Maximum Relevance Research Articles

Development and validation of a radiomics-based nomogram for predicting pathological grade of upper urinary tract urothelial carcinoma

BackgroundUpper urinary tract urothelial carcinoma (UTUC) is a rare and highly aggressive malignancy characterized by poor prognosis, making the accurate identification of high-grade (HG) UTUC essential for subsequent treatment strategies. This study aims to develop and validate a nomogram model using computed tomography urography (CTU) images to predict HG UTUC.MethodsA retrospective cohort study was conducted to include patients with UTUC who underwent radical nephroureterectomy and received a CTU examination prior to surgery. In the CTU images, tumor lesions located in the renal calyces, renal pelvis and ureter were segmented, and radiomics features from the unenhanced, medullary, and excretory phases were extracted. The maximum relevance minimum redundancy algorithm, least absolute shrinkage and selection operator, and various machine learning (ML) algorithms—including random forest, support vector machine, and eXtreme gradient boosting—were employed to select radiomics features and calculate radiomics scores. Logistic regression (LR) analysis was performed to identify the independent influencing factors of clinical baseline characteristics. Multiple datasets of radiomics features were constructed by integrating single-phase radiomics features with the most significant independent factor. Both LR and ML algorithms were utilized to develop predictive models. The area under the receiver operating characteristic curve (AUC values), accuracy, sensitivity, and specificity were assessed for model performance evaluation. Decision curve analysis was conducted to evaluate the clinical net benefits.ResultsA total of 167 patients were enrolled in this study. Among them, 56 were diagnosed with low-grade UTUC (including papillary urothelial neoplasms with low malignant potential and low-grade urothelial carcinoma) as confirmed by postoperative pathological examination results, and 111 were of HG. These patients were randomly allocated to the training set and the validation set at a ratio of 7:3. The training set comprised 116 patients with a mean age of 63.5 ± 9.38 years and 38 males. The validation set comprised 51 patients with a mean age of 65.6 ± 11.1 years and 18 males. Hydronephrosis was identified as the most significant independent factor in the clinical baseline features. Models that include mixed-phase development achieve better performance compared to models that rely simply on single-phase development. The nomogram model had excellent predictive ability for HG UTUC, with AUC values of 0.844 and an accuracy of 0.793 in the validation sets. The nomogram model can enhance accuracy by 14.1% (79.3% vs. 65.2%) and sensitivity by 32.8% (93.2% vs. 60.4%) compared to urinary cytology.ConclusionsThis study developed a nomogram model, which significantly improved the diagnostic ability for HG UTUC compared to urinary cytology. Furthermore, the results of the decision curve analysis showed that the model had a net benefit and could provide a non-invasive and potentially diagnostic reference tool for HG UTUC.

Enhanced Sequence-to-Sequence Attention-Based PM2.5 Concentration Forecasting Using Spatiotemporal Data

Severe air pollution problems continue to increase because of accelerated industrialization and urbanization. Specifically, fine particulate matter (PM2.5) causes respiratory and cardiovascular diseases, and according to the World Health Organization (WHO), millions of premature deaths and significant health burdens annually. Therefore, PM2.5 concentration forecasting is essential. This study proposed a method to forecast PM2.5 concentrations one hour after using Sequence-to-Sequence Attention (Seq2Seq-attention). The proposed method selects neighboring stations using minimum redundancy maximum relevance (mRMR) and integrates their data using a convolutional neural network (CNN). The proposed attention score and Seq2Seq are used on the integrated data to forecast PM2.5 concentration after one hour. The performance of the proposed method is validated through two case studies. The first comparison evaluated the performance of the conventional attention score against the proposed attention scores. The second comparison evaluated the forecasting results with and without considering neighboring stations. The first study showed that the proposed attention score improved the performance index (Root Mean Square Error (RMSE): 3.48%p, Mean Absolute Error (MAE): 8.60%p, R2: 0.49%p, relative Root Mean Square Error (rRMSE): 3.64%p, Percent Bias (PBIAS): 59.29%p). The second case study showed that considering neighboring stations’ data can be more effective in forecasting than considering that of a standalone station (RMSE: 5.49%p, MAE: 0.51%p, R2: 0.67%p, rRMSE: 5.44%p, PBIAS: 46.56%p). This confirmed that the proposed method can effectively forecast the PM2.5 concentration after one hour.

Robust audio copy-move detection and localisation based on Smoothed Pseudo Wigner-Ville Distribution and adaptive processing

Audio Copy-Move forgery is a copy and paste operation within the same audio source that alters the original audio content in order to deceive or mislead listeners. Existing audio copy-move forgery detection algorithms show good robustness against audio that has undergone a single post-processing (noise addition, compression, median filtering, etc.), but perform poorly in detecting copy-move that have undergone a mixed post-processing attack, where multiple post-processes are present in the audio simultaneously. To overcome this problem, we propose a new adaptive processing method based on Variational Modal Decomposition and Minimum Redundancy Maximum Relevance. First, the forged audio is decomposed into multiple Intrinsic Mode Functions located in different frequency domains. Scores are then computed based on correlation and redundancy. The calculated scores are used as weights and reconstructed by weighting to obtain the final audio signal. Through feature selection and weighted reconstruction, the new audio signal is provided with adaptive attributes. We chose the Smoothed Pseudo Wigner-Weil Distribution as the feature for performing the final copy-move detection. For the experimental part, we created hybrid post-processing datasets and applied adaptive processing to existing representative copy-move detection methods, all of which have improved detection results. Finally, we tested the robustness of the proposed method and the performance is optimal in comparison with other methods. The experimental results show that the proposed method is robust to both single and hybrid post-processing attacks.

Enhanced Intrusion Detection in Software-Defined Networking using Advanced Feature Selection: The EMRMR Approach

Most traditional IP networks face serious security and management challenges due to their rapid increase in complexity. SDN resolves these issues by the separation of control and data planes, hence enabling programmability for centralized management with flexibility. On the other hand, its centralized architecture makes SDN very prone to DDoS attacks, hence necessitating the use of advanced and efficient IDSs. This study focuses on improving IDS performance in SDN environments through the integration of deep learning techniques and novel feature selection methods. This study presents an Enhanced Maximum Relevance Minimum Redundancy (EMRMR) approach that incorporates a Mutual Information Feature Selection (MIFS) strategy and a new Contextual Redundancy Coefficient Upweighting (CRCU) strategy to optimize feature selection for early attack detection. Experiments on the inSDN dataset showed that EMRMR achieved better precision, recall, F1-score, and accuracy compared to the state-of-the-art approaches, especially when fewer features are selected. These results highlight the efficiency of the proposed EMRMR approach in the selection of relevant features with minimal computational overhead, which enhances the real-time capability for IDS in SDN environments.

Classification of cyclic alternating patterns of sleep using EEG signals

Cyclic alternating patterns (CAP) occur in electroencephalogram (EEG) signals during non-rapid eye movement sleep. The analysis of CAP can offer insights into various sleep disorders. The first step is the identification of phases A and B for the CAP cycles. In this work, we develop an easy-to-implement accurate system to differentiate between CAP A and CAP B. Small segments of the EEG signal are processed using Gaussian filters to obtain sub-band components. Features are extracted using some statistical characteristics of these signal components. Minimum redundancy maximum relevance test is employed to identify the more significant features. Three different machine learning classifiers are considered and their performance is compared. The results are analyzed for both the balanced and unbalanced datasets. The k-nearest neighbour (kNN) classifier achieves 79.14% accuracy and F-1 score of 79.24% for the balanced dataset. The proposed method outperforms the existing methods for CAP classification. It is easy-to-implement and can be considered as a candidate for real-time deployment.

PSR-LeafNet: A Deep Learning Framework for Identifying Medicinal Plant Leaves Using Support Vector Machines

In computer vision, recognizing plant pictures has emerged as a multidisciplinary area of interest. In the last several years, much research has been conducted to determine the type of plant in each image automatically. The challenges in identifying the medicinal plants are due to the changes in the effects of image light, stance, and orientation. Further, it is difficult to identify the medicinal plants due to factors like variations in leaf shape with age and changing leaf color in response to varying weather conditions. The proposed work uses machine learning techniques and deep neural networks to choose appropriate leaf features to determine if the leaf is a medicinal or non-medicinal plant. This study presents a neural network design based on PSR-LeafNet (PSR-LN). PSR-LeafNet is a single network that combines the P-Net, S-Net, and R-Net, all intended for leaf feature extraction using the minimum redundancy maximum relevance (MRMR) approach. The PSR-LN helps obtain the shape features, color features, venation of the leaf, and textural features. A support vector machine (SVM) is applied to the output achieved from the PSR network, which helps classify the name of the plant. The model design is named PSR-LN-SVM. The advantage of the designed model is that it suits more considerable dataset processing and provides better results than traditional neural network models. The methodology utilized in the work achieves an accuracy of 97.12% for the MalayaKew dataset, 98.10% for the IMP dataset, and 95.88% for the Flavia dataset. The proposed models surpass all the existing models, having an improvement in accuracy. These outcomes demonstrate that the suggested method is successful in accurately recognizing the leaves of medicinal plants, paving the way for more advanced uses in plant taxonomy and medicine.

Interpretable machine learning model based on CT semantic features and radiomics features to preoperatively predict Ki-67 expression in gastrointestinal stromal tumors

To develop and validate a machine learning (ML) model which combined computed tomography (CT) semantic and radiomics features to preoperatively predict Ki-67 expression in gastrointestinal stromal tumors (GISTs) patients. We retrospectively collected the clinical, imaging and pathological data of 149 GISTs patients. We randomly assigned the patients in a ratio of 7:3 to a training set (104 cases) and a validation (45 cases) set. We divided the patients into low and high Ki-67 expression group according to postoperative pathology. CT semantic features were analyzed from preoperative enhancement CT images and radiomics features were extracted from venous phase-enhanced images. We used intraclass correlation coefficient, maximal relevance and minimal redundancy and least absolute shrinkage and selection operator method to screen radiomics features and build radiomics label. 6 ML models were used for model construction. Receiver operating characteristic curves were used to evaluate the predictive efficiency of ML models. SHAP analysis was used to explain the contribution of different variables and their risk threshold. AUC of radscores in predicting Ki-67 expression of GIST patients were 0.749 and 0.729 in training and validation set. Among the 6 ML models, SVM exhibited best prediction accuracy. AUC of SVM model in predicting Ki-67 expression of GIST patients were 0.840, 0.767 and 0.832 in training, validation and test set. SHAP analysis showed that radscores and tumor diameter had highly positive contribution to the model. Therefore, the interpretable SVM model can predict Ki-67 expression of GISTs patients individually before surgery, which can provide reliable imaging biomarkers for clinical treatment decisions.

The value of radiomics based on 2-[18 F]FDG PET/CT in predicting WHO/ISUP grade of clear cell renal cell carcinoma

BackgroundThe aim is to develop and validate radiomics based on 2-[18F]fluoro-D-glucose positron emission tomography/computed tomography (2-[18F]FDG PET/CT) parameters for predicting the World Health Organization/International Society of Urological Pathology (WHO/ISUP) grade of clear cell renal cell carcinoma (ccRCC).MethodsA total of 209 patients with 214 lesions, who underwent 2-[18F]FDG PET/CT scans between December 2016 to December 2023, were included in our study. All ccRCC lesions were categorized into low grade (WHO/ISUP grade I-II) and high grade (WHO/ISUP grade III-IV). The lesions were allocated into a training group and a testing group in a ratio of 7:3. The radiomics features were extracted by a serious of maximum standardized uptake value (SUVmax) thresholds (0,2.5%,25%,40%) with the utilization of the minimum redundancy and maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) regression algorithm. The clinical, radiomics and combined models were constructed. The receiver operating characteristic (ROC) curve, decision curve and calibration curves were plotted to assess the predicting performance.ResultsThe area under curve (AUC) of PET-0, PET-2.5%, PET-25%, PET-40% model in the training group were 0.881(95% CI: 0.822–0.940),0.883(95% CI: 0.825–0.942),0.889(95% CI: 0.831–0.946),0.887(95% CI: 0.826–0.948); and 0.878(95% CI: 0.777–0.978),0.876(95% CI: 0.776–0.977),0.871(95% CI: 0.769–0.972),0.882(95% CI: 0.786–0.979) in the testing group. Due to perfect prediction and verification performance, the volume of interest (VOI) from PET images with SUVmax threshold of 40% were selected to construct the radiomics model and combined model. The AUC of the clinical model and radiomics model was 0.859 (sensitivity = 0.846, specificity = 0.747) and 0.909 (sensitivity = 0.808, specificity = 0.751) in the training group, respectively; 0.882 (sensitivity = 0.857, specificity = 0.857) and 0.901 (sensitivity = 0.905, specificity = 0.833) in the testing group, respectively. In combined models, the AUC was 0.916, the sensitivity was 0.923 and the specificity was 0.808 in the training group; the AUC was 0.916, the sensitivity was 0.881 and the specificity was 0.792 in the training group.ConclusionRadiomics based on 2-[18F]FDG PET/CT can be helpful to predict WHO/ISUP grade of ccRCC.

Classification of Asphalt Pavement Defects for Sustainable Road Development Using a Novel Hybrid Technology Based on Clustering Deep Features

In the rapid development of urbanization, the sustained and healthy development of transportation infrastructure has become a widely discussed topic. The inspection and maintenance of asphalt pavements not only concern road safety and efficiency but also directly impact the rational allocation of resources and environmental sustainability. To address the challenges of modern transportation infrastructure management, this study innovatively proposes a hybrid learning model that integrates deep convolutional neural networks (DCNNs) and support vector machines (SVMs). Specifically, the model initially employs a ShuffleNet architecture to autonomously extract abstract features from various defect categories. Subsequently, the Maximum Relevance Minimum Redundancy (MRMR) method is utilized to select the top 25% of features with the highest relevance and minimal redundancy. After that, SVMs equipped with diverse kernel functions are deployed to perform training and prediction based on the selected features. The experimental results reveal that the model attains a high classification accuracy of 94.62% on a self-constructed asphalt pavement image dataset. This technology not only significantly improves the accuracy and efficiency of pavement inspection but also effectively reduces traffic congestion and incremental carbon emissions caused by pavement distress, thereby alleviating environmental burdens. It is of great significance for enhancing pavement maintenance efficiency, conserving resource consumption, mitigating environmental pollution, and promoting sustainable socio-economic development.

Multimodal AI/ML for discovering novel biomarkers and predicting disease using multi-omics profiles of patients with cardiovascular diseases

Cardiovascular diseases (CVDs) are complex, multifactorial conditions that require personalized assessment and treatment. Advancements in multi-omics technologies, namely RNA sequencing and whole-genome sequencing, have provided translational researchers with a comprehensive view of the human genome. The efficient synthesis and analysis of this data through integrated approach that characterizes genetic variants alongside expression patterns linked to emerging phenotypes, can reveal novel biomarkers and enable the segmentation of patient populations based on personalized risk factors. In this study, we present a cutting-edge methodology rooted in the integration of traditional bioinformatics, classical statistics, and multimodal machine learning techniques. Our approach has the potential to uncover the intricate mechanisms underlying CVD, enabling patient-specific risk and response profiling. We sourced transcriptomic expression data and single nucleotide polymorphisms (SNPs) from both CVD patients and healthy controls. By integrating these multi-omics datasets with clinical demographic information, we generated patient-specific profiles. Utilizing a robust feature selection approach, we identified a signature of 27 transcriptomic features and SNPs that are effective predictors of CVD. Differential expression analysis, combined with minimum redundancy maximum relevance feature selection, highlighted biomarkers that explain the disease phenotype. This approach prioritizes both biological relevance and efficiency in machine learning. We employed Combination Annotation Dependent Depletion scores and allele frequencies to identify variants with pathogenic characteristics in CVD patients. Classification models trained on this signature demonstrated high-accuracy predictions for CVD. The best performing of these models was an XGBoost classifier optimized via Bayesian hyperparameter tuning, which was able to correctly classify all patients in our test dataset. Using SHapley Additive exPlanations, we created risk assessments for patients, offering further contextualization of these predictions in a clinical setting. Across the cohort, RPL36AP37 and HBA1 were scored as the most important biomarkers for predicting CVDs. A comprehensive literature review revealed that a substantial portion of the diagnostic biomarkers identified have previously been associated with CVD. The framework we propose in this study is unbiased and generalizable to other diseases and disorders.

Development of New Electricity System Marginal Price Forecasting Models Using Statistical and Artificial Intelligence Methods

The System Marginal Price (SMP) is the cost of the last unit of electricity supplied to the grid, reflecting the supply–demand equilibrium and serving as a key indicator of market conditions. Accurate SMP forecasting is essential for ensuring market stability and economic efficiency. This study addresses the challenges of SMP prediction in Turkey by proposing a comprehensive forecasting framework that integrates machine learning, deep learning, and statistical models. Advanced feature selection techniques, such as Minimum Redundancy Maximum Relevance (mRMR) and Maximum Likelihood Feature Selector (MLFS), are employed to refine model inputs. The framework incorporates time series methods like Multilayer Perceptron (MLP), Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), and Convolutional LSTM (ConvLSTM) to capture complex temporal patterns, alongside models such as Support Vector Machine (SVM), Extreme Gradient Boosting (XGBoost), and Extreme Learning Machine (ELM) for modeling non-linear relationships. Model performance was evaluated using the Mean Absolute Percentage Error (MAPE) across regular weekdays, weekends, and public holidays. XGBoost combined with MLFS consistently achieved the lowest MAPE values, demonstrating exceptional accuracy and robustness. Among all of the models, XGBoost combined with MLFS consistently achieved the lowest MAPE values, demonstrating superior accuracy and robustness. The results highlight the inadequacy of traditional models like ARIMA and SARIMA in capturing non-linear and highly volatile patterns, reinforcing the necessity of using advanced techniques for effective SMP forecasting. Overall, this study presents a novel and comprehensive approach tailored for complex electricity markets, significantly enhancing predictive reliability by incorporating economic indicators and sophisticated feature selection methods.

Preoperative Prediction of Lymph Node Metastases in Nonfunctional Pancreatic Neuroendocrine Tumors Using a Combined CT Radiomics-Clinical Model.

PanNETs are a rare group of pancreatic tumors that display heterogeneous histopathological and clinical behavior. Nodal disease has been established as one of the strongest predictors of patient outcomes in PanNETs. Lack of accurate preoperative assessment of nodal disease is a major limitation in the management of these patients, in particular those with small (< 2cm) low-grade tumors. The aim of the study was to evaluate the ability of radiomic features (RF) to preoperatively predict the presence of nodal disease in pancreatic neuroendocrine tumors (PanNETs). An institutional database was used to identify patients with nonfunctional PanNETs undergoing resection. Pancreas protocol computed tomography was obtained, manually segmented, and RF were extracted. These were analyzed using the minimum redundancy maximum relevance analysis for hierarchical feature selection. Youden index was used to identify the optimal cutoff for predicting nodal disease. A random forest prediction model was trained using RF and clinicopathological characteristics and validated internally. Of the 320 patients included in the study, 92 (28.8%) had nodal disease based on histopathological assessment of the surgical specimen. A radiomic signature based on ten selected RF was developed. Clinicopathological characteristics predictive of nodal disease included tumor grade and size. Upon internal validation the combined radiomics and clinical feature model demonstrated adequate performance (AUC 0.80) in identifying nodal disease. The model accurately identified nodal disease in 85% of patients with small tumors (< 2cm). Non-invasive preoperative assessment of nodal disease using RF and clinicopathological characteristics is feasible.

High-speed rail passenger flow prediction based on crossformer and quantile regression: A deep learning approach assisted by internet search indices

Understanding potential fluctuations in passenger volume is crucial for the operation and management of high-speed rail, especially during peak times. The uncertainty caused by multiple factors is the main obstacle to accurate prediction. To quantify and mitigate the impact of these uncertainties, Internet search indices are utilized as insightful resources to grasp dynamic trends in passenger flow. Leveraging minimum redundancy maximum relevance, we identify the top search index features based on their predictive contribution to high-speed rail passenger flow. A two-level decomposition strategy is then established based on variational modal decomposition to extract significant influencing factors hidden in the Internet index and capture the dynamic uncertainty of passenger flow. By integrating Crossformer with quantile regression, we construct the upper and lower bounds of the prediction interval. Furthermore, the obtained upper and lower bounds are corrected by the error of point prediction, which allows for dynamic adjustment of the prediction intervals width based on fluctuations in uncertainty, thereby refining the precision of the prediction interval. Finally, the developed approaches effectiveness is validated through two real-world experiments, and the experimental results indicate that this method can more accurately capture variations in high-speed rail passenger flow, improving both management and service quality.

Diagnostic precision in thyroid-associated ophthalmopathy using multi-center radiomics with 99mTc-DTPA SPECT/CT

To explore the performance of 99mTc-diethylene triaminepentaacetic acid (DTPA) SPECT/CT texture analysis in evaluating the activity of thyroid-associated ophthalmopathy (TAO) . This retrospective study examined 115 TAO patients from a single institution as an internal cohort and 58 TAO patients from another institution as an external validation set. Patients in the internal cohort were randomly divided into training (n = 81) and internal validation sets (n = 34). Radiomics signatures were constructed with the minimal redundancy maximal relevance and least absolute shrinkage and selection operator algorithms in training set. Multivariate logistic regression analysis was used to develop a clinical model and a combined clinical–radiomics model. Diagnostic performance of models was evaluated using receiver operating characteristic curve analysis, calibration curves and decision curve analysis. Compared with CT and SPECT radiomics models, Rad-scoreSPECT/CT demonstrated the best performance with areas under the receiver operating characteristic curve of 0.94 and 0.91 in the training and test sets, respectively. The combined clinical-radiomics model exhibited significantly better performance in evaluating TAO activity. Our results demonstrate the validity of a multimodal radiomic model of 99mTc-DTPA-SPECT/CT to assess TAO activity. The combined clinical-radiomics model exhibited significantly better diagnostic performance than the clinical model.

A computed tomography‑based radio‑clinical model for the prediction of microvascular invasion in gastric cancer.

The objective of the present study was to build and validate a radio-clinical model integrating radiological features and clinical characteristics based on information available before surgery for prediction of microvascular invasion (MI) in gastric cancer. The retrospective study included a cohort of 534 patients (n=374 for the training set and n=160 for the test set) who were diagnosed with gastric cancer. All patients underwent contrast-enhanced computed tomography within one month before surgery. The focal area was mapped by ITK-SNAP. Radiomics features were extracted from portal venous phase CT images. Principal component analysis was used to reduce dimensionality, maximum relevance minimum redundancy, and least absolute shrinkage and selection operator to screen features most associated with MI. The radiomics signature was subsequently computed based on the coefficient weight assigned to it. The independent risk factors for MI of gastric cancer were determined using univariate analysis and multivariate logistic regression analysis. Univariate logistic regression analysis was used to assess the association between clinical characteristics and MI status. A radio-clinical model was constructed by employing multi-variable logistic regression analysis, incorporating radiomic features with clinical characteristics. Receiver operating characteristic (ROC) curve analysis, decision curve analysis (DCA) and calibration curves were employed for the analysis and evaluation of the model's performance. The radiomics signature model had moderate recognition ability, with an area under ROC curve (AUC) of 0.77 for the training set and 0.73 for the test set. The radio-clinical model, consisting of rad-score and clinical features, could well discriminate the training set and test set (AUC=0.88 and 0.80, respectively). The calibration curves and DCA further validated the favorable fit and clinical applicability of the radio-clinical model. In conclusion, the radio-clinical model combining the radiomics signature and clinical characteristics may be used to individually predict MI in gastric cancer to aid in the development of a clinical treatment strategy.

Wheat grain classification using hyperspectral imaging: Concatenating Vis-NIR and SWIR Data for single and bulk grains

Wheat variety identification is an essential component of the official inspection of wheat grains to determine their quality and trade value. Traditionally, wheat class identification requires deep knowledge about the appearance of plants and kernels, their history and distribution. Therefore, there is an urgent need for an automatic technology to standardize the nomenclature of wheat varieties to enable both growers and producers to identify their grains practically. The present study investigated the feasibility of Visible-Near Infrared (Vis-NIR) and Short-Wave Infrared (SWIR) spectral imaging to identify wheat classes. Both sides of the wheat kernel as single kernel measurements and bulk grains were used to acquire hyperspectral data and vertical and horizontal data concatenation was implemented to explore performance differences. In addition, the classification of bulk kernels from single kernel data was investigated. Spectral data was pre-treated by standard normal variate (SNV), Savitzky-Golay first (SG-1) and second derivatives (SG-2) and linear discriminant analysis (LDA), support vector machine (SVM) and artificial neural network (ANN) were applied as the classification algorithms. Furthermore, feature selection was utilised using the minimum redundancy maximum relevance (mRMR) algorithm to investigate the performance difference between data concatenation and feature selection. The results showed that ventral (up) side data showed better classification performances than reverse (down) side data, while Vis-NIR data achieved higher classification accuracies than SWIR data. However, the best classification performance for single kernels was obtained by LDA-SNV using up and down data in the Vis-NIR and SWIR regions vertically and horizontally concatenated with an accuracy of 93.72% for 10-fold cross-validation and 94.93% for test sets. Models based on a hundred features did not achieve the accuracy of models based on concatenated data. Moreover, classification performances of bulk samples were higher than single kernels, which achieved 100% accuracy for both cross-validation and test sets. The study demonstrates that spectral imaging has a high potential to identify wheat classes non-destructively.

A Feature-Weighted Support Vector Regression Machine Based on Hilbert–Schmidt Independence Criterion Least Absolute Shrinkage and Selection Operator

Support vector regression (SVR) is a powerful kernel-based regression prediction algorithm that performs excellently in various application scenarios. However, for real-world data, the general SVR often fails to achieve good predictive performance due to its inability to assess feature contribution accurately. Feature weighting is a suitable solution to address this issue, applying correlation measurement methods to obtain reasonable weights for features based on their contributions to the output. In this paper, based on the idea of a Hilbert–Schmidt independence criterion least absolute shrinkage and selection operator (HSIC LASSO) for selecting features with minimal redundancy and maximum relevance, we propose a novel feature-weighted SVR that considers the importance of features to the output and the redundancy between features. In this approach, the HSIC is utilized to effectively measure the correlation between features as well as that between features and the output. The feature weights are obtained by solving a LASSO regression problem. Compared to other feature weighting methods, our method takes much more comprehensive consideration of weight calculation, and the obtained weighted kernel function can lead to more precise predictions for unknown data. Comprehensive experiments on real datasets from the University of California Irvine (UCI) machine learning repository demonstrate the effectiveness of the proposed method.

Predictive Value of Multi-Phase CT Radiomics in Delineating Early and Advanced T Staging of Colon Cancer

This study leveraged the power of multi-phase CT radiomics, which extracts detailed features from multi-phase images, to explore the distinction between early (T1-T2) and advanced (T3-T4) colon cancer stages in each of multiple phases, providing valuable insights to healthcare professionals and enhancing their decision-making capabilities. A total of 191 patients with surgically confirmed primary colon cancer were retrospectively included, and multi-phase CT scans (non-enhanced contrast phase, arterial phase, portal venous phase, and delayed phase) were conducted within one week before surgery. Three-dimensional segmentation of colonic tumors was performed on the images of the four phases, and radiomics features of each colonic tumor were automatically extracted. Minimum redundancy maximum relevance (mRMR) was applied to features selection in each of the four phases. The least absolute shrinkage and selection operator logistic regression was conducted to determine the association between radiomics features and early (T1-T2) and advanced (T3-T4) stages of colon cancer. Additionally, diagnostic performance comparison was carried out in four phases. The mean (±SD) age of the 191 individuals was 61.87±13.137 years, with females comprising 43.5% of the cohort. The selected features for the non-enhanced, arterial, portal venous, and delayed phases numbered 7, 11, 6, and 10, respectively. In the test set, the AUC values for the on-enhanced, arterial, portal venous, and delayed phases were 0.86 (0.76-0.96), 0.84 (0.73-0.94), 0.82 (0.71-0.93), and 0.86 (0.75-0.97), with corresponding accuracies of 0.84, 0.80, 0.75, and 0.88, sensitivities of 0.73, 0.64, 0.86, and 0.68, and specificities of 0.91, 0.91, 0.68, and 1.00, respectively. DeLong's test revealed no statistically significant differences in the AUC values between the four phases within the test sets (P = 0.3233~ 0.9912). Multi-phase CT radiomics demonstrated substantial value in differentiating between early (T1-T2) and advanced (T3-T4) stages of patients with colon cancer.

Enhancing the accuracy of blockage detection in centrifugal pump using the majority voting classifier on an unbalanced dataset

Blockage in a centrifugal pump can adversely impact its performance. In this study, an experimental facility is developed to simulate three types of blockages (suction, discharge, and simultaneous suction and discharge). To classify these faults and detect their severity, a methodology involving the application of majority voting classifier (MVC) to the pump’s discharge pressure signals is presented. An unbalanced dataset is constructed, where the number of samples for a specific blockage condition decreases with increasing fault severity. Statistical features, entropy features, and entropies meta features are extracted from the signal and ranked using XGBoost and minimum redundancy maximum relevance (MRMR). Subsequently, the optimal features are selected based on the best performing model (Linear Discriminant Analysis) among ten different models. Best four models are selected and ensembled to form MVC. Results show that MVC achieves an accuracy of 89.90% and 88.26% for features selected by XGBoost and MRMR, respectively. Finally, the unbalanced dataset is balanced using synthetic minority oversampling and it is shown that MVC achieves an accuracy of 100% on this balanced dataset for the features selected using both approaches.

Improving Alzheimer's Disease Prediction with Different Machine Learning Approaches and Feature Selection Techniques.

Machine learning (ML) has increasingly been utilized in healthcare to facilitate disease diagnosis and prediction. This study focuses on predicting Alzheimer's disease (AD) through the development and comparison of ML models using Support Vector Machine (SVM), Random Forest (RF), and Logistic Regression (LR) algorithms. Additionally, feature selection techniques including Minimum Redundancy Maximum Relevance (mRMR) and Mutual Information (MI) were employed to enhance the model performance. The research methodology involved training and testing these models on the OASIS-2 dataset, evaluating their predictive accuracies. Notably, LR combined with mRMR achieved the highest accuracy of 99.08% in predicting AD. These findings underscore the efficacy of ML algorithms in AD prediction and highlight the utility of the feature selection methods in improving the model performance. This study contributes to the ongoing efforts to leverage ML for more accurate disease prognosis and underscores the potential of these techniques in advancing clinical decision-making.