Cross-validation Accuracy Research Articles

ECG-Based Driving Fatigue Detection Using Heart Rate Variability Analysis with Mutual Information

One of the WHO’s strategies to reduce road traffic injuries and fatalities is to enhance vehicle safety. Driving fatigue detection can be used to increase vehicle safety. Our previous study developed an ECG-based driving fatigue detection framework with AdaBoost, producing a high cross-validated accuracy of 98.82% and a testing accuracy of 81.82%; however, the study did not consider the driver’s cognitive state related to fatigue and redundant features in the classification model. In this paper, we propose developments in the feature extraction and feature selection phases in the driving fatigue detection framework. For feature extraction, we employ heart rate fragmentation to extract non-linear features to analyze the driver’s cognitive status. These features are combined with features obtained from heart rate variability analysis in the time, frequency, and non-linear domains. In feature selection, we employ mutual information to filter redundant features. To find the number of selected features with the best model performance, we carried out 28 combination experiments consisting of 7 possible selected features out of 58 features and 4 ensemble learnings. The results of the experiments show that the random forest algorithm with 44 selected features produced the best model performance testing accuracy of 95.45%, with cross-validated accuracy of 98.65%.

An innovative tool for automating classification of stellar variability through nonlinear data analytics

Though Classical Cepheids, δ-Scuti, Eclipsing Binary, Long-Period variables, and RRLyraes are abundant in most of the clusters, automating the classification of the objects faces challenges. Since the rate at which the data has been getting accumulated is enormous, this automation of classification is paramount for carrying out appropriate analysis of the objects depending on the class it belongs to. Our results prove that the proposed tool for automating stellar classification not only reduces misclassification by up to 94.79% (in case of classification between multimode subclass of δ-Scuti and Mira subclass of Long-Period variables) but also improves reliability by as high as 78.35% (in case of conventionally misclassified pair of RRab subclass of RRLyrae and Fundamental Mode subclass of Classical Cepheids). Our random forest model has achieved a cross-validation accuracy of 0.88 with conventional statistical parameters coupled with tools of Nonlinear Dynamical Theory. It has achieved the highest precision and recalls for Long-Period variables of the Mira subclass (i.e., 0.99 & 0.99) and the lowest for Eclipsing Binary of subclass contact (i.e., 0.81 & 0.77). A positive improvement in accuracy rate by 7.3% is observed when compared with a model based on a conventional statistical platform. This proves the significance of introducing the proposed tools in devising an automated classification model for stellar variables.

Hybrid Whale Archimedes Optimization-based MLPNN model for soil nutrient classification and pH prediction.

Soil fertility and environmental factors play an important role in improving productivity and cropland quality in the agricultural sector. A new prediction and classification model for the potential of soil nutrients and hydrogen (pH) levels is proposed. The proposed model, Hybrid Whale Archimedes Optimization-based Multilayer Perceptron Neural Network (HWAO-MLPNN), is utilized for soil features classification of pH levels, and the soils are collected from the villages such as phosphorous (P), organic carbon (OC), boron (B), and potassium (K). The village-wise soil fertility prediction and classification model aims to improve soil health, reduce harmful fertilizer usage, enhance environmental quality, and achieve more profits. The proposed model combines the Multilayer Perceptron Neural Network (MLPNN) model and the Hybrid Whale Archimedes Optimization (HWAO) algorithm to enhance the classification performance on the validation data. The Marathwada dataset is selected to validate the soil nutrient prediction and classification model, and various measuring units such as cross-validation accuracy, Area Under Curve (AUC), accuracy, Mean Squared Error (MSE), G-mean, precision, specificity, and sensitivity are used for evaluation. The comparative study of this paper shows that the proposed HWAO-MLPNN achieved more classification accuracy of 98.1%, cross-validation accuracy of 98.3% for pH classification, and cross-validation accuracy of 97.9% for soil nutrient classification. The proposed model can be utilized to accurately classify soil nutrients and pH levels, which can have a significant impact on improving soil health, reducing harmful fertilizer usage, enhancing environmental quality, and ultimately increasing profitability in the agricultural sector.

Prediction of Progression After Cervix Cancer Radiotherapy Using a Machine-Learning Model on Pre-Treatment MRI

MRI may be useful to identify women with cervical cancer at high risk of disease progression to test strategies of treatment intensification. The purpose of this study was to determine the value of a machine-learning model built on pre-treatment MRI for prediction of risk of progression after radiation therapy. MagneticResonance Imaging (MRI) data for women with cervical cancer was collected from The Cancer Genome Atlas Cervical Squamous Cell Carcinoma and Endocervical Adenocarcinoma Collection (TCGA-CESC) on the Cancer Imaging Archive (TCIA), which reported clinical, treatment, and imaging data from a single institution. 27 patients who had received radiation for cervical cancer were selected for input into a custom 3-D Residual Neural Network (ResNet) model with added custom layers specific to DICOM data in tensorflow python package. One T2 MRI per patient was used to predict recurrence free survival after radiation treatment, where patients were predicted to be "high risk" or "low risk" for disease recurrence as the output of the model. All slices of the T2 MRI were used. The model was validated using five-fold cross validation; 80% of the data was used to train each fold and 20% was used for testing. Final model statistical significance was confirmed through shuffle test at the p < 0.01 level. The clinical outcomes of patients and the model's "low-risk" and "high-risk" prediction were compared. There were 27 patients in the study with mean age of 51 years (range 29-79). 20 patients had squamous cell carcinoma and 7 patients had adenocarcinoma. The stage breakdown consisted of 9 women IB, 2 IIA, 9 IIB, 2 IIIA, 2 IIIB, and 3 stage IV. 10 women were treated with radiation alone and 17 with chemo-radiation. 5 women received surgery in addition to radiation or chemoradiation. 21 patients received brachytherapy. Median follow-up of patients was 29 months (range 3-64). The model predicted 7 patients as "high risk" for recurrence; all 7 developed a recurrence during follow up. None of the 20 patients predicted to be "low risk" developed disease recurrence. Among all patients in the study, the two-year progression free survival (PFS) was 82.0%. Patients identified as "low risk" and "high risk" by model had two-year PFS of 100% and 43%, respectively. Among patients with recurrence, 3 developed local recurrence and 4 developed distant metastases. The ResNet model achieved cross-validated accuracy of 92% for prediction of progression-free survival (p<0.01). A 3-D ResNet machine-learning model using pretreatment MRI image data can accurately predict clinical outcomes for cervical cancer following radiation therapy. Future work to confirm generalizability should focus on validation with a larger clinical dataset.

Can lung airway geometry be used to predict autism? A preliminary machine learning-based study.

The goal of this study is to assess the feasibility of airway geometry as a biomarker for autism spectrum disorder (ASD). Chest computed tomography images of children with a documented diagnosis of ASD as well as healthy controls were identified retrospectively. Fifty-four scans were obtained for analysis, including 31 ASD cases and 23 controls. A feature selection and classification procedure using principal component analysis and support vector machine achieved a peak cross validation accuracy of nearly 89% using a feature set of eight airway branching angles. Sensitivity was 94%, but specificity was only 78%. The results suggest a measurable difference in airway branching angles between children with ASD and the control population.

Development of Machine Learning Algorithms for Riverside Land Cover Classification Using Synthetic Aperture Radar Satellite Imagery and Terrain Data

Riverine environments play a crucial role in maintaining the stability of river ecosystems as well as biodiversity. Furthermore, the appropriate management of small rivers has a significant impact not only on stable water supplies but also on water resource management. Wide monitoring of the riverside environment including land covers and their changes is an important issue in water resource management. This study aims to develop a high-resolution (10 m) model for classifying riverside land cover by integrating Sentinel-1 synthetic aperture radar (SAR) data and terrestrial characteristics using machine learning algorithms. We constructed a total of 3,284 landcover reference point datasets near the four major rivers of South Korea with five classes: water, barren, grass, forest, and built-up. The Random Forest and Light Gradient Boosting Machine classification models were developed using eight input variables derived from SAR signal and digital terrain data. The models showed an overall cross-validation accuracy exceeding 80% while maintaining consistent spatial distributions, except for the barren class. The false alarms on barren would be corrected through additional sampling processes and incorporating optical characteristics in further study. The high-resolution riverside land cover maps are expected to contribute to the establishment of a comprehensive management system for water resources such as riverside land cover change detection, river ecosystem monitoring, and flood hazard management. Furthermore, the utilization of the next generation medium satellite 5 (C-band SAR) would improve the performance of riverside land cover classification algorithm in the future.

Authentication of aged beef in terms of aging time and aging type by 1H NMR spectroscopy

Meat authenticity addresses parameters such as species, breed, sex, housing system and postmortem treatment. Seventy-four beef backs from two breeds (‘Fleckvieh’ and ‘Schwarzbunt’) and three cattle types (heifer, cow, young bull) were dry-aged and wet-aged up to 28 days and analyzed by 1H NMR spectroscopy. Statistical models based on partial least squares regression and discriminant analysis were performed to classify the beef samples by breed, cattle type, aging time, and aging type based on their 1H NMR spectra. The aging time of beef samples can be predicted with an error ± 2.28 days. The cattle type model has an accuracy of cross-validation of 99.2 %, the breed models of 100 % and the aging type model for 28-days aged samples of 99.6 %. These models allow the authentication of beef samples in terms of breed, cattle type, aging time, and aging type with a single 1H NMR measurement.

A cross-cohort computational framework to trace tumor tissue-of-origin based on RNA sequencing

Carcinoma of unknown primary (CUP) is a type of metastatic cancer with tissue-of-origin (TOO) unidentifiable by traditional methods. CUP patients typically have poor prognosis but therapy targeting the original cancer tissue can significantly improve patients’ prognosis. Thus, it’s critical to develop accurate computational methods to infer cancer TOO. While qPCR or microarray-based methods are effective in inferring TOO for most cancer types, the overall prediction accuracy is yet to be improved. In this study, we propose a cross-cohort computational framework to trace TOO of 32 cancer types based on RNA sequencing (RNA-seq). Specifically, we employed logistic regression models to select 80 genes for each cancer type to create a combined 1356-gene set, based on transcriptomic data from 9911 tissue samples covering the 32 cancer types with known TOO from the Cancer Genome Atlas (TCGA). The selected genes are enriched in both tissue-specific and tissue-general functions. The cross-validation accuracy of our framework reaches 97.50% across all cancer types. Furthermore, we tested the performance of our model on the TCGA metastatic dataset and International Cancer Genome Consortium (ICGC) dataset, achieving an accuracy of 91.09% and 82.67%, respectively, despite the differences in experiment procedures and pipelines. In conclusion, we developed an accurate yet robust computational framework for identifying TOO, which holds promise for clinical applications. Our code is available at http://github.com/wangbo00129/classifybysklearn.

Deep Learning-Based Detection of Bone Tumors around the Knee in X-rays of Children.

Even though tumors in children are rare, they cause the second most deaths under the age of 18 years. More often than in other age groups, underage patients suffer from malignancies of the bones, and these mostly occur in the area around the knee. One problem in the treatment is the early detection of bone tumors, especially on X-rays. The rarity and non-specific clinical symptoms further prolong the time to diagnosis. Nevertheless, an early diagnosis is crucial and can facilitate the treatment and therefore improve the prognosis of affected children. A new approach to evaluating X-ray images using artificial intelligence may facilitate the detection of suspicious lesions and, hence, accelerate the referral to a specialized center. We implemented a Vision Transformer model for image classification of healthy and pathological X-rays. To tackle the limited amount of data, we used a pretrained model and implemented extensive data augmentation. Discrete parameters were described by incidence and percentage ratio and continuous parameters by median, standard deviation and variance. For the evaluation of the model accuracy, sensitivity and specificity were computed. The two-entity classification of the healthy control group and the pathological group resulted in a cross-validated accuracy of 89.1%, a sensitivity of 82.2% and a specificity of 93.2% for test groups. Grad-CAMs were created to ensure the plausibility of the predictions. The proposed approach, using state-of-the-art deep learning methodology to detect bone tumors on knee X-rays of children has achieved very good results. With further improvement of the algorithm, enlargement of the dataset and removal of potential biases, this could become a useful additional tool, especially to support general practitioners for early, accurate and specific diagnosis of bone lesions in young patients.

Effective automatic detection of anterior cruciate ligament injury using convolutional neural network with two attention mechanism modules

BackgroundTo develop a fully automated CNN detection system based on magnetic resonance imaging (MRI) for ACL injury, and to explore the feasibility of CNN for ACL injury detection on MRI images.MethodsIncluding 313 patients aged 16 – 65 years old, the raw data are 368 pieces with injured ACL and 100 pieces with intact ACL. By adding flipping, rotation, scaling and other methods to expand the data, the final data set is 630 pieces including 355 pieces of injured ACL and 275 pieces of intact ACL. Using the proposed CNN model with two attention mechanism modules, data sets are trained and tested with fivefold cross-validation.ResultsThe performance is evaluated using accuracy, precision, sensitivity, specificity and F1 score of our proposed CNN model, with results of 0.8063, 0.7741, 0.9268, 0.6509 and 0.8436. The average accuracy in the fivefold cross-validation is 0.8064. For our model, the average area under curves (AUC) for detecting injured ACL has results of 0.8886.ConclusionWe propose an effective and automatic CNN model to detect ACL injury from MRI of human knees. This model can effectively help clinicians diagnose ACL injury, improving diagnostic efficiency and reducing misdiagnosis and missed diagnosis.

Comparing Pertubagens from Differential Gene Expression Data Analysis of ASD using Random Forest and Statistical Test

Abstract Differentially Expressed Genes (DEGs) are treated as candidate biomarkers, and a small set of DEGs might be identified as biomarkers using either biological knowledge or data-driven approaches like machine learning and statistical analysis. In this study, we used a combination of the machine learning algorithm and statistical tests to identify the top 300 genes that are differentially expressed in ASD compared to Typically Developed (TD). Initially, we extracted microarray gene expression data of 15 ASD and 15 TD from NCBI GEO database and used a standard pipeline to preprocess the data. Further, Random Forest (RF) was used to discriminate genes between ASD and TD. We, then analyzed the upregulated and downregulated genes using the logFC value to gain insights into their potential roles in the development of ASD. We further used drug-gene interaction analysis from ConnectivityMap to identify drugs that can inhibit the expression of these genes. Our results show that the proposed RF model yields average 5-fold cross-validation accuracy, sensitivity, and specificity of 96.67%. Further, we obtained precision and F-measure scores of 97.5% and 96.57%, respectively. Our analysis identified several novel genes that are dysregulated in ASD, including genes (such as proliferation-inducing protein 38 and germinal centre expressed transcript 32) involved in synaptic transmission, neural development, and immune function. We also identified several drugs (such as ATPase_Inhibitor, kinase inhibitors, and histone deacetylase inhibitors) that can potentially be used to treat ASD. Our findings provide new insights into the molecular mechanisms of ASD and suggest potential targets for drug development. These findings may lead to new therapeutic approaches for the treatment of ASD.

A Prospective Study Demonstrating Early Prediction of Skin Toxicity From Radiation Therapy Using Radiomic Features From Optical and Infrared Images

Approximately 90% of patients undergoing breast cancer radiation therapyexperience skin toxicities that are difficult to classify and predict ahead of time. A prediction oftoxicityat the early stages of the treatment would provide clinicians with a prompt to intervene. The objectives of this study were to evaluate the correlation between skin toxicity and radiomic features extracted from optical and infrared (thermal) images of skin, and to develop a model for predicting a patient's skin response to radiation. Optical and infrared breast and chest-wall images were acquired daily during the course of radiation therapy, as well as weekly for 3 weeks after the end of treatment for 20 patients with breast cancer. Skin-toxicity assessments were conducted weekly until the patients' finalvisit. Skin color and temperature trends from histogram-based and texture-based radiomic features, extracted from the treatment area, were analyzed, reduced, and used in a cross-validation machine learning model to predict the patients' skin toxicity grades. A set of 9 independent color and temperature features with significant correlation to skin toxicitywere identified from 108 features. The cross-validation accuracy of a cubic Support Vector Machine remained >85% and area under the receiver operating characteristic curve remained>0.75,when reducing the input imaging data to include onlythe sessions with a biologically effective dosenot exceeding30 Gy(approximately the first third to first half of the total treatment dose). The quantitative analysis of radiomic featuresextracted from opticaland infrared (thermal) images of skinwas shown to be promising for predicting skin toxicities.

Targeted gene expression profiling for accurate endometrial receptivity testing

Expressional profiling of the endometrium enables the personalised timing of the window of implantation (WOI). This study presents and evaluates a novel analytical pipeline based on a TAC-seq (Targeted Allele Counting by sequencing) method for endometrial dating. The expressional profiles were clustered, and differential expression analysis was performed on the model development group, using 63 endometrial biopsies spanning over proliferative (PE, n = 18), early-secretory (ESE, n = 18), mid-secretory (MSE, n = 17) and late-secretory (LSE, n = 10) endometrial phases of the natural cycle. A quantitative predictor model was trained on the development group and validated on sequenced samples from healthy women, consisting of 52 paired samples taken from ESE and MSE phases and five LSE phase samples from 31 individuals. Finally, the developed test was applied to 44 MSE phase samples from a study group of patients diagnosed with recurrent implantation failure (RIF). In validation samples (n = 57), we detected displaced WOI in 1.8% of the samples from fertile women. In the RIF study group, we detected a significantly higher proportion of the samples with shifted WOI than in the validation set of samples from fertile women, 15.9% and 1.8% (p = 0.012), respectively. The developed model was evaluated with an average cross-validation accuracy of 98.8% and an accuracy of 98.2% in the validation group. The developed beREADY screening model enables sensitive and dynamic detection of selected transcriptome biomarkers, providing a quantitative and accurate prediction of endometrial receptivity status.

TROLLOPE: A novel sequence-based stacked approach for the accelerated discovery of linear T-cell epitopes of hepatitis C virus.

Hepatitis C virus (HCV) infection is a concerning health issue that causes chronic liver diseases. Despite many successful therapeutic outcomes, no effective HCV vaccines are currently available. Focusing on T cell activity, the primary effector for HCV clearance, T cell epitopes of HCV (TCE-HCV) are considered promising elements to accelerate HCV vaccine efficacy. Thus, accurate and rapid identification of TCE-HCVs is recommended to obtain more efficient therapy for chronic HCV infection. In this study, a novel sequence-based stacked approach, termed TROLLOPE, is proposed to accurately identify TCE-HCVs from sequence information. Specifically, we employed 12 different sequence-based feature descriptors from heterogeneous perspectives, such as physicochemical properties, composition-transition-distribution information and composition information. These descriptors were used in cooperation with 12 popular machine learning (ML) algorithms to create 144 base-classifiers. To maximize the utility of these base-classifiers, we used a feature selection strategy to determine a collection of potential base-classifiers and integrated them to develop the meta-classifier. Comprehensive experiments based on both cross-validation and independent tests demonstrated the superior predictive performance of TROLLOPE compared with conventional ML classifiers, with cross-validation and independent test accuracies of 0.745 and 0.747, respectively. Finally, a user-friendly online web server of TROLLOPE (http://pmlabqsar.pythonanywhere.com/TROLLOPE) has been developed to serve research efforts in the large-scale identification of potential TCE-HCVs for follow-up experimental verification.

Classification of concrete corrosion states by GPR with machine learning

The evaluation of rebar corrosion in reinforced concrete by using ground penetrating radar (GPR) and machine learning (ML) is a complex process. In this paper, a multi-variate method is presented. It uses full-volume data obtained from the amplitude domain in a regular GPR x-y scanning exercise, and the shape of the rebar’s reflection to categorise different corrosion phases. This method allows multi-dimensional analysis with quantifiable GPR attributes. GPR data were extracted from the field and laboratory and then labelled according to the ground truths and reference specimens. A classic ML algorithm, logistic regression, was applied. The cross-validation accuracy (sensitivity and specificity) of individual corrosion phases was high (>99%), and the false alarm rate was low (<1%). This work shows that GPR as an evaluation tool can assess unseen data like doing blind tests. Nonetheless, continuous expansion of the training database is suggested to increase its diversity in the future.

Spatiotemporal Weighted for Improving the Satellite-Based High-Resolution Ground PM2.5 Estimation Using the Light Gradient Boosting Machine

Surface fine particulate matter (PM) with a diameter of less than 2.5 microns (PM2.5) negatively impacts human health and the economy. However, due to data and model limitations, obtaining high-quality, high-spatial-resolution surface PM2.5 concentration data is a challenging task, and it is difficult to accurately assess the temporal and spatial changes in PM2.5 levels at a small regional scale. Here, we combined multi-angle implementation of atmospheric correction (MAIAC) aerosol products, ERA5 reanalysis data, etc., to construct an STW-LightGBM model that considers the spatiotemporal characteristics of air pollution and estimate the PM2.5 concentration of China’s surface at 1 km resolution from 2015 to 2020. Our model performed well, and the fitting accuracy of the 10-fold cross-validation between years was 0.877–0.917. The fitting accuracy of the model was &gt;0.85 at different time scales (month, season, and year). The average slope of the regression prediction was 0.9 annually. The results showed that PM2.5 pollution improved from 2015 to 2020. The average PM2.5 concentration decreased by 4.55 μg/m3, and the maximum decrease reached 90.51 μg/m3. The areas with high PM2.5 concentrations were predominantly in the North China Plain, Sichuan Basin, and Xinjiang in the west, and the levels in areas with elevated PM2.5 levels were consistent across most study years. The standard deviation ellipse for PM2.5 in China showed a ‘northeast–southwest’ spatial distribution. From an interannual perspective, the average values of the four seasonal stations in the country showed a downward trend from 2015 to 2020, with the most obvious decline in winter, from 70.67 μg/m3 in 2015 to 46.75 μg/m3 in 2020. Compared to earlier inversion studies, this work provides a more stable and accurate method for obtaining high-resolution PM2.5 data, which is necessary for local air governance and environmental ecological construction at a fine scale.

A new computer vision workflow to assess yield quality traits in bush bean (Phaseolus vulgaris L.)

Quality assessments of horticultural products are still often carried out manually in breeding contexts, although computer vision systems have been reported to be able to overcome the limitations of manual assessments, e.g. in automated food processing. Here, a new computer vision workflow for quality trait assessment of bush bean pods (Phaseolus vulgaris) is introduced to replace physical measurements and visual scorings of expert breeders, while increasing consistency, accuracy, and objectivity of the measurements. A closed imaging box was used to take images of bean pods from 40 different varieties to develop and validate computer vision workflows to assess breeding relevant shape and color traits of bean pods. For the detection of beaks and peduncles via a neural network approach (Mask R-CNN) accuracies of 95.5% were reached. Computer vision estimations and manual reference measurements of length and caliber were highly correlated (R=0.99). Also, curvature and brightness of green bean pods well- correlated with visual scorings of expert breeders (R=0.81, R=-0.87). A Random Forest Classifier was trained to distinguish yellow and extremely rare bicolored pods and a cross validation accuracy of 83 ±7% was reached. An additional backlight LED panel enabled non-destructive analysis of seed formation inside the pod and promising results were achieved using a Faster R-CNN model. This new computer vision workflow provides the opportunity to replace well-established manual workflows for quality trait assessment of bush bean pods as it is more objective, reliable, and considerably faster.

Self-supervised ensembled learning for autism spectrum classification

PurposeDeep learning has made remarkable progress in classifying autism spectrum disorder (ASD) using neuroimaging data. However, the current methods rely mainly on supervised learning, which requires a large amount of manually labeled data, making it an expensive and difficult task to scale. MethodsTo overcome this limitation, we propose a novel ensemble-based framework that learns a transferable and generalizable visual representation from different self-supervised features for the downstream task of ASD classification. This framework dynamically learns a superior representation by aggregating complementary information in the frequency domain from independent self-supervised features with limited data. Additionally, to address the information loss caused by the dimensionality reduction of 3D fMRI data, we propose a thresholding algorithm to optimally extract the most discriminant features from 2D rs-fMRI data. ResultsExperimental results demonstrate that the proposed method outperforms previous state-of-the-art methods by 19.69% on the ABIDE-1 dataset with a 10-fold cross-validation accuracy of 94.51%. ConclusionThe proposed method learns a transferrable and generalizable ensembled representation by leveraging complementary information encoded in different self-supervised representations for ASD classification.

Blunted Neurobehavioral Loss Anticipation Predicts Relapse to Stimulant Drug Use

BackgroundPatients with stimulant use disorder experience high rates of relapse. While neurobehavioral mechanisms involved in initiating drug use have been studied extensively, less research has focused on relapse. MethodsTo assess motivational processes involved in relapse and diagnosis, we acquired functional magnetic resonance imaging responses to nondrug (monetary) gains and losses in detoxified patients with stimulant use disorder (n = 68) and community control participants (n = 42). In a prospective multimodal design, we combined imaging of brain function, brain structure, and behavior to longitudinally track subsequent risk for relapse. ResultsAt the 6-month follow-up assessment, 27 patients remained abstinent, but 33 had relapsed. Patients with blunted anterior insula (AIns) activity during loss anticipation were more likely to relapse, an association that remained robust after controlling for potential confounds (i.e., craving, negative mood, years of use, age, and gender). Lower AIns activity during loss anticipation was associated with lower self-reported negative arousal to loss cues and slower behavioral responses to avoid losses, which also independently predicted relapse. Furthermore, AIns activity during loss anticipation was associated with the structural coherence of a tract connecting the AIns and the nucleus accumbens, as was functional connectivity between the AIns and nucleus accumbens during loss processing. However, these neurobehavioral responses did not differ between patients and control participants. ConclusionsTaken together, the results of the current study show that neurobehavioral markers predicted relapse above and beyond conventional self-report measures, with a cross-validated accuracy of 72.7%. These findings offer convergent multimodal evidence that implicates blunted avoidance motivation in relapse to stimulant use and may therefore guide interventions targeting individuals who are most vulnerable to relapse.

A novel MRI-based deep learning networks combined with attention mechanism for predicting CDKN2A/B homozygous deletion status in IDH-mutant astrocytoma.

To develop a high-accuracy MRI-based deep learning method for predicting cyclin-dependent kinase inhibitor 2A/B (CDKN2A/B) homozygous deletion status in isocitrate dehydrogenase (IDH)-mutant astrocytoma. Multiparametric brain MRI data and corresponding genomic information of 234 subjects (111 positives for CDKN2A/B homozygous deletion and 123 negatives for CDKN2A/B homozygous deletion) were obtained from The Cancer Imaging Archive (TCIA) and The Cancer Genome Atlas (TCGA) respectively. Two independent multi-sequence networks (ResFN-Net and FN-Net) are built on the basis of ResNet and ConvNeXt network combined with attention mechanism to classify CDKN2A/B homozygous deletion status using MR images including contrast-enhanced T1-weighted imaging (CE-T1WI) and T2-weighted imaging (T2WI). The performance of the network is summarized by three-way cross-validation; ROC analysis is also performed. The average cross-validation accuracy (ACC) of ResFN-Net is 0.813. The average cross-validation area under curve (AUC) of ResFN-Net is 0.8804. The average cross-validation ACC and AUC of FN-Net is 0.9236 and 0.9704, respectively. Comparing all sequence combinations of the two networks (ResFN-Net and FN-Net), the sequence combination of CE-T1WI and T2WI performed the best, and the ACC and AUC were 0.8244, 0.8975 and 0.8971, 0.9574, respectively. The FN-Net deep learning networks based on ConvNeXt network achieved promising performance for predicting CDKN2A/B homozygous deletion status of IDH-mutant astrocytoma. A novel deep learning network (FN-Net) based on preoperative MRI was developed to predict the CDKN2A/B homozygous deletion status. This network has the potential to be a practical tool for the noninvasive characterization of CDKN2A/B in glioma to support personalized classification and treatment planning. • CDKN2A/B homozygous deletion status is an important marker for glioma grading and prognosis. • An MRI-based deep learning approach was developed to predict CDKN2A/B homozygous deletion status. • The predictive performance based on ConvNeXt network was better than that of ResNet network.