Classification Precision Research Articles

Transformer-based deep learning networks for fault detection, classification, and location prediction in transmission lines

ABSTRACT Fault detection, classification, and location prediction are crucial for maintaining the stability and reliability of modern power systems, reducing economic losses, and enhancing system protection sensitivity. This paper presents a novel Hierarchical Deep Learning Approach (HDLA) for accurate and efficient fault diagnosis in transmission lines. HDLA leverages two-stage transformer-based classification and regression models to perform Fault Detection (FD), Fault Type Classification (FTC), and Fault Location Prediction (FLP) directly from synchronized raw three-phase current and voltage samples. By bypassing the need for feature extraction, HDLA significantly reduces computational complexity while achieving superior performance compared to existing deep learning methods. The efficacy of HDLA is validated on a comprehensive dataset encompassing various fault scenarios with diverse types, locations, resistances, inception angles, and noise levels. The results demonstrate significant improvements in accuracy, recall, precision, and F1-score metrics for classification, and Mean Absolute Errors (MAEs) and Root Mean Square Errors (RMSEs) for prediction, showcasing the effectiveness of HDLA for real-time fault diagnosis in power systems.

Coherent Feature Extraction with Swarm Intelligence Based Hybrid Adaboost Weighted ELM Classification for Snoring Sound Classification.

For patients suffering from obstructive sleep apnea and sleep-related breathing disorders, snoring is quite common, and it greatly interferes with the quality of life for them and for the people surrounding them. For diagnosing obstructive sleep apnea, snoring is used as a screening parameter, so the exact detection and classification of snoring sounds are quite important. Therefore, automated and very high precision snoring analysis and classification algorithms are required. In this work, initially the features are extracted from six different domains, such as time domain, frequency domain, Discrete Wavelet Transform (DWT) domain, sparse domain, eigen value domain, and cepstral domain. The extracted features are then selected using three efficient feature selection techniques, such as Golden Eagle Optimization (GEO), Salp Swarm Algorithm (SSA), and Refined SSA. The selected features are finally classified with the help of eight traditional machine learning classifiers and two proposed classifiers, such as the Firefly Algorithm-Weighted Extreme Learning Machine hybrid with Adaboost model (FA-WELM-Adaboost) and the Capuchin Search Algorithm-Weighted Extreme Learning Machine hybrid with Adaboost model (CSA-WELM-Adaboost). The analysis is performed on the MPSSC Interspeech dataset, and the best results are obtained when the DWT features with the refined SSA feature selection technique and FA-WELM-Adaboost hybrid classifier are utilized, reporting an Unweighted Average Recall (UAR) of 74.23%. The second-best results are obtained when DWT features are selected with the GEO feature selection technique and a CSA-WELM-Adaboost hybrid classifier is utilized, reporting an UAR of 73.86%.

A hierarchical deep learning approach for diagnosing impacted canine-induced root resorption via cone-beam computed tomography

ObjectivesCanine-induced root resorption (CIRR) is caused by impacted canines and CBCT images have shown to be more accurate in diagnosing CIRR than panoramic and periapical radiographs with the reported AUCs being 0.95, 0.49, and 0.57, respectively. The aim of this study was to use deep learning to automatically evaluate the diagnosis of CIRR in maxillary incisors using CBCT images.MethodsA total of 50 cone beam computed tomography (CBCT) images and 176 incisors were selected for the present study. The maxillary incisors were manually segmented and labeled from the CBCT images by two independent radiologists as either healthy or affected by root resorption induced by the impacted canines. We used five different strategies for training the model: (A) classification using 3D ResNet50 (Baseline), (B) classification of the segmented masks using the outcome of a 3D U-Net pretrained on the 3D MNIST, (C) training a 3D U-Net for the segmentation task and use its outputs for classification, (D) pretraining a 3D U-Net for the segmentation and transfer of the model, and (E) pretraining a 3D U-Net for the segmentation and fine-tuning the model with only the model encoder. The segmentation models were evaluated using the mean intersection over union (mIoU) and Dice coefficient (DSC). The classification models were evaluated in terms of classification accuracy, precision, recall, and F1 score.ResultsThe segmentation model achieved a mean intersection over union (mIoU) of 0.641 and a DSC of 0.901, indicating good performance in segmenting the tooth structures from the CBCT images. For the main classification task of detecting CIRR, Model C (classification of the segmented masks using 3D ResNet) and Model E (pretraining on segmentation followed by fine-tuning for classification) performed the best, both achieving 82% classification accuracy and 0.62 F1-scores on the test set. These results demonstrate the effectiveness of the proposed hierarchical, data-efficient deep learning approaches in improving the accuracy of automated CIRR diagnosis from limited CBCT data compared to the 3D ResNet baseline model.ConclusionThe proposed approaches are effective at improving the accuracy of classification tasks and are helpful when the diagnosis is based on the volume and boundaries of an object. While the study demonstrated promising results, future studies with larger sample size are required to validate the effectiveness of the proposed method in enhancing the medical image classification tasks.

DCM-GIFT: An Android malware dynamic classification method based on gray-scale image and feature-selection tree

Context:The boom of Android market makes mobile products more popular and convenient. However, in the face of the complex Android application market, how to efficiently and accurately identify malware has become one of the focuses of research. Various new types of disguised malware lurk in the web pages, links and major application malls. Therefore, people’s privacy and property security have become a major obstacle to the continued development of mobile devices. Objective:Most of the existing malware classification methods are fixed on one or several types of characteristics of Android devices, such as static characteristics, dynamic characteristics and traffic characteristics. Single feature detection or fixed feature fusion models limit the dimension of detection software, and also cause imbalanced classification results. This paper proposes an Android Malware Dynamic Classification Method based on Gray-scale Image and Feature-selection Tree (DCM-GIFT), which aims to improve and stabilize the precision of Android software classification and enhance the robustness of malware classification. Method:In this paper, we construct gray-scale images for the original Android traffic to retain the characteristics of the time series and spatial structure of the original network traffic. At the same time, we take the dynamic information and static information of Android software as auxiliary features to build a feature selection tree. The feature-selection algorithm helps the classifier dynamically select the optimal feature fusion scheme, and the resulting fusion feature vector will be trained and predicted using machine learning clusters for model training. Results:We evaluate the performance of DCM-GIFT on multiple datasets published at the Canadian Institute for Cybersecurity, the area under the accuracy, precision, recall and F1measure. The results show that the proposed DCM-GIFT model has significantly better prediction performance compared to other software classification models. Conclusion:It can be concluded that: (1) In terms of accuracy, precision, recall and F1measure, the DCM-GIFT model has a higher average value. (2) The DCM-GIFT model effectively solves the problem of imbalanced classification results in Android software. (3) The DCM-GIFT model achieves the goal of dynamic feature fusion and significantly improves the utilization of system resources.

Radar-Based Elderly Fall Detection Using Smoothed Pseudo Wigner Ville Distribution and XGBoost Learning.

Falls among the elderly population pose significant health risks, often leading to morbidity and decreased quality of life. Traditional fall detection methods, namely wearable devices and cameras, have limitations such as lighting conditions and privacy concerns. Radar-based fall detection has emerged as a promising alternative, offering unobtrusive technique. In this study, an attempt has been made to classify fall detection using smoothed pseudo wigner-ville distribution (SPWVD) images and XGBoost learning. For this, online publicly available radar database (N=15) is considered. Radar signals is employed to SPWVD for time-frequency representation images. Ten features are extracted and applied to XGBoost learning. Experiments are performed and performance is evaluated using 10-fold cross validation. The proposed approach is able to discriminate elderly fall. Using XGBoost learning, the approach yields a maximum average classification accuracy, f1-score, precision, sensitivity, specificity, and kappa scores of 87.47%, 87.38%, 88.12%, 86.81%, 88.31% and 74.94% respectively. The combination of conventional features with concentration measures and median frequency obtained the second best performance. Thus, the proposed framework could be utilized for accurate and efficient detection of falls among the elderly population in their private spaces.

Enhancing acceleration data cleaning through time–frequency feature integration using deep learning

In recent years, the safety issues of bridge has garnered significant attention from various sectors of society. Bridge structural health monitoring (SHM) systems have emerged as essential tools for real-time monitoring and assessing the operational safety of bridges. Despite the substantial implementation and operation of monitoring systems, challenges have arisen from abnormal data caused by equipment malfunctions, power outages, and environmental interference. These anomalies are challenging to manually remove within the vast dataset. To address the challenge of automated detection and classification of anomalies, this paper focuses on acceleration data and proposes a data cleaning method that integrates time–frequency features and one-dimensional convolutional neural network (1D CNN). The method utilizes extreme value envelopes and wavelet scattering transforms to extract time–frequency features from acceleration data, thereby achieving data dimensionality reduction and feature extraction goals. Subsequently, these features are employed to train and test a 1D CNN. The proposed method’s validation is conducted using acceleration monitoring data obtained from a large-span arch bridge. Through extensive testing, the model demonstrates a classification accuracy exceeding 98% on both the test set and unseen data, showcasing high classification precision and generalization capabilities. This approach provides an effective solution for automated data cleaning and anomaly identification within the context of bridge health monitoring systems.

A stacking ensemble machine learning based approach for classification of plant diseases through leaf images

Diseases and pests in plants/crops are major causes of significant agricultural losses with economic, social and ecological impacts. Therefore, there is a need for early identification of plant diseases and pests through automated systems. Recently, machine learning-based methods have become popular in solving agricultural problems such as plant diseases faced by technically-noob farmers. This work proposes a novel method based on stacking ensemble machine learning to detect plant diseases in Uradbean precisely. Two classifiers: support vector machine (SVM), random forest (RF) are trained on a dataset consists of Uradbean infected and healthy leaf images. These classifiers are stacked with logistic regression (LR) classifier. In the diverse ensemble, LR classifier is used as a meta-learner which enhanced the precision of the disease classification. The fuzzy C-Means clustering with particle swarm optimization is used for image segmentation. Haralick, Hu Moments and color histogram methods are used in feature extraction. During the tests, the proposed model is also compared with pre-trained networks: DenseNet-201, ResNet-50, and VGG19. It achieved an impressive classification accuracy of 96.82 % which is higher than the individual classifiers and pre-trained networks. To validate model performance, it is evaluated on a benchmark public dataset consists of Apple leaf images and achieved 98.30% accuracy. It is observed that ensemble method reflects an advantage over individual models in increasing the classification rates and reducing the computational overhead in comparison to pre-trained networks which struggle due to the issues such as irrelevant features, generation of pertinent characteristics, and noise

Middle ear-acquired cholesteatoma diagnosis based on CT scan image mining using supervised machine learning models

BackgroundDistinguishing between middle ear cholesteatoma and chronic suppurative otitis media (CSOM) is an ongoing challenge. While temporal bone computed tomography (CT) scan is highly accurate for diagnosing middle ear conditions, its specificity in discerning between cholesteatoma and CSOM is only moderate. To address this issue, we utilized trained machine learning models to enhance the specificity of temporal bone CT scan in diagnosing middle ear cholesteatoma. Our database consisted of temporal bone CT scan native images from 122 patients diagnosed with middle ear cholesteatoma and a control group of 115 patients diagnosed with CSOM, with both groups labeled based on surgical findings. We preprocessed the native images to isolate the region of interest and then utilized the Inception V3 convolutional neural network for image embedding into data vectors. Classification was performed using machine learning models including support vector machine (SVM), k-nearest neighbors (k-NN), random forest, and neural network. Statistical metrics employed to interpret the results included classification accuracy, precision, recall, F1 score, confusion matrix, area under the receiver operating characteristic curve (AUC), and FreeViz diagram.ResultsOur training dataset comprised 5390 images, and the testing dataset included 125 different images. The neural network, k-NN, and SVM models demonstrated significantly higher relevance in terms of classification accuracy, precision, and recall compared to the random forest model. For instance, the F1 scores were 0.974, 0.987, and 0.897, respectively, for the former three models, in contrast to 0.661 for the random forest model.ConclusionThe performance metrics of the presented trained machine learning models hold promising prospects as potentially clinically useful aids.

Personalized identification of autism-related bacteria in the gut microbiome using explainable artificial intelligence

Autism spectrum disorder (ASD) affects social interaction and communication. Emerging evidence links ASD to gut microbiome alterations, suggesting that microbial composition may play a role in the disorder. This study employs explainable artificial intelligence (XAI) to examine the contributions of individual microbial species to ASD. By using local explanation embeddings and unsupervised clustering, the research identifies distinct ASD subgroups, underscoring the disorder's heterogeneity. Specific microbial biomarkers associated with ASD are revealed, and the best classifiers achieved an AU-ROC of 0.965± 0.005 and an AU-PRC of 0.967± 0.008. The findings support the notion that gut microbiome composition varies significantly among individuals with ASD. This work's broader significance lies in its potential to inform personalized interventions, enhancing precision in ASD management and classification. These insights highlight the importance of individualized microbiome profiles for developing tailored therapeutic strategies for ASD.

A Hybrid Lung Cancer Model for Diagnosis and Stage Classification from Computed Tomography Images

Pulmonary cancers at early stages is difficult but crucial for patient survival. Therefore, it is essential to develop an intelligent, autonomous, and accurate lung cancer detection system that shows great reliability compared to previous systems and research. In this study, we have developed an innovative lung cancer detection system known as the Hybrid Lung Cancer Stage Classifier and Diagnosis Model (Hybrid-LCSCDM). This system simplifies the complex task of diagnosing lung cancer by categorizing patients into three classes: normal, benign, and malignant, by analyzing computed tomography (CT) scans using a two-part approach: First, feature extraction is conducted using a pre-trained model called VGG-16 for detecting key features in lung CT scans indicative of cancer. Second, these features are then classified using a machine learning technique called XGBoost, which sorts the scans into three categories. A dataset, IQ-OTH/NCCD - Lung Cancer, is used to train and evaluate the proposed model to show its effectiveness. The dataset consists of the three aforementioned classes containing 1190 images. Our suggested strategy achieved an overall accuracy of 98.54 %, while the classification precision among the three classes was 98.63 %. Considering the accuracy, recall, and precision as well as the F1-score evaluation metrics, the results indicated that when using solely computed tomography scans, the proposed (Hybrid-LCSCDM) model outperforms all previously published models.

Noise robustness evaluation of image processing algorithms for eye blinking detection

Robust algorithms for eye blinking detection are required due to the effects of noisy environments and varying light conditions on image-based detection methods. This paper compares five non-supervised image-based algorithms for eye blinking detection, evaluating their robustness to additive Gaussian noise. The algorithms were tested on a video dataset acquired using a smartphone and an ophthalmology chin rest. Through Monte Carlo simulation that introduces Gaussian noise at different intensities, we evaluate the algorithms’ precision, sensitivity, and F1-scores for frame classification, True Positive Rate (TPR), False Discovery Rate (FDR) and F1-score for the event detector. The results of experimental tests reveal significant variations in algorithms’ performance with increasing noise levels. Notably, the Image Correlation (IC) algorithm demonstrates superior eye blinking detection capabilities under various noise conditions, emerging as the most robust algorithm among those tested. This distinction highlights the potential of IC for reliable blink detection in noisy environments.

Breaking Barriers in Cancer Diagnosis: Super‐Light Compact Convolution Transformer for Colon and Lung Cancer Detection

ABSTRACTAccording to the World Health Organization, lung and colon cancers are known for their high mortality rates which necessitate the diagnosis of these cancers at an early stage. However, the limited availability of data such as histopathology images used for diagnosis of these cancers, poses a significant challenge while developing computer‐aided detection system. This makes it necessary to keep a check on the number of parameters in the artificial intelligence (AI) model used for the detection of these cancers considering the limited availability of the data. In this work, a customised compact and efficient convolution transformer architecture, termed, C3‐Transformer has been proposed for the diagnosis of colon and lung cancers using histopathological images. The proposed C3‐Transformer relies on convolutional tokenisation and sequence pooling approach to keep a check on the number of parameters and to combine the advantage of convolution neural network with the advantages of transformer model. The novelty of the proposed method lies in efficient classification of colon and lung cancers using the proposed C3‐Transformer architecture. The performance of the proposed method has been evaluated on the ‘LC25000’ dataset. Experimental results shows that the proposed method has been able to achieve average classification accuracy, precision and recall value of 99.30%, 0.9941 and 0.9950, in classifying the five different classes of colon and lung cancer with only 0.0316 million parameters. Thus, the present computer‐aided detection system developed using proposed C3‐Transformer can efficiently detect the colon and lung cancers using histopathology images with high detection accuracy.

Towards Robust Pansharpening: A Large-Scale High-Resolution Multi-Scene Dataset and Novel Approach

Pansharpening, a pivotal task in remote sensing, involves integrating low-resolution multispectral images with high-resolution panchromatic images to synthesize an image that is both high-resolution and retains multispectral information. These pansharpened images enhance precision in land cover classification, change detection, and environmental monitoring within remote sensing data analysis. While deep learning techniques have shown significant success in pansharpening, existing methods often face limitations in their evaluation, focusing on restricted satellite data sources, single scene types, and low-resolution images. This paper addresses this gap by introducing PanBench, a high-resolution multi-scene dataset containing all mainstream satellites and comprising 5898 pairs of samples. Each pair includes a four-channel (RGB + near-infrared) multispectral image of 256 × 256 pixels and a mono-channel panchromatic image of 1024 × 1024 pixels. To avoid irreversible loss of spectral information and achieve a high-fidelity synthesis, we propose a Cascaded Multiscale Fusion Network (CMFNet) for pansharpening. Multispectral images are progressively upsampled while panchromatic images are downsampled. Corresponding multispectral features and panchromatic features at the same scale are then fused in a cascaded manner to obtain more robust features. Extensive experiments validate the effectiveness of CMFNet.

Application Of SMOTE To Address Class Imbalance In Diabetes Disease Classification Utilizing C5.0, Random Forest, And SVM

The implementation of SMOTE to tackle class imbalance in classification frequently results in suboptimal outcomes, owing to the intricacy of the dataset and the multitude of attributes at play. Consequently, alternative classification models were explored through experimentation to gauge their precision. This research aims to compare the precision of C5.0, Random Forest, and SVM classification models both with and without SMOTE. The methodology encompasses dataset selection, an overview of classification algorithms (C5.0, Random Forest, SVM), SMOTE technique, validation via split validation, preprocessing involving min-max normalization, and execution evaluation utilizing confusion matrices and AUC analysis. The dataset was sourced by Kaggle, specifically to rectify class imbalance in a diabetes dataset using SMOTE, consisting of 768 instances, with 268 samples for diabetic cases and 500 samples for non-diabetic cases. Prior to SMOTE application, the classification precision for C5.0, Random Forest, and SVM were 0.714, 0.733, and 0.746 respectively, with corresponding AUC values of 0.745, 0.824, and 0.799. Post-SMOTE, the precision depicts for the same techniques were 0.603, 0.727, and 0.727, with AUC values of 0.734, 0.831, and 0.794 respectively. It can be inferred that there's minimal impact post-SMOTE across the three classification models due to potential overfitting on the dataset, leading to excessive reliance on synthesized data for minority classes, resulting in diminished model execution, precision, and AUC scores.

Utilizing multitemporal indices and spectral bands of Sentinel-2 to enhance land use and land cover classification with random forest and support vector machine

Multitemporal imagery offers a critical advantage by capturing seasonal variations, which are essential for differentiating between land use and land cover (LULC) types. While these types may appear similar when examined at one specific time, they exhibit distinct phenological patterns across different seasons. This temporal depth is crucial for enhancing model accuracy, particularly in heterogeneous landscapes where LULC transitions are frequent and complex. This paper made use of spectral bands and indices of Sentinel-2 from April to September 2020 for LULC classification using two advanced machine learning models: Random forest (RF) and support vector machine (SVM). The spectral indices include the normalized difference vegetation index (NDVI), normalized difference built-up index (NDBI), and modified normalized water index (MNDWI). The dataset was divided into four temporal feature sets: April-May, June-July, August-September, and the entire period from April-September. For each two-month period, the median values of the spectral bands and indices were used. Both models were evaluated based on overall accuracy, F1-score, Kappa coefficient, precision, and recall. Results indicate that incorporating multitemporal features enhanced the performance of the chosen models, with overall accuracy increasing from 82.4% to 94.03% for RF and from 75.4% to 93.54% for SVM. Additionally, the RF algorithm demonstrated higher accuracy than the SVM model across various time periods, with notable improvements in other performance metrics. These improvements also underscore the ability of the models to leverage the rich multitemporal data provided by Sentinel-2 for accurate LULC classification. This study highlights the importance of incorporating the dynamics of features in remote sensing applications to enhance the precision and reliability of LULC classification.

Skin cancer classification using EfficientNet architecture

Skin cancer is one of the most common deadly diseases worldwide. Hence, skin cancer classification is becoming increasingly important because treatment in the early stages of skin cancer is much more effective and efficient. This study focuses on the classification of three common types of skin cancer, namely basal cell carcinoma (BCC), squamous cell carcinoma (SCC), and melanoma using EfficientNet architecture. The dataset is preprocessed and each image in the dataset is resized to 256×256 pixels prior to incorporation in later stages. We then train all types of EfficientNet starting from EfficientNet-B0 to EfficientNet-B7 and compare their performances. Based on the test results, all trained EfficientNet models are capable of producing good accuracy, precision, recall, and F1-score in skin cancer classification. Particularly, our designed EfficientNet-B4 model achieves 79.69% accuracy, 81.67% precision, 76.56% recall, and 79.03% F1-score as the highest among others. These results confirm that EfficientNet architecture can be utilized to classify skin cancer properly.

Machine Learning Models for Predicting Risky Pregnancies in Early Clinical Interventions

Background: Risky pregnancies present significant challenges in maternal healthcare, often requiring accurate prediction to prevent adverse outcomes. Machine learning (ML) models offer a promising approach for predicting such risks, enabling timely interventions. This study evaluates five machine learning models—Logistic Regression, Decision Tree, K-Nearest Neighbors (KNN), Naive Bayes, and Support Vector Machine (SVM)—for their effectiveness in predicting risky pregnancies using clinical datasets. Methods: The study developed and evaluated five ML models, each implemented using Python’s scikit-learn library. The dataset was split into 75% for training and 25% for testing. Standard classification metrics, including accuracy, precision, recall, and F1-score, were used to assess model performance. Hyperparameter tuning was conducted using grid search and cross-validation to optimize model parameters. The models’ performance was compared to identify the most suitable for clinical applications. Results: The Decision Tree model achieved the highest accuracy (100% on training data, 95.6% on testing data), along with excellent precision, recall, and F1-scores for both classes, making it the most accurate and interpretable model for predicting risky pregnancies. Logistic Regression also performed well, particularly in identifying high-risk cases, with testing accuracy of 82%. KNN and SVM provided moderate accuracy, with KNN achieving 78% testing accuracy and SVM 80%. Naive Bayes, however, performed poorly, achieving only 43.2% accuracy due to its assumption of feature independence, which was not suitable for the dataset. Conclusion: The Decision Tree and Logistic Regression models emerged as the most effective for predicting risky pregnancies, offering high accuracy and interpretability, crucial for clinical decision-making.

Efficient multi-level cross-modal fusion and detection network for infrared and visible image

With the rapid development of uncrewed aerial vehicle (UAV) technology, detecting aerial images has found significant applications across various domains. However, existing algorithms overlook the impact of illumination on target detection, resulting in less satisfactory detection performance under low-light conditions. We propose EfficientFuseDet, a visible and infrared image fusion detection network to overcome this issue. First, an effective multilevel cross-modal fusion network called EfficientFuse is presented to combine complementary information from both modalities better. EfficientFuse captures local dependencies and global contextual information in shallow and deep layers, seamlessly combining complimentary local and global features throughout the network. The generated fused images can exhibit clear target contours and abundant texture information. Second, we propose a detection network called AFI-YOLO, which employs an inverted residual vision transformer backbone (IRViT) to effectively address the challenges associated with background interference in fused images. We design an efficient feature pyramid network (EFPN) that efficiently integrates multiscale information, enhancing multiscale detection capability using aerial images. A reparameterization decoupling head (RepHead) is proposed to further improve target classification and localization precision. Finally, experiments on the DroneVehicle dataset indicate that the detection accuracy using fused images can reach 47.2 %, which is higher than that observed with visible light images of 45 %. Compared to state-of-the-art detection algorithms, EfficientFuseDet exhibits a slight decrease in speed. However, it demonstrates superior detection capabilities and effectively enhances the detection accuracy using aerial images under low-light conditions.

Impact of database choice and confidence score on the performance of taxonomic classification using Kraken2

AbstractAccurate taxonomic classification is essential to understanding microbial diversity and function through metagenomic sequencing. However, this task is complicated by the vast variety of microbial genomes and the computational limitations of bioinformatics tools. The aim of this study was to evaluate the impact of reference database selection and confidence score (CS) settings on the performance of Kraken2, a widely used k-mer-based metagenomic classifier. In this study, we generated simulated metagenomic datasets to systematically evaluate how the choice of reference databases, from the compact Minikraken v1 to the expansive nt- and GTDB r202, and different CS (from 0 to 1.0) affect the key performance metrics of Kraken2. These metrics include classification rate, precision, recall, F1 score, and accuracy of true versus calculated bacterial abundance estimation. Our results show that higher CS, which increases the rigor of taxonomic classification by requiring greater k-mer agreement, generally decreases the classification rate. This effect is particularly pronounced for smaller databases such as Minikraken and Standard-16, where no reads could be classified when the CS was above 0.4. In contrast, for larger databases such as Standard, nt and GTDB r202, precision and F1 scores improved significantly with increasing CS, highlighting their robustness to stringent conditions. Recovery rates were mostly stable, indicating consistent detection of species under different CS settings. Crucially, the results show that a comprehensive reference database combined with a moderate CS (0.2 or 0.4) significantly improves classification accuracy and sensitivity. This finding underscores the need for careful selection of database and CS parameters tailored to specific scientific questions and available computational resources to optimize the results of metagenomic analyses.

Attention-enhanced dilated convolution for Parkinson’s disease detection using transcranial sonography

BackgroundTranscranial sonography (TCS) plays a crucial role in diagnosing Parkinson's disease. However, the intricate nature of TCS pathological features, the lack of consistent diagnostic criteria, and the dependence on physicians' expertise can hinder accurate diagnosis. Current TCS-based diagnostic methods, which rely on machine learning, often involve complex feature engineering and may struggle to capture deep image features. While deep learning offers advantages in image processing, it has not been tailored to address specific TCS and movement disorder considerations. Consequently, there is a scarcity of research on deep learning algorithms for TCS-based PD diagnosis.MethodsThis study introduces a deep learning residual network model, augmented with attention mechanisms and multi-scale feature extraction, termed AMSNet, to assist in accurate diagnosis. Initially, a multi-scale feature extraction module is implemented to robustly handle the irregular morphological features and significant area information present in TCS images. This module effectively mitigates the effects of artifacts and noise. When combined with a convolutional attention module, it enhances the model's ability to learn features of lesion areas. Subsequently, a residual network architecture, integrated with channel attention, is utilized to capture hierarchical and detailed textures within the images, further enhancing the model's feature representation capabilities.ResultsThe study compiled TCS images and personal data from 1109 participants. Experiments conducted on this dataset demonstrated that AMSNet achieved remarkable classification accuracy (92.79%), precision (95.42%), and specificity (93.1%). It surpassed the performance of previously employed machine learning algorithms in this domain, as well as current general-purpose deep learning models.ConclusionThe AMSNet proposed in this study deviates from traditional machine learning approaches that necessitate intricate feature engineering. It is capable of automatically extracting and learning deep pathological features, and has the capacity to comprehend and articulate complex data. This underscores the substantial potential of deep learning methods in the application of TCS images for the diagnosis of movement disorders.Graphical