Pneumonia is a microorganism infection that causes chronic inflammation of the human lung cells. Chest X-ray imaging is the most well-known screening approach used for detecting pneumonia in the early stages. While chest-Xray images are mostly blurry with low illumination, a strong feature extraction approach is required for promising identification performance. A new hybrid explainable deep learning framework is proposed for accurate pneumonia disease identification using chest X-ray images. The proposed hybrid workflow is developed by fusing the capabilities of both ensemble convolutional networks and the Transformer Encoder mechanism. The ensemble learning backbone is used to extract strong features from the raw input X-ray images in two different scenarios: ensemble A (i.e., DenseNet201, VGG16, and GoogleNet) and ensemble B (i.e., DenseNet201, InceptionResNetV2, and Xception). Whereas, the Transformer Encoder is built based on the self-attention mechanism with multilayer perceptron (MLP) for accurate disease identification. The visual explainable saliency maps are derived to emphasize the crucial predicted regions on the input X-ray images. The end-to-end training process of the proposed deep learning models over all scenarios is performed for binary and multi-class classification scenarios. The proposed hybrid deep learning model recorded 99.21% classification performance in terms of overall accuracy and F1-score for the binary classification task, while it achieved 98.19% accuracy and 97.29% F1-score for multi-classification task. For the ensemble binary identification scenario, ensemble A recorded 97.22% accuracy and 97.14% F1-score, while ensemble B achieved 96.44% for both accuracy and F1-score. For the ensemble multiclass identification scenario, ensemble A recorded 97.2% accuracy and 95.8% F1-score, while ensemble B recorded 96.4% accuracy and 94.9% F1-score. The proposed hybrid deep learning framework could provide promising and encouraging explainable identification performance comparing with the individual, ensemble models, or even the latest AI models in the literature. The code is available here: https://github.com/chiagoziemchima/Pneumonia_Identificaton.

Stress has become a dangerous health problem in our life, especially in student education journey. Accordingly, previous methods have been conducted to detect mental stress based on biological and biochemical effects. Moreover, hormones, physiological effects, and skin temperature have been extensively used for stress detection. However, based on the recent literature, biological, biochemical, and physiological-based methods have shown inconsistent findings, which are initiated due to hormones’ instability. Therefore, it is crucial to study stress using different mechanisms such as Electroencephalogram (EEG) signals. In this research study, the frontal lobes EEG spectrum analysis is applied to detect mental stress. Initially, we apply a Fast Fourier Transform (FFT) as a feature extraction stage to measure all bands’ power density for the frontal lobe. After that, we used two type of classifications such as subject wise and mix (mental stress vs. control) using Support Vector Machine (SVM) and Naive Bayes (NB) machine learning classifiers. Our obtained results of the average subject wise classification showed that the proposed technique has better accuracy (98.21%). Moreover, this technique has low complexity, high accuracy, simple and easy to use, no over fitting, and it could be used as a real-time and continuous monitoring technique for medical applications.

In the modern world, wearable smart devices are continuously used to monitor people’s health. This study aims to develop an automatic mental stress detection system for researchers based on Electrocardiogram (ECG) signals from smart T-shirts using machine learning classifiers. We used 20 subjects, including 10 from mental stress (after twelve hours of continuous work in the laboratory) and 10 from normal (after completing the sleep or without any work). We also applied three scoring techniques: Chalder Fatigue Scale (CFS), Specific Fatigue Scale (SFS), Depression, Anxiety, and Stress Scale (DASS), to confirm the mental stress. The total duration of ECG recording was 1800 min, including 1200 min during mental stress and 600 min during normal. We calculated two types of features, such as demographic and extracted by ECG signal. In addition, we used Decision Tree (DT), Naive Bayes (NB), Random Forest (RF), and Logistic Regression (LR) to classify the intra-subject (mental stress and normal) and inter-subject classification. The DT leave-one-out model has better performance in terms of recall (93.30%), specificity (96.70%), precision (94.40%), accuracy (93.30%), and F1 (93.50%) in the intra-subject classification. Additionally, The classification accuracy of the system in classifying inter-subjects is 94.10% when using a DT classifier. However, our findings suggest that the wearable smart T-shirt based on the DT classifier may be used in big data applications and health monitoring. Mental stress can lead to mitochondrial dysfunction, oxidative stress, blood pressure, cardiovascular disease, and various health problems. Therefore, real-time ECG signals help assess cardiovascular and related risk factors in the initial stage based on machine learning techniques.

Early risk identification of an unexpected sudden cardiac death (SCD) in a person who is suffering malignant ventricular arrhythmias is highly significant for timely intervention and increasing the survival rate. For this purpose, we have presented an automated strategy for prediction of SCD with a high-level accuracy by using measurable arrhythmic markers in this paper. The set of arrhythmic parameters includes three repolarization interval ratios, such as T _p T _e /QT, JT _p /JT _e , and T _p T _e /JT _p and two conduction-repolarization markers, such as T _p T _e /QRS and T _p T _e /(QT×QRS). Each of them is calculated directly from the detected QRS complex waves and T-wave of electrocradiogram (ECG) signals. Then, all calculated markers are used for the automatical classification of normal and SCD risk groups by employing machine learning classifiers, such as k-nearest neighbor (KNN), decision tree (DT), Naive Bayes (NB), support vector machine (SVM), and random forest (RF). The effectiveness and usefulness of the proposed method is evaluated using a database of measured ECG data acquired from 28 SCD and 18 normal patients. For the automated strategy, the set of five arrhythmic risk markers can predict SCD in less than one second with an average accuracy of 98.91% (KNN), 98.70% (SVM), 98.99% (DT), 97.46% (NB), and 99.49% (RF) for 30 minutes before the occurrence of SCD. Moreover, a practical and straightforward SCD index (SCDI) through a judicious integration of these markers is also proposed by using the Student's t-test. The obtained SCDIs are 1.2058 ± 0.0795 and 1.7619 ± 0.1902 for normal and SCD patients, respectively, which provide a sufficient discrimination degree with a p-value of 6.5061e-35. The present results show that both the automated classifier and the integrated SCDI can predict the SCD up to 30 minutes earlier, and that these predictions could be more practical and efficient if applied in portable smart devices with real-time requirements in hospital settings or at home.

CT vertebral segmentation plays an essential role in various clinical applications, such as computer‐assisted surgical interventions, assessment of spinal abnormalities, and vertebral compression fractures. Automatic CT vertebral segmentation is challenging due to the overlapping shadows of thoracoabdominal structures such as the lungs, bony structures such as the ribs, and other issues such as ambiguous object borders, complicated spine architecture, patient variability, and fluctuations in image contrast. Deep learning is an emerging technique for disease diagnosis in the medical field. This study proposes a patch‐based deep learning approach to extract the discriminative features from unlabeled data using a stacked sparse autoencoder (SSAE). 2D slices from a CT volume are divided into overlapping patches fed into the model for training. A random under sampling (RUS)‐module is applied to balance the training data by selecting a subset of the majority class. SSAE uses pixel intensities alone to learn high‐level features to recognize distinctive features from image patches. Each image is subjected to a sliding window operation to express image patches using autoencoder high‐level features, which are then fed into a sigmoid layer to classify whether each patch is a vertebra or not. We validate our approach on three diverse publicly available datasets: VerSe, CSI‐Seg, and the Lumbar CT dataset. Our proposed method outperformed other models after configuration optimization by achieving 89.9% in precision, 90.2% in recall, 98.9% in accuracy, 90.4% in F‐score, 82.6% in intersection over union (IoU), and 90.2% in Dice coefficient (DC). The results of this study demonstrate that our model’s performance consistency using a variety of validation strategies is flexible, fast, and generalizable, making it suited for clinical application.

Lack of sleep causes many sleep disorders such as nocturnal frontal lobe epilepsy, narcolepsy, bruxism, sleep apnea, insomnia, periodic limb movement disorder, and rapid eye movement behavioral disorder. Out of all, bruxism is a common behavior, which is found in 8-31% of the population. Bruxism is a sleep disorder in which individuals involuntarily grinds and clenches the teeth. The main aim of this work is to detect sleep bruxism by analyzing the electroencephalogram (EEG) spectrum analysis of the change in the domain of different stages of sleep. The present research was performed in different stages such as collection of the data, preprocessing of the EEG signal, analysis of the C4-P4 and C4-A1 channels, comparison between healthy humans and bruxism patients, and classification using decision tree method. In this study, the channels C4-P4 and C4-A1 of the EEG signal were combined for the detection of bruxism by using Welch technique, which mainly focused on two sleep stages such as S1 and rapid eye movement. The total number of EEG channels of healthy humans and bruxism patients analyzed in this work were 15 and 18, respectively. The results showed that the individual accuracy of the C4-P4 and C4-A1 channels was 81.70% and 74.11%, respectively. The combined accuracy of both C4-P4 and C4-A1 channels was 81.25%. The specificity of combined result was higher than individual. In addition, the value of theta activity during detection is consistent throughout the period, and the accuracy of S1 stage is better than rapid eye movement stage. We proposed that the theta activity of S1 could be taken for the detection of bruxism. The proposed approach in the detection of the bruxism is negligible in noise as it is in mathematical form and has taken very less time as compared with the traditional systems. The present research work would provide a fast and effective detection system of the sleep bruxism with high accuracy for medical big data applications.