Artificial intelligence of medical things for disease detection using ensemble deep learning and attention mechanism

Abstract

In this paper, we present a novel paradigm for disease detection. We build an artificial intelligence based system where various biomedical data are retrieved from distributed and homogeneous sensors. We use different deep learning architectures (VGG16, RESNET, and DenseNet) with ensemble learning and attention mechanisms to study the interactions between different biomedical data to detect and diagnose diseases. We conduct extensive testing on biomedical data. The results show the benefits of using deep learning technologies in the field of artificial intelligence of medical things to diagnose diseases in the healthcare decision‐making process. For example, the disease detection rate using the proposed methodology achieves 92%, which is greatly improved compared to the higher‐level disease detection models.

Deep Learning Enabled Perceptive Wearable Sensor: An Interactive Gadget for Tracking Movement Disorder

“…The deep neural network contains several layers of nodes, and these are like neurons which are grouped into layers. [ 57,58 ] Accuracy is one of the important metric that is used to quantify the measurement of correct predictions/classifications. It can be calculated by the following equation $$$$\begin{eqnarray} {\rm{Accuracy}} &=& {{{\rm{Number}}\,\,{\rm{of}}\,\,{\rm{correct}}\,\,{\rm{predictions}}} \over {{\rm{Total}}\,\,{\rm{number}}\,\,{\rm{of}}\,\,{\rm{predictions}}}}\nonumber\\ &=& {{\sum {\rm{TP}} + {\rm{TN}}} \over {\sum {\rm{TP}} + {\rm{TN}} + {\rm{FP}} + {\rm{FN}}}}\end{eqnarray}$$$$…”

“…The deep neural network contains several layers of nodes, and these are like neurons which are grouped into layers. [57,58] Accuracy is one of the important metric that is used to quantify the measurement of correct predictions/classifications. The classification using the deep neural network with artificial neurons provides an efficiency of ≈94% (Figure 7c).…”

Predicting Childhood Obesity Based on Single and Multiple Well-Child Visit Data Using Machine Learning Classifiers

“…Ahmed et al developed seven ensemble models combining different ML algorithms, including naïve Bayes, decision tree, random tree, random forest, support vector classifier, logistic regression, adaptive boosting, gradient boosting, support vector machine with sequential minimal optimization, and JRip, to detect Android ransomware and to mitigate adversarial evasion attacks and found that the ensemble methods performed better than the individual ML algorithms [ 53 ]. Djenouri et al used several deep learning architectures such as VGG16, RESNET, and DenseNet with an efficient ensemble learning and attention mechanism and achieved a disease detection rate of 92% [ 54 ]. Once we obtain additional data with significantly more subjects, we will investigate hybrid ensemble methods and check whether they perform better than the ML algorithms proposed in this paper.…”

Machine learning and internet of things applications in enterprise architectures: Solutions, challenges, and open issues

“…It is possible to use IoT devices to distantly capture and relay this information to significant health data collection and diagnostic centers (e.g., cardiovascular, blood pressure, and anatomy data) (Luong et al (2016), Sharma, Al‐Wanain, et al (2022)). The most critical risks to healthcare facilities are data protection and privacy of the patient data collected for medical purposes (Djenouri et al (2022), Tariq, Asim, and Al‐Obeidat (2019)).…”