Voice pathology disorders can be effectively detected using computer-aided voice pathology classification tools. These tools can diagnose voice pathologies at an early stage and offering appropriate treatment. This study aims to develop a powerful feature extraction voice pathology detection tool based on Deep Learning. In this paper, a pre-trained Convolutional Neural Network (CNN) was applied to a dataset of voice pathology to maximize the classification accuracy. This study also proposes a distinguished training method combined with various training strategies in order to generalize the application of the proposed system on a wide range of problems related to voice disorders. The proposed system has tested using a voice database, namely the Saarbrücken voice database (SVD). The experimental results show the proposed CNN method for speech pathology detection achieves accuracy up to 95.41%. It also obtains 94.22% and 96.13% for F1-Score and Recall. The proposed system shows a high capability of the real-clinical application that offering a fast-automatic diagnosis and treatment solutions within 3 s to achieve the classification accuracy.

In healthcare applications, numerous sensors and devices produce massive amounts of data which are the focus of critical tasks. Their management at the edge of the network can be done by Fog computing implementation. However, Fog Nodes suffer from lake of resources That could limit the time needed for final outcome/analytics. Fog Nodes could perform just a small number of tasks. A difficult decision concerns which tasks will perform locally by Fog Nodes. Each node should select such tasks carefully based on the current contextual information, for example, tasks’ priority, resource load, and resource availability. We suggest in this paper a Multi-Agent Fog Computing model for healthcare critical tasks management. The main role of the multi-agent system is mapping between three decision tables to optimize scheduling the critical tasks by assigning tasks with their priority, load in the network, and network resource availability. The first step is to decide whether a critical task can be processed locally; otherwise, the second step involves the sophisticated selection of the most suitable neighbor Fog Node to allocate it. If no Fog Node is capable of processing the task throughout the network, it is then sent to the Cloud facing the highest latency. We test the proposed scheme thoroughly, demonstrating its applicability and optimality at the edge of the network using iFogSim simulator and UTeM clinic data.

In the last decade, the developments in healthcare technologies have been increasing progressively in practice. Healthcare applications such as ECG monitoring, heartbeat analysis, and blood pressure control connect with external servers in a manner called cloud computing. The emerging cloud paradigm offers different models, such as fog computing and edge computing, to enhance the performances of healthcare applications with minimum end-to-end delay in the network. However, many research challenges exist in the fog-cloud enabled network for healthcare applications. Therefore, in this paper, a Critical Healthcare Task Management (CHTM) model is proposed and implemented using an ECG dataset. We design a resource scheduling model among fog nodes at the fog level. A multi-agent system is proposed to provide the complete management of the network from the edge to the cloud. The proposed model overcomes the limitations of providing interoperability, resource sharing, scheduling, and dynamic task allocation to manage critical tasks significantly. The simulation results show that our model, in comparison with the cloud, significantly reduces the network usage by 79%, the response time by 90%, the network delay by 65%, the energy consumption by 81%, and the instance cost by 80%.

Biomarkers including fasting blood sugar, heart rate, electrocardiogram (ECG), blood pressure, etc. are essential in the heart disease (HD) diagnosing. Using wearable sensors, these measures are collected and applied as inputs to a deep learning (DL) model for HD diagnosis. However, it is observed that model accuracy weakens when the data gathered are scarce or imbalanced. Therefore, this work proposes two DL-based frameworks, GAN-1D-CNN, and GAN-Bi-LSTM. These frameworks contain: (1) a generative adversarial network (GAN) and (2) a one-dimensional convolutional neural network (1D-CNN) or bi-directional long short-term memory (Bi-LSTM). The GAN model is utilized to augment the small and imbalanced dataset, which is the Cleveland dataset. The 1D-CNN and Bi-LSTM models are then trained using the enlarged dataset to diagnose HD. Unlike previous works, the proposed frameworks increase the dataset first to avoid the prediction bias caused by the limited data. The GAN-1D-CNN achieved 99.1% accuracy, specificity, sensitivity, F1-score, and 100% area under the curve (AUC). Similarly, the GAN-Bi-LSTM obtained 99.3% accuracy, 99.2% specificity, 99.3% sensitivity, 99.2% F1-score, and 100% AUC. Furthermore, time complexity of proposed frameworks is investigated with and without principal component analysis (PCA). The PCA method reduced prediction times for 61 samples using GAN-1D-CNN and GAN-Bi-LSTM to 68.8 and 74.8 ms, respectively. These results show that it is reliable to use our frameworks for augmenting limited data and predicting heart disease.