EDITORIAL

published: 04 March 2022

doi: 10.3389/fpubh.2022.852840

Editorial: Artificial Intelligence for

eHealth
Deepak Gupta 1*, Joel J. P. C. Rodrigues 2 , Sheng-Lung Peng 3 and Nhu Nguyen 4
1
Department of Computer Science and Engineering, Maharaja Agrasen Institute of Technology, New Delhi, India, 2 Senac
Faculty of Ceará, Fortaleza, Brazil, 3 Department of Creative Technologies and Product Design, National Taipei University of
Business, Taipei, Taiwan, 4 Duy Tan University, Da Nang, Vietnam

Keywords: artificial intelligence (AI), eHealth, computational intelligence, COVID-19, neural network

Editorial on the Research Topic

Editorial: Artificial Intelligence for eHealth

Artificial intelligence has grown extensively in recent times and is changing the healthcare industry
from many perspectives: Clinical Diagnosis, suggests treatment and follow up. Clinical Decision
Support (CDS) is a major topic of AI in medicine to assist clinicians at point of care. Existing
techniques used for processing health data can be broadly classified into two categories: (a)
non-Artificial Intelligence (AI) systems and (b) Artificial Intelligence systems. Even though non-AI
techniques are less complex in nature, most of the systems suffer from the drawbacks of inaccuracy
and lack of convergence. Hence, these systems are generally replaced by AI based systems which are
much superior to the conventional systems. AI techniques are mostly hybrid in nature and include
Edited by:
Artificial Neural Networks (ANN), fuzzy theory, and evolutionary algorithms. AI increases the
Bruno Bonnechère, ability for healthcare professionals to better understand the day-to-day patterns and needs of the
University of Oxford, United Kingdom people they care for, and with that understanding they are able to provide better feedback, guidance
Reviewed by: and support for staying healthy.
Dalibor Stanimirovic, AI-based CDS uses inference and logics, while non-AI-based CDS relies on machine learning
National Institute for Public to perform the same functions. There are many clinical duties that CDS may assist with, but it is
Health, Slovenia essential that CDS is correctly integrated into the clinical workflow and health records. CDS can
*Correspondence: be used to assist clinicians in the interpretation of medical pictures through the use of Computer
Deepak Gupta Aided Diagnosis (CAD). CAD incorporates AI as well as computer vision, signal processing, and
[email protected] other components relevant to medicine. Breast cancer, lung cancer, colon cancer, coronary artery
disease, and Alzheimer’s disease are just a few of the conditions that can benefit from CADs.
Specialty section: There are certain societal concerns about the expanding use of AI in healthcare, including the
This article was submitted to possibility of bias, lack of transparency for some AI algorithms, privacy problems for data used for
Digital Public Health,
AI model training, and security and implementation responsibilities in clinical settings.
a section of the journal
Frontiers in Public Health
All areas of artificial intelligence (AI) in the fields of health informatics, biomedical informatics,
and medical image analysis are covered in this special issue. Based on the reviews, eight papers were
Received: 11 January 2022
chosen from a total of fifteen submissions to this special collection. At least two reviewers and at
Accepted: 14 February 2022
Published: 04 March 2022
least two rounds of review were required for each paper. Listed below are some of the papers that
made important contributions to this discussion.
Citation:
Gupta D, Rodrigues JJPC, Peng S-L
Using data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) and independent
and Nguyen N (2022) Editorial: latent class analysis (LCA and LTA) and latent transition analysis (LTA) over a three-year period,
Artificial Intelligence for eHealth. the authors (Alashwal et al.), describe their findings in the first paper of this special issue.
Front. Public Health 10:852840. Researchers found that LCA was a better predictor of AD progression than typical clinical cut-off
doi: 10.3389/fpubh.2022.852840 measures on neuropsychological exams when it came to defining and recognizing the disease.

Frontiers in Public Health | www.frontiersin.org 1 March 2022 | Volume 10 | Article 852840

Gupta et al. Editorial: Artificial Intelligence for eHealth

Researchers (Mishra et al.) recommend data preparation to in the world in the next several days (ARIMA) (Dansana
eliminate the problem of data skewing by employing sample et al.). In addition to the 120,000 confirmed fatalities predicted
approaches. The K-Nearest Neighbor algorithm is used to classify by the ARIMA model until April 1, 2020, we also evaluated
the data using three different sampling techniques: Resampling, the total number of confirmed cases, the total number of
Spread Subsampling, and SMOTE. Evaluation of classification’s fatalities predicted, the autocorrelation function, and the white
performance is done using a variety of performance indicators to noise time series for the COVID-19 outbreak’s confirmed and
determine classification’s efficiency. fatalities cases.
The authors (Iwendi et al.) in this research offer a fine-tuned Patients infected with the COVID-19 virus can have their
AdaBoost algorithm-boosted Random Forest model. Data from medical issues diagnosed using a data mining model built on a
the COVID-19 patients is used to create a model that can estimate hybrid deep learning framework (Khadidos et al.). Convolution
the severity of a patient’s condition and whether or not they will neural networks (CNNs) and recurrent neural networks (RNNs)
recover or die. On the dataset used, the model has an accuracy of combine to form the DeepSense technique, a hybrid deep
94% and an F1 Score of 0.86. The study of the data shows a link learning model. In comparison to other deep learning and
between the gender of the patients and their mortality, as well as machine learning classifiers, DeepSense’s accuracy was shown
the fact that the vast majority of patients are between the ages of to be significantly higher. A patient’s prognosis for COVID-
20 and 70. 19 infections can be improved by knowing the accuracy of the
They are trying to find out if heat massaging of the spinal diagnostic approach used.
column can reduce muscle discomfort and increase antioxidant To summarize, eight of the fifteen papers submitted to
function in this study (Kim et al.). There were 60 people in the this special issue were accepted for publication in this
study who had lower back discomfort. Both an experimental special edition. We, as guest editors, believe that this special
and a control group were given spinal column heat massage issue’s research contributions and conclusions will assist
and normal rehabilitative treatment, respectively, as part of readers by expanding their knowledge and inspiring them to
their rehabilitation. According to the results of the study, work on a variety of elements of Artificial Intelligence for
spinal column thermal massage decreases pain more efficiently eHealth themselves.
and improves impairment levels. Because of this, thermal
massage may be beneficial in the treatment and prevention of AUTHOR CONTRIBUTIONS
oxidative disorders.
In the next paper (Song et al.) of this special collection All authors listed have made a substantial, direct, and intellectual
shows that the pectolinarin triggers apoptotic cell death in contribution to the work and approved it for publication.
PC12 cells by DNA fragmentation and the production of
apoptotic bodies via the activation of ER stress sensors (eIF2 ACKNOWLEDGMENTS
phosphorylation and ATF6 fragmentation) in PC12 cells. The
treatment of PC12 cells with 50 µM pectolinarin for 24 h We would like to thank the editor-in-chief for giving us the
increased the mRNA expression of ATF6, PERK, and IRE1 by up opportunity to put together this issue. A special thanks to the
to 1.6, 1.7, and 1.4 times, respectively, compared to the control. editing staffs for their assistance. We would also want to express
Pectolinarin administration enhanced ATF6 fragmentation by our gratitude to the authors and reviewers who contributed to
roughly twofold compared to the control, and phosphorylation this special edition.
of eIF2 by 2.5 fold. As a result of these findings, future natural
medicines and health supplements targeting disorders caused Conflict of Interest: The authors declare that the research was conducted in the
absence of any commercial or financial relationships that could be construed as a
by apoptosis could benefit from a better understanding of the
potential conflict of interest.
molecular pathways involved.
For the purpose of estimating the number of people who Publisher’s Note: All claims expressed in this article are solely those of the authors
will die from COVID-19-related causes in India over the next and do not necessarily represent those of their affiliated organizations, or those of
decade, the authors (Dhamodharavadhani et al.) conducted an the publisher, the editors and the reviewers. Any product that may be evaluated in
investigation into the suitability of Statistical Neural Network this article, or claim that may be made by its manufacturer, is not guaranteed or
(SNN) models and their hybrid version. These SNN models, endorsed by the publisher.
including the Probabilistic Neural Network and a Radial Basis
Function Neural Network, are used to construct the COVID-19 Copyright © 2022 Gupta, Rodrigues, Peng and Nguyen. This is an open-access article
Mortality Rate Prediction model (MRP) in India. MRP models distributed under the terms of the Creative Commons Attribution License (CC BY).
The use, distribution or reproduction in other forums is permitted, provided the
based on PNN and RBFNN were found to perform better than
original author(s) and the copyright owner(s) are credited and that the original
other models in COVID-19 datasets D2 and D1. publication in this journal is cited, in accordance with accepted academic practice.
Based on an autoregressive integrated moving average, a No use, distribution or reproduction is permitted which does not comply with these
model was developed to predict an epidemic of COVID-19 terms.

Frontiers in Public Health | www.frontiersin.org 2 March 2022 | Volume 10 | Article 852840