Search

Abstract Mental workload refers to the cognitive effort required to perform tasks, and it is an important factor in various fields, including system design, clinical medicine, and industrial applications. In this paper, we propose innovative methods to assess mental workload from EEG data that use effective brain connectivity for the purpose of extracting features, a hierarchical feature selection algorithm to select the most significant features, and finally machine learning models. We have used the Simultaneous Task EEG Workload (STEW) dataset, an open-access collection of raw EEG data from 48 subjects. We extracted brain-effective connectivities by the direct directed transfer function and then selected the top 30 connectivities for each standard frequency band. Then we applied three feature selection algorithms (forward feature selection, Relief-F, and minimum-redundancy-maximum-relevance) on the top 150 features from all frequencies. Finally, we applied sevenfold cross-validation on four machine learning models (support vector machine (SVM), linear discriminant analysis, random forest, and decision tree). The results revealed that SVM as the machine learning model and forward feature selection as the feature selection method work better than others and could classify the mental workload levels with accuracy equal to 89.53% (± 1.36).

Abstract This study aims to develop a machine learning approach leveraging clinical data and blood parameters to predict non-alcoholic steatohepatitis (NASH) based on the NAFLD Activity Score (NAS). Using a dataset of 181 patients, we performed preprocessing including normalization and categorical encoding. To identify predictive features, we applied sequential forward selection (SFS), chi-square, analysis of variance (ANOVA), and mutual information (MI). The selected features were used to train machine learning classifiers including SVM, random forest, AdaBoost, LightGBM, and XGBoost. Hyperparameter tuning was done for each classifier using randomized search. Model evaluation was performed using leave-one-out cross-validation over 100 repetitions. Among the classifiers, random forest, combined with SFS feature selection and 10 features, obtained the best performance: Accuracy: 81.32% ± 6.43%, Sensitivity: 86.04% ± 6.21%, Specificity: 70.49% ± 8.12% Precision: 81.59% ± 6.23%, and F1-score: 83.75% ± 6.23% percent. Our findings highlight the promise of machine learning in enhancing early diagnosis of NASH and provide a compelling alternative to conventional diagnostic techniques. Consequently, this study highlights the promise of machine learning techniques in enhancing early and non-invasive diagnosis of NASH based on readily available clinical and blood data. Our findings provide the basis for developing scalable approaches that can improve screening and monitoring of NASH progression.

Abstract This study aimed to investigate the diagnostic performance of machine learning-based radiomics analysis to diagnose coronary artery disease status and risk from rest/stress Myocardial Perfusion Imaging (MPI) single-photon emission computed tomography (SPECT). A total of 395 patients suspicious of coronary artery disease who underwent 2-day stress-rest protocol MPI SPECT were enrolled in this study. The left ventricle myocardium, excluding the cardiac cavity, was manually delineated on rest and stress images to define a volume of interest. Added to clinical features (age, sex, family history, diabetes status, smoking, and ejection fraction), a total of 118 radiomics features, were extracted from rest and stress MPI SPECT images to establish different feature sets, including Rest-, Stress-, Delta-, and Combined-radiomics (all together) feature sets. The data were randomly divided into 80% and 20% subsets for training and testing, respectively. The performance of classifiers built from combinations of three feature selections, and nine machine learning algorithms was evaluated for two different diagnostic tasks, including 1) normal/abnormal (no CAD vs. CAD) classification, and 2) low-risk/high-risk CAD classification. Different metrics, including the area under the ROC curve (AUC), accuracy (ACC), sensitivity (SEN), and specificity (SPE), were reported for models’ evaluation. Overall, models built on the Stress feature set (compared to other feature sets), and models to diagnose the second task (compared to task 1 models) revealed better performance. The Stress-mRMR-KNN (feature set-feature selection-classifier) reached the highest performance for task 1 with AUC, ACC, SEN, and SPE equal to 0.61, 0.63, 0.64, and 0.6, respectively. The Stress-Boruta-GB model achieved the highest performance for task 2 with AUC, ACC, SEN, and SPE of 0.79, 0.76, 0.75, and 0.76, respectively. Diabetes status from the clinical feature family, and dependence count non-uniformity normalized, from the NGLDM family, which is representative of non-uniformity in the region of interest were the most frequently selected features from stress feature set for CAD risk classification. This study revealed promising results for CAD risk classification using machine learning models built on MPI SPECT radiomics. The proposed models are helpful to alleviate the labor-intensive MPI SPECT interpretation process regarding CAD status and can potentially expedite the diagnostic process.

Abstract Prediction of response to Repetitive Transcranial Magnetic Stimulation (rTMS) can build a very effective treatment platform that helps Major Depressive Disorder (MDD) patients to receive timely treatment. We proposed a deep learning model powered up by state-of-the-art methods to classify responders (R) and non-responders (NR) to rTMS treatment. Pre-treatment Electro-Encephalogram (EEG) signal of public TDBRAIN dataset and 46 proprietary MDD subjects were utilized to create time–frequency representations using Continuous Wavelet Transform (CWT) to be fed into the two powerful pre-trained Convolutional Neural Networks (CNN) named VGG16 and EfficientNetB0. Equipping these Transfer Learning (TL) models with Bidirectional Long Short-Term Memory (BLSTM) and attention mechanism for the extraction of most discriminative spatiotemporal features from input images, can lead to superior performance in the prediction of rTMS treatment outcome. Five brain regions named Frontal, Central, Parietal, Temporal, and occipital were assessed and the highest evaluated performance in 46 proprietary MDD subjects was acquired for the Frontal region using the TL-LSTM-Attention model based on EfficientNetB0 with accuracy, sensitivity, specificity, and Area Under the Curve (AUC) of 97.1%, 97.3%, 97.0%, and 0.96 respectively. Additionally, to test the generalizability of the proposed models, these TL-BLSTM-Attention models were evaluated on a public dataset called TDBRAIN and the highest accuracy of 82.3%, the sensitivity of 80.2%, the specificity of 81.9% and the AUC of 0.83 were obtained. Therefore, advanced deep learning methods using a time–frequency representation of EEG signals from the frontal brain region and the convolutional recurrent neural networks equipped with the attention mechanism can construct an accurate platform for the prediction of response to the rTMS treatment.