Combined Model Research Articles

A hybrid model with combined feature selection based on optimized VMD and improved multi-objective coati optimization algorithm for short-term wind power prediction

With the continuous global increase in installed wind power capacity and subsequent surge in power generation, the contradiction between the safe operation of the grid and the efficient consumption of new energy after large-scale grid connection has become increasingly prominent. This paper presents a hybrid model prediction method to further improve the stability and accuracy of wind power prediction. Firstly, variational modal decomposition (VMD) optimized by the coati optimization algorithm (COA) is employed to decompose original wind power, mitigating the non-stationary characteristics of the power sequence. Subsequently, permutation entropy (PE) is used to recombine the decomposed components, and the combined feature selection method is achieved by integrating the Spearman correlation coefficient (SCC) and the autocorrelation function (ACF). Then, the multivariate combined model is constructed, and the improved multi-objective coati optimization algorithm (IMOCOA) determines the weight coefficients of each model to enhance the performance of the hybrid model. Finally, research and analysis are conducted from multiple scenarios and time scales using historical operating data from a wind farm in Xinjiang. The experimental results show that the proposed prediction model effectively improves the accuracy and stability of the wind power prediction compared with other popular prediction models.

Prognostication in Lymph Node-Positive Prostate Cancer with No PSA Persistence After Radical Prostatectomy.

This study aimed to create a prognostic model to predict disease recurrence among patients with lymph node involvement but no prostate-specific antigen (PSA) persistence and to explore its clinical utility. The study analyzed patients with lymph node involvement after pelvic lymph node dissection with radical prostatectomy in whom no PSA persistence was observed between 2006 and 2019 at 33 institutions. Prognostic factors for recurrence-free survival (RFS) were analyzed by the Cox proportional hazards model. Among 231 patients, 127 experienced disease recurrence. The factors prognostic for RFS were PSA level at diagnosis (≥ 20 vs. < 20ng/mL: hazard ratio [HR], 1.66; 95% confidence interval [CI], 1.09-2.52; P = 0.017), International Society of Urological Pathology grade group at radical prostatectomy (RP) specimen (group ≥ 4 vs. ≤ 3: HR, 1.63; 95% CI 1.12-2.37; P = 0.010), pathologic T-stage (pT3b/4 vs. pT2/3a: HR, 1.70; 95% CI 1.20-2.42; P = 0.0031), and surgical margin status (positive vs. negative: HR, 1.60; 95% CI 1.13-2.28; P = 0.0086). The prognostic model using four parameters were associated with RFS and metastasis-free survival. The prognostic model in combination with postoperative PSA value and number of lymph nodes is clinically useful for discussing treatment choice with patients.

Development of a multi-phase CT-based radiomics model to differentiate heterotopic pancreas from gastrointestinal stromal tumor

BackgroundTo investigate whether CT-based radiomics can effectively differentiate between heterotopic pancreas (HP) and gastrointestinal stromal tumor (GIST), and whether different resampling methods can affect the model’s performance.MethodsMulti-phase CT radiological data were retrospectively collected from 94 patients. Of these, 40 with HP and 54 with GISTs were enrolled between April 2017 and November 2021. One experienced radiologist manually delineated the volume of interest and then resampled the voxel size of the images to 0.5 × 0.5 × 0.5 mm3, 1 × 1 × 1 mm3, and 2 × 2 × 2 mm3, respectively. Radiomics features were extracted using PyRadiomics, resulting in 1218 features from each phase image. The datasets were randomly divided into training set (n = 66) and validation set (n = 28) at a 7:3 ratio. After applying multiple feature selection methods, the optimal features were screened. Radial basis kernel function-based support vector machine (RBF-SVM) was used as the classifier, and model performance was evaluated using the area under the receiver operating curve (AUC) analysis, as well as accuracy, sensitivity, and specificity.ResultsThe combined phase model performed better than the other phase models, and the resampling method of 0.5 × 0.5 × 0.5 mm3 achieved the highest performance with an AUC of 0.953 (0.881-1), accuracy of 0.929, sensitivity of 0.938, and specificity of 0.917 in the validation set. The Delong test showed no significant difference in AUCs among the three resampling methods, with p > 0.05.ConclusionsRadiomics can effectively differentiate between HP and GISTs on CT images, and the diagnostic performance of radiomics is minimally affected by different resampling methods.

Research on time series prediction of multi-process based on deep learning

Aiming at the problem of data fluctuation in multi-process production, a Soft Update Dueling Double Deep Q-learning (SU-D3QN) network combined with soft update strategy is proposed. Based on this, a time series combination forecasting model SU-D3QN-G is proposed. Firstly, based on production data, Gate Recurrent Unit (GRU) is used for prediction. Secondly, based on the model, SU-D3QN algorithm is used to learn and add bias to it, and the prediction results of GRU are corrected, so that the prediction value of each time node fits in the direction of reducing the absolute error. Thirdly, experiments were carried out on the dataset of a company. The data sets of four indicators, namely, the outlet temperature of drying silk, the loose moisture return water, the outlet temperature of feeding leaves and the inlet water of leaf silk warming and humidification, are selected, and more than 1000 real production data are divided into training set, inspection set and test set according to the ratio of 6:2:2. The experimental results show that the SU-D3QN-G combined time series prediction model has a great improvement compared with GRU, LSTM and ARIMA, and the MSE index is reduced by 0.846–23.930%, 5.132–36.920% and 10.606–70.714%, respectively. The RMSE index is reduced by 0.605–10.118%, 2.484–14.542% and 5.314–30.659%. The MAE index is reduced by 3.078–15.678%, 7.94–15.974% and 6.860–49.820%. The MAPE index is reduced by 3.098–15.700%, 7.98–16.395% and 7.143–50.000%.

Prediction of Microwave Ablation Recurrence in Pulmonary Malignancies Using Preoperative Computed Tomography Radiomics Models.

Assessing the early efficacy of microwave ablation (MWA) for pulmonary malignancies is a challenge for interventionalists. However, performing an accurate efficacy assessment at an earlier stage can significantly enhance clinical intervention and improve the patient's prognosis. This research aimed to create and assess non-invasive diagnostic techniques using pre-operative computed tomography (CT) radiomics models to predict the recurrence of MWA in pulmonary malignancies. We retrospectively enrolled 116 eligible patients with pulmonary malignancies treated with MWA. we separated the patients into two groups: a recurrence group (n = 28) and a non-recurrence group (n = 88), following the modified Response Evaluation Criteria in Solid Tumors (m-RECIST) criteria. We segmented the preoperative tumor area manually. We expanded outward the tumor boundary 4 times, with a width of 3 mm, using the tumor boundary as the baseline. Five groups of radiomics features were extracted and screened using max-relevance and min-redundancy (mRMR) and least absolute shrinkage and selection operator (LASSO) regression. Weight coefficients of the aforementioned features were used to calculate the Radscore and construct radiomics models for both tumoral and peritumoral areas. The Radscore from the radiomics model was combined with clinical risk factors to construct a combined model. The performance and clinical usefulness of the combined models were assessed through the evaluation of receiver operating characteristic (ROC) curves, the Delong test, calibration curves, and decision curve analysis (DCA) curves. The clinical risk factor for recurrence after MWA was tumor diameter (P < 0.05). Both tumoral and four peritumoral radiomics models exhibited high diagnostic efficacy. Furthermore, the combined 1 (C1)-RO model and the combined 2 (C2)-RO model showed higher efficacy with area under the curve (AUCs) of 0.89 and 0.89 in the training cohort, and 0.93 and 0.94 in the validation cohort, respectively. Both combined models demonstrated excellent predictive accuracy and clinical benefit. Preoperative CT radiomics models for both tumoral and peritumoral regions are capable of accurately predicting the recurrence of pulmonary malignancies after MWA. The combination of both models may lead to better performance and may aid in devising more effective preoperative treatment strategies.

Remotely sensed desertification modeling using ensemble of machine learning algorithms

Due to having a sensitive and fragile ecosystem, dry areas are constantly exposed to land degradation and desertification. Therefore, it is necessary to formulate appropriate strategies for quantitative assessment of desertification that are highly accurate. In this research, desertification of the region was first evaluated using MEDALUS model, then according to the results of MEDALUS model and reviewing the results of other researchers, 8 indicators remote sensing that had the highest correlation with field data were selected for modeling. Four machine learning methods Support Vector Machine (SVM), Gradient Boosting Machine (GBM), Generalized Linear Models (GLM) and Random Forests (RF) were used to model the risk of desertification in northeastern Iran. Finally, the weighted average of the ensemble model in the SDM statistical package was used to predict the desertification of the region. Based on the results obtained from MEDALUS model, the indicators of drought resistance (score 162), conservation operations (score 158) and soil salinity (score 155), in the working units of abandoned lands, wetland lands, and Salty lands located in the north East of the region, have increased the process of desertification. The results of modeling using machine learning methods showed that in 2002, the SVM model (AUC = 0.91, TSS = 0.93, and Kappa = 0.86) and in 2021, the RF model (AUC = 0.94, TSS = 0.94, and Kappa = 0.90) have performed best. The forecast of the combined model for desertification in 2021 in the studied area showed that the northeastern and sporadically in the central parts of the studied area are affected by the progress of the desertification process. Therefore, by considering the results of the combined model (as a model with the least uncertainty), it is possible to reduce the progress of the desertification process by planning, optimal management and applying corrective methods in the areas affected by desertification.

Multiphase MRI-Based Radiomics for Predicting Histological Grade of Hepatocellular Carcinoma.

Hepatocellular carcinoma (HCC) is a highly heterogeneous cancer. Accurate preoperative prediction of histological grade holds potential for improving clinical management and disease prognostication. To evaluate the performance of a radiomics signature based on multiphase MRI in assessing histological grade in solitary HCC. Retrospective. A total of 405 patients with histopathologically confirmed solitary HCC and with liver gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced MRI within 1 month of surgery. Contrast-enhanced T1-weighted spoiled gradient echo sequence (LAVA) at 1.5 or 3.0 T. Tumors were graded (low/high) according to results of histopathology. Basic clinical characteristics (including age, gender, serum alpha-fetoprotein (AFP) level, history of hepatitis B, and cirrhosis) were collected and tumor size measured. Radiomics features were extracted from Gd-EOB-DTPA-enhanced MRI data. Three feature selection strategies were employed sequentially to identify the optimal features: SelectFromModel (SFM), SelectPercentile (SP), and recursive feature elimination with cross-validation (RFECV). Probabilities of five single-phase radiomics-based models were averaged to generate a radiomics signature. A combined model was built by combining the radiomics signature and clinical predictors. Pearson χ2 test/Fisher exact test, Wilcoxon rank sum test, interclass correlation coefficient (ICC), univariable/multivariable logistic regression analysis, area under the receiver operating characteristic (ROC) curve (AUC), DeLong test, calibration curve, Brier score, decision curve, Kaplan-Meier curve, and log-rank test. A P-value <0.05 was considered statistically significant. High-grade HCCs were present in 33.8% of cases. AFP levels (odds ratio [OR] 1.89) and tumor size (>5 cm; OR 2.33) were significantly associated with HCC grade. The combined model had excellent performance in assessing HCC grade in the test dataset (AUC: 0.801), and demonstrated satisfactory calibration and clinical utility. A model that combined a radiomics signature derived from preoperative multiphase Gd-EOB-DTPA-enhanced MRI and clinical predictors showed good performance in assessing HCC grade. 3 TECHNICAL EFFICACY: Stage 5.

Multimodality radiomics for tumor prognosis in nasopharyngeal carcinoma.

The prognosis of nasopharyngeal carcinoma (NPC) is challenging due to late-stage identification and frequently undetectable Epstein-Barr virus (EBV) DNA. Incorporating radiomic features, which quantify tumor characteristics from imaging, may enhance prognosis assessment. To investigate the predictive power of radiomic features on overall survival (OS), progression-free survival (PFS), and distant metastasis-free survival (DMFS) in NPC. A retrospective analysis of 183 NPC patients treated with chemoradiotherapy from 2010 to 2019 was conducted. All patients were followed for at least three years. The pretreatment CT images with contrast medium, MR images (T1W and T2W), as well as gross tumor volume (GTV) contours, were used to extract radiomic features using PyRadiomics v.2.0. Robust and efficient radiomic features were chosen using the intraclass correlation test and univariate Cox proportional hazard regression analysis. They were then combined with clinical data including age, gender, tumor stage, and EBV DNA level for prognostic evaluation using Cox proportional hazard regression models with recursive feature elimination (RFE) and were optimized using 20 repetitions of a five-fold cross-validation scheme. Integrating radiomics with clinical data significantly enhanced the predictive power, yielding a C-index of 0.788 ± 0.066 to 0.848 ± 0.079 for the combined model versus 0.745 ± 0.082 to 0.766 ± 0.083 for clinical data alone (p<0.05). Multimodality radiomics combined with clinical data offered the highest performance. Despite the absence of EBV DNA, radiomics integration significantly improved survival predictions (C-index ranging from 0.770 ± 0.070 to 0.831 ± 0.083 in combined model versus 0.727 ± 0.084 to 0.734 ± 0.088 in clinical model, p<0.05). The combination of multimodality radiomic features from CT and MR images could offer superior predictive performance for OS, PFS, and DMFS compared to relying on conventional clinical data alone.

Ensemble learning based transmission line fault classification using phasor measurement unit (PMU) data with explainable AI (XAI).

A large volume of data is being captured through the Phasor Measurement Unit (PMU), which opens new opportunities and challenges to the study of transmission line faults. To be specific, the Phasor Measurement Unit (PMU) data represents many different states of the power networks. The states of the PMU device help to identify different types of transmission line faults. For a precise understanding of transmission line faults, only the parameters that contain voltage and current magnitude estimations are not sufficient. This requirement has been addressed by generating data with more parameters such as frequencies and phase angles utilizing the Phasor Measurement Unit (PMU) for data acquisition. The data has been generated through the simulation of a transmission line model on ePMU DSA tools and Matlab Simulink. Different machine learning models have been trained with the generated synthetic data to classify transmission line fault cases. The individual models including Decision Tree (DT), Random Forest (RF), and K-Nearest Neighbor (K-NN) have outperformed other models in fault classification which have acquired a cross-validation accuracy of 99.84%, 99.83%, and 99.76% respectively across 10 folds. Soft voting has been used to combine the performance of these best-performing models. Accordingly, the constructed ensemble model has acquired a cross-validation accuracy of 99.88% across 10 folds. The performance of the combined models in the ensemble learning process has been analyzed through explainable AI (XAI) which increases the interpretability of the input parameters in terms of making predictions. Consequently, the developed model has been evaluated with several performance matrices, such as precision, recall, and f1 score, and also tested on the IEEE 14 bus system. To sum up, this article has demonstrated the classification of six scenarios including no fault and fault cases from transmission lines with a significant number of training parameters and also interpreted the effect of each parameter to make predictions of different fault cases with great success.

LCEMILCP: Design of a Low-complexity Energy Harvesting Model Via Incremental Learning and Continuous Power Quality Optimization Process in Wsn

Optimization of energy harvesting requires design of low complexity &amp; high efficiency models that can work with maximum power gain levels. To design such models, researchers have proposed multiple techniques, that can assist in improving power quality via selection of optimum harvesting sources in multisource environments. But these models require continuous reconfiguration of static rules, which limits their efficiency when applied to large-scale network scenarios. Moreover, most of these models also showcase higher complexity due to reconfiguration, which reduces their scalability performance. To overcome this limitation, a novel Low-Complexity Energy Harvesting Model via Incremental Learning and Continuous Power Quality Optimization process is discussed in this text. The proposed model initially uses a Q-Learning based power evaluation method, that is capable of generating high-efficiency configurations of multisource harvesting devices. This is cascaded with design of a Particle Swarm Optimizer (PSO), that assists in performing continuous power quality optimizations. The combined model is capable of selecting hybrid harvesting source configurations, and incrementally tune it for optimum harvesting performance. This is achieved via modelling a reward function that incorporates power gain along with low-complexity source selection process. The selection process is further enhanced via PSO based continuous learning for improving harvesting source configurations. The proposed model was tested on a wide variety of network scenarios, and its QoS efficiency levels were compared with different state-of-the-art methods. Based on this comparison, it was observed that the proposed model is capable of improving power gain by 8.3%, while minimizing harvesting delay by 6.5%, and improving harvesting throughput by 5.9%, which makes it useful for large-scale multisource harvesting applications.

Pain and Cognitive Load As Predictors of Absenteeism in Industry

Purpose: This study delves into the impact of pain and cognitive load on industrial absenteeism, aiming to highlight the importance of these factors in workforce management and productivity enhancement. Theoretical framework: Anchored in the biopsychosocial model, this research contends that absenteeism is intricately linked to the interplay between pain and cognitive load. Utilizing the Numeric Rating Scale (NRS) and the NASA Task Load Index (NASA-TLX), it seeks to elucidate their combined influence on absenteeism. Method/design/approach: In a pilot study, NRS and the NASA-TLX were utilized for data collection, with a third-party dataset incorporating seven predictor variables and absenteeism as the dependent variable. Statistical analysis involved the Shapiro-Wilk Test for data normality and the Mann-Whitney U Test for non-normal distributions. Generalized Linear Models were applied to analyze the impact of variables on absenteeism, including multivariate, univariate and combined models. Model performance was evaluated using BIC, AIC, pseudo-R², LogLik, and AUC, with the Delong method for AUC comparison. Results and conclusion: Pain emerged as a significant absenteeism predictor. The integrated model displayed enhanced discriminatory capacity, underscoring the pronounced interrelation between pain, cognitive load, and absenteeism. These insights advocate for holistic management strategies encompassing both health and occupational aspects. Research implications: This study reaffirms the crucial role of pain and cognitive load in industrial absenteeism and highlights the importance of a multifaceted, data-driven approach for effective intervention strategies. Originality/value: This study offers unique perspectives on absenteeism by merging pain and cognitive load assessments, presenting a comprehensive method for understanding and mitigating its impact on the industrial workforce.

Risk evaluation of mega infrastructure construction supply chain in engineering-procurement-construction projects: an integrated fuzzy AHP and fuzzy DEMATEL approach

PurposeThe paper covers mega infrastructure construction supply chain (MICSC) in Engineering-Procurement-Construction (EPC) projects, where the frequent occurrence of risk incidents has greatly affected human life. The research aims to establish a risk evaluation index system for MICSC in EPC projects, exploring what risk factors lead to risk incidents and measure the importance and causality of all these risk factors.Design/methodology/approachThe research applies a combination of quantitative and qualitative analysis methodology to process data sequentially. In the first place, risk factors for MICSC in EPC projects are extracted and identified from literature survey and expert interviews. In the second place, an integration model fuzzy Analytic Hierarchy Process (f-AHP) and fuzzy Decision-making Trial and Evaluation Laboratory (f-DEMATEL) is constructed to comprehensively analyze all these risk factors.Findings12 primary risk factors and 36 secondary risk factors comprise the risk evaluation index system for MICSC in EPC projects from 178 literature and 5 professionals. The results indicate that Political Situation (F1), Social Security (F2) and Management Mode (F8) are critical risk factors, where F1 and F2 are cause factors and F8 is an effect factor.Originality/valueThere are three main contributions of this paper. First and foremost, from the perspective of the research content, no other study has been able to assess risk factors for MICSC in EPC projects, while embedding nine phases of the whole project life cycle and six subjects of stakeholders into a risk evaluation index system. Additionally, from the perspective of research method, a combined model incorporating f-AHP and f-DEMATEL is constructed to avoid the one-sidedness of a single model. Last but not least, from the perspective of practical significance, focusing on the critical risk factors, a series of effective measures are formulated to make appropriate management decisions for nodal enterprises of MICSC, which can improve their risk management capabilities.

Life events and life satisfaction: Estimating effects of multiple life events in combined models

How do life events affect life satisfaction? Previous studies focused on a single event or separate analyses of several events. However, life events are often grouped non-randomly over the lifespan, occur in close succession, and are causally linked, raising the question of how to best analyze them jointly. Here, we used representative German data (SOEP; N = 40,121 individuals; n = 41,402 event occurrences) to contrast three fixed-effects model specifications: First, individual event models in which other events were ignored, which are thus prone to undercontrol bias; second, combined event models which controlled for all events, including subsequent ones, which may induce overcontrol bias; and third, our favored combined models that only controlled for preceding events. In this preferred model, the events of new partner, cohabitation, marriage, and childbirth had positive effects on life satisfaction, while separation, unemployment, and death of partner or child had negative effects. Model specification made little difference for employment- and bereavement-related events. However, for events related to romantic relationships and childbearing, small but consistent differences arose between models. Thus, when estimating effects of new partners, separation, cohabitation, marriage, and childbirth, care should be taken to include appropriate controls (and omit inappropriate ones) to minimize bias.

A Deeper Insight into the Supramolecular Activation of Oxidative Addition Reactions Involving Pincer-Rhodium(I) Complexes.

The factors governing the acceleration of the oxidative addition of methyl iodide to pincer rhodium(I)-complexes induced by coronene have been computationally explored in detail using quantum chemical methods. Both the parent reaction and the coronene-mediated process proceed via a stepwise SN2-type mechanism. It is found that the acceleration of the process derives from the formation of an initial supramolecular complex, mainly stabilized by electrostatic and π-π interactions, which significantly increases the electron richness of the complex. The impact of this effect on the reaction barrier has been quantitatively analyzed by applying the activation strain model in combination with the energy decomposition analysis method. In addition, the influence of other polycyclic aromatic hydrocarbons on the oxidative reaction has been also considered.

Quantifying the Contribution of Soil Heavy Metals to Ecological and Health Risk Sources

Quantifying the risk of soil heavy metal sources can identify the main pollution sources. It can provide a scientific basis for reducing the ecological and human health risks of soil heavy metals. Taking the shallow soil in a Pb-Zn mine watershed in northern Guangxi as a research object， ecological and human health risk assessments were conducted using potential ecological risk assessment （RI） and human health risk assessment （HRA）， and the source apportionment of soil heavy metals was completed using the absolute principal component-multiple linear regression receptor （APCS-MLR） model and random forest （RF） model. Then， a combined risk assessment model， consisting of RI， HRA， and APCS-MLR， was used to quantify the risk of soil heavy metal sources. The results showed that the contents of Pb， Zn， Cu， and Cd exceeded the environmental screening values for agricultural land with mean values of 342.77， 693.34， 61.27， and 3.08 mg·kg-1， respectively， and there was a certain degree of contamination. Pb， Cr， and As were the main health risk impact factors， with higher health risks for children than for adults. Three sources were identified： mining activities （Source Ⅰ）， soil parent material sources and original formation （Source Ⅱ）， and unknown sources. Pb， Zn， Cu， and Cd were mainly derived from Source Ⅰ， and Cr and As were controlled by unknown sources and Source Ⅱ. The source risk assessment results of soil heavy metals indicated that the potential ecological risk and non-carcinogenic risk were mainly from Source Ⅰ and Source Ⅱ， and carcinogenic risk was mainly from unknown sources. The unknown sources had a high proportion in source apportionment and risk assessment， and should be further researched to provide scientific basis for soil heavy metal control. The combined risk assessment model based on source analysis， focusing on the risk characteristics of different sources， can accurately identify high-risk pollution sources. It is a more reasonable and reliable risk assessment method.

Stability Analysis of the Rotor-Journal Bearing System Considering Shear and Gaseous Cavitation

Part of the gas phase within the bearing emanates from the gaseous lubricating medium generated by the phase transition of the liquid lubricant under low pressure, while the remaining portion originates from the expansion of gases, such as air, present in the lubricant. This study delves into the impact of vapor and gas cavitation on the stability of the rotor-journal bearing system. Utilizing computational fluid dynamics (CFD), a 3D transient lubrication model is developed for the rotor-journal bearing system. This model integrates a combined cavitation approach, encompassing both vaporous and gaseous cavitation phenomena. Based on a new structured dynamic mesh method, the journal orbits are obtained when the journal moves in the rotor-journal bearing system. In vaporous and gaseous cavitation, shear stress and non-condensable gases (NCG) are incorporated successively. Compared with the combined cavitation model, the basic cavitation model journal orbit amplitude is significantly larger than the combined cavitation model. The carrying capacity of journal bearings under the basic cavitation model is overestimated, leading to a more conservative prediction for system stability.

Comparative study of stretched-exponential and kurtosis models of diffusion-weighted imaging in renal assessment to distinguish patients with primary aldosteronism from healthy controls.

To compare the ability of diffusion parameters obtained by stretched-exponential and kurtosis models of diffusion-weighted imaging (DWI) to distinguish between patients with primary aldosteronism (PA) and healthy controls (HCs) in renal assessment. A total of 44 participants (22 patients and 22 HCs) underwent renal MRI with an 11 b-value DWI sequence and a 3 b-value diffusion kurtosis imaging (DKI) sequence from June 2021 to April 2022. Binary logistic regression was used to construct regression models combining different diffusion parameters. Receiver-operating characteristic (ROC) curve analysis and comparisons were used to evaluate the ability of single diffusion parameters and combined diffusion models to distinguish between the two groups. A total of six diffusion parameters (including the cortical anomalous exponent term [α_Cortex], medullary fractional anisotropy [FA_Medulla], cortical FA [FA_Cortex], cortical axial diffusivity [Da_Cortex], medullary mean diffusivity [MD_Medulla] and medullary radial diffusivity [Dr_Medulla]) were included, and 10 regression models were studied. The area under the curve (AUC) of Dr_Medulla was 0.855, comparable to that of FA_Cortex and FA_Medulla and significantly higher than that of α_Cortex, Da_Cortex and MD_Medulla. The AUC of the Model_all parameters was 0.967, comparable to that of Model_FA (0.946) and Model_DKI (0.966) and significantly higher than that of the other models. The sensitivity and specificity of Model_all parameters were 87.2% and 95%, respectively. The Model_all parameters, Model_FA and Model_DKI were valid for differentiating between PA patients and HCs with similar differentiation efficacy and were superior to single diffusion parameters and other models.

Predicting the 90-day prognosis of stereotactic brain hemorrhage patients by multiple machine learning using radiomic features combined with clinical features.

Hypertensive Intracerebral Hemorrhage (HICH) is one of the most common types of cerebral hemorrhage with a high mortality and disability rate. Currently, preoperative non-contrast computed tomography (NCCT) scanning-guided stereotactic hematoma removal has achieved good results in treating HICH, but some patients still have poor prognoses. This study collected relevant clinical and radiomic data by retrospectively collecting and analyzing 432 patients who underwent stereotactic hematoma removal for HICH from January 2017 to December 2020 at the Liuzhou Workers Hospital. The prognosis of patients after 90 days was judged by the modified Rankin Scale (mRS) scale and divided into the good prognosis group (mRS ≤ 3) and the poor prognosis group (mRS > 3). The 268 patients were randomly divided into training and test sets in the ratio of 8:2, with 214 patients in the training set and 54 patients in the test set. The least absolute shrinkage and selection operator (Lasso) was used to screen radiomics features. They were combining clinical features and radiomic features to build a joint prediction model of the nomogram. The AUCs of the clinical model for predicting different prognoses of patients undergoing stereotactic HICH were 0.957 and 0.922 in the training and test sets, respectively, while the AUCs of the radiomics model were 0.932 and 0.770, respectively, and the AUCs of the combined prediction model for building a nomogram were 0.987 and 0.932, respectively. Compared with a single clinical or radiological model, the nomogram constructed by fusing clinical variables and radiomic features could better identify the prognosis of HICH patients undergoing stereotactic hematoma removal after 90 days.

Autoregressive Feature Extraction with Topic Modeling for Aspect-based Sentiment Analysis of Arabic as a Low-resource Language

This paper proposes an approach for aspect-based sentiment analysis of Arabic social data, especially the considerable text corpus generated through communications on X (formerly known as Twitter) for expressing opinions in Arabic-language tweets during the COVID-19 pandemic. The proposed approach examines the performance of several pre-trained predictive and autoregressive language models; namely, Bidirectional Encoder Representations from Transformers (BERT) and XLNet, along with topic modeling algorithms; namely, Latent Dirichlet Allocation (LDA) and Non-negative Matrix Factorization (NMF), for aspect-based sentiment analysis of online Arabic text. In addition, Bidirectional Long Short Term Memory (Bi-LSTM) deep learning model is used to classify the extracted aspects from online reviews. Obtained experimental results indicate that the combined XLNet-NMF model outperforms other implemented state-of-the-art methods through improving the feature extraction of unstructured social media text with achieving values of 0.946 and 0.938, for average sentiment classification accuracy and F-measure, respectively.

Clinical application of machine learning models in patients with prostate cancer before prostatectomy

BackgroundTo build machine learning predictive models for surgical risk assessment of extracapsular extension (ECE) in patients with prostate cancer (PCa) before radical prostatectomy; and to compare the use of decision curve analysis (DCA) and receiver operating characteristic (ROC) metrics for selecting input feature combinations in models.MethodsThis retrospective observational study included two independent data sets: 139 participants from a single institution (training), and 55 from 15 other institutions (external validation), both treated with Robotic Assisted Radical Prostatectomy (RARP). Five ML models, based on different combinations of clinical, semantic (interpreted by a radiologist) and radiomics features computed from T2W-MRI images, were built to predict extracapsular extension in the prostatectomy specimen (pECE+). DCA plots were used to rank the models’ net benefit when assigning patients to prostatectomy with non-nerve-sparing surgery (NNSS) or nerve-sparing surgery (NSS), depending on the predicted ECE status. DCA model rankings were compared with those drived from ROC area under the curve (AUC).ResultsIn the training data, the model using clinical, semantic, and radiomics features gave the highest net benefit values across relevant threshold probabilities, and similar decision curve was observed in the external validation data. The model ranking using the AUC was different in the discovery group and favoured the model using clinical + semantic features only.ConclusionsThe combined model based on clinical, semantic and radiomic features may be used to predict pECE + in patients with PCa and results in a positive net benefit when used to choose between prostatectomy with NNS or NNSS.