorporated to improve prediction accuracy. The research adopts multiple evaluation metrics, including Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and R-squared (R²) values, to assess the performance of different models. Initial findings suggest that ensemble learning methods such as Random Forest and XGBoost outperform traditional regression models in terms of predictive accuracy. However, deep learning models, specifically artificial neural networks (ANNs), show promising results when trained on large datasets. The study also highlights the importance of hyper parameter tuning in refining model performance, addressing issues such as over fitting and computational efficiency. One of the key contributions of this research is the comparative analysis of different machine learning algorithms to identify the best-suited approach for house price prediction. Traditional regression techniques, while interpretable, often fail to capture complex nonlinear relationships within the data. Decision tree-based methods, such as Random Forest and XGBoost, demonstrate robust performance by learning intricate patterns and reducing variance through ensemble learning. Deep learning models, on the other hand, excel in capturing high-dimensional interactions but require extensive computational resources and larger training datasets. Beyond predictive accuracy, the study examines the interpretability of machine learning models, which is crucial for real-world applications. Feature importance analysis is conducted to understand which variables have the most significant impact on house prices. Factors such as location, property size, and market demand emerge as primary determinants. The study also investigates the effect of external factors, including economic indicators like interest rates, inflation, and employment rates, on housing prices. Despite the promising results, challenges such as data availability, regional market variations, and the need for real-time data integration remain. Future research can focus on enhancing prediction models by incorporating geospatial analytics, satellite imagery, and deep learning advancements. Additionally, integrating real-time market data and leveraging blockchain technology for transparent property transactions can further improve predictive modelling in real estate.

Integration of Artificial Neural Network Regression and Principal Component Analysis for Indoor Visible Light Positioning.

The advancement of artificial intelligence has brought visible-light positioning (VLP) to the forefront of indoor positioning research, enabling precise localization without additional infrastructure. However, the complex interplay between light propagation phenomena and environmental factors in indoor spaces presents significant challenges for VLP systems. Additionally, the pose of the light-emitting diodes is prior unknown, adding another layer of complexity to the positioning process. Dynamic indoor environments further complicate matters due to user mobility and obstacles, which can affect system accuracy. In this study, user movement is simulated using a constructed dataset with systematically varied receiver positions, reflecting realistic motion patterns rather than real-time movement. While the experimental setup considers a fixed obstacle scenario, the training and testing datasets incorporate position variations to emulate user displacement. Given these dataset characteristics, it is crucial to employ robust positioning techniques that can handle environmental variations. Conventional methods, such as received signal strength (RSS)-based techniques, face practical implementation hurdles due to fluctuations in transmitted optical power and modeling imperfections. Leveraging machine learning techniques, particularly regression-based artificial neural networks (ANNs), offer a promising alternative. ANNs excel at modeling the intricate relationships within data, making them well-suited for handling the complex dynamics of indoor lighting environments. To address the computational complexities arising from high-dimensional data, this research incorporates principal component analysis (PCA) as a method for reducing dimensionality. PCA eases the computational burden, accelerates training speeds by normalizing the data, and reduces loss rates, thereby enhancing the overall efficacy and feasibility of the proposed VLP framework. Rigorous experimentation and validation demonstrate the potential of employing principal components. Experimental results show significant improvements across multiple evaluation metrics for a constellation comprising eight LEDs mounted in a rectangular structure measuring a room dimension of 12 m × 18 m × 6.8 m, with a photodiode (PD) receiver. Specifically, the mean squared error (MSE) values for the training and testing samples are 0.0062 and 0.0456 cm, respectively. Furthermore, the R-squared values of 99.31% and 94.74% for training and testing, respectively, signify a robust predictive performance of the model with low model loss. These findings underscore the efficacy of the proposed PCA-ANN regression model in optimizing VLP systems and providing reliable indoor positioning services.

Top-DTI: Integrating Topological Deep Learning and Large Language Models for Drug Target Interaction Prediction.

The accurate prediction of drug-target interactions (DTI) is a crucial step in drug discovery, providing a foundation for identifying novel therapeutics. Traditional drug development is both costly and time-consuming, often spanning over a decade. Computational approaches help narrow the pool of compound candidates, offering significant starting points for experimental validation. In this study, we propose Top-DTI framework for predicting DTI by integrating topological data analysis (TDA) with large language models (LLMs). Top-DTI leverages persistent homology to extract topological features from protein contact maps and drug molecular images. Simultaneously, protein and drug LLMs generate semantically rich embeddings that capture sequential and contextual information from protein sequences and drug SMILES strings. By combining these complementary features, Top-DTI enhances predictive performance and robustness. Experimental results on the public BioSNAP and Human DTI benchmark datasets demonstrate that the proposed Top-DTI model outperforms state-of-the-art approaches across multiple evaluation metrics, including AUROC, AUPRC, sensitivity, and specificity. Furthermore, the Top-DTI model achieves superior performance in the challenging cold-split scenario, where the test and validation sets contain drugs or targets absent from the training set. This setting simulates real-world scenarios and highlights the robustness of the model. Notably, incorporating topological features alongside LLM embeddings significantly improves predictive performance, underscoring the value of integrating structural and sequence-based representations. The data and source code of Top-DTI is available at https://github.com/bozdaglab/Top_DTI under Creative Commons Attribution Non Commercial 4.0 International Public License.

Multi-Modality Fusion and Tumor Sub-Component Relationship Ensemble Network for Brain Tumor Segmentation.

Deep learning technology has been widely used in brain tumor segmentation with multi-modality magnetic resonance imaging, helping doctors achieve faster and more accurate diagnoses. Previous studies have demonstrated that the weighted fusion segmentation method effectively extracts modality importance, laying a solid foundation for multi-modality magnetic resonance imaging segmentation. However, the challenge of fusing multi-modality features with single-modality features remains unresolved, which motivated us to explore an effective fusion solution. We propose a multi-modality and single-modality feature recalibration network for magnetic resonance imaging brain tumor segmentation. Specifically, we designed a dual recalibration module that achieves accurate feature calibration by integrating the complementary features of multi-modality with the specific features of a single modality. Experimental results on the BraTS 2018 dataset showed that the proposed method outperformed existing multi-modal network methods across multiple evaluation metrics, with spatial recalibration significantly improving the results, including Dice score increases of 1.7%, 0.5%, and 1.6% for the enhanced tumor core, whole tumor, and tumor core regions, respectively.

Estimating water surface elevation for a wetland using integrated multi-sourced remote sensing data

Surface water plays an important role in understanding the hydrological behaviour of a wetland and is crucial for the sustainability of wetland ecosystems. Remote sensing increasingly is used for the estimation of surface water levels in larger inland waterbodies. However, there are few investigations that have employed multi-sourced remote sensing data for water level predictions in wetlands, which was the motivation for undertaking this study. Sentinel-2 and Landsat-8 are among the latest satellites providing optical imagery with high spatial resolution and coverage that are available in the public domain. Different water indices have been applied to estimate surface water levels using these satellite image sources; however, what index to use for a particular application requires thorough, site-specific analysis. In this study, the Normalized Difference Water Index (NDWI), two versions of the Modified Normalized Difference Water Index (MNDWI), and the Water Ratio Index (WRI) were used to estimate water levels in a constructed wetland, as part of an effort to better guide regulation and decision-making for a local management agency. The satellite data were complemented with high resolution aerial photogrammetric images and LiDAR data to assess the accuracy of water level predictions provided by the satellite images. The photogrammetric images were used as reference datasets while the LiDAR data supported the development of area-elevation curves for the wetland. Accuracy assessment between the satellite and reference images was performed using the Kappa co-efficient (K). MNDWI performed better than the other water indices for both satellite data sources; however, the optimum threshold was different for each satellite (− 0.35 for Sentinel-2 and − 0.25 for Landsat-8). K values for the optimum threshold ranged between 0.72 and 0.77 for Sentinel-2 and 0.73 and 0.87 for Landsat-8. The water levels estimated using the remotely sensed data were assessed against in situ, continuously measured water levels using multiple efficiency evaluation metrics including R2, RMSE, and SSE. Estimated water levels with Sentinel-2 and Landsat-8 resulted in an R2 of 0.86 and 0.88, RMSE of 0.04 m and 0.06 m, and an SSE of 0.02 m and 0.06 m, respectively. These results show that even for a small wetland, it is possible to use satellite imagery to estimate water levels with high accuracy.

Multimodal Metaverse Healthcare: A Collaborative Representation and Adaptive Fusion Approach for Generative Artificial-Intelligence-Driven Diagnosis.

The metaverse enables immersive virtual healthcare environments, presenting opportunities for enhanced care delivery. A key challenge lies in effectively combining multimodal healthcare data and generative artificial intelligence abilities within metaverse-based healthcare applications, which is a problem that needs to be addressed. This paper proposes a novel multimodal learning framework for metaverse healthcare, MMLMH, based on collaborative intra- and intersample representation and adaptive fusion. Our framework introduces a collaborative representation learning approach that captures shared and modality-specific features across text, audio, and visual health data. By combining modality-specific and shared encoders with carefully formulated intrasample and intersample collaboration mechanisms, MMLMH achieves superior feature representation for complex health assessments. The framework's adaptive fusion approach, utilizing attention mechanisms and gated neural networks, demonstrates robust performance across varying noise levels and data quality conditions. Experiments on metaverse healthcare datasets demonstrate MMLMH's superior performance over baseline methods across multiple evaluation metrics. Longitudinal studies and visualization further illustrate MMLMH's adaptability to evolving virtual environments and balanced performance across diagnostic accuracy, patient-system interaction efficacy, and data integration complexity. The proposed framework has a unique advantage in that a similar level of performance is maintained across various patient populations and virtual avatars, which could lead to greater personalization of healthcare experiences in the metaverse. MMLMH's successful functioning in such complicated circumstances suggests that it can combine and process information streams from several sources. They can be successfully utilized in next-generation healthcare delivery through virtual reality.

Laryngeal cancer diagnosis based on improved YOLOv8 algorithm

Abstract Laryngeal cancer is the most common malignant tumor in the head and neck region. The larynx, also known as the voice box, plays a crucial role in voice production and ventilation. Enhancing the diagnosis and treatment of laryngeal cancer can significantly improve patients’ prognosis and quality of life. Artificial intelligence (AI) technology shows promise as a valuable tool for diagnosing laryngeal cancer. It not only reduces the burden on endoscopists in interpreting images but also performs screening and diagnosis efficiently and accurately. However, due to the hidden and diverse nature of laryngeal cancer lesions, achieving accuracy and efficiency in AI-based diagnosis presents poses challenges. This study introduces an improved YOLOv8 algorithm named MSEC-YOLO, specifically designed for the detection and classification tasks of laryngeal cancer in endoscopic images. A novel multiscale enhanced convolution module has been introduced to improve the model’s feature extraction capabilities for small-sized targets. Additionally, a tiny fully convolutional network architecture has been employed, reducing the number of model parameters and computational costs while maintaining or enhancing performance, which is crucial for real-time medical imaging analysis. The experiments utilized a real-world endoscopic image dataset from the hospital, and the results indicated that MSEC-YOLO outperformed the original YOLOv8 model and its multi-kernel versions across multiple evaluation metrics, especially in critical categories such as malignant tumors, polyps, and papillomas, demonstrating extremely high precision and recall rates.