Patient Satisfaction: The Role of Artificial Intelligence in Healthcare

Abstract

Applications of artificial intelligence (AI) can be seen in almost every aspect of the healthcare system, as it has potential to affect almost every facet of the healthcare, from detection of ailments and serious or complex chronic diseases to their control, prevention and cure. With technological innovations, upgradation and adoption in the field of healthcare, healthcare professionals are required to be well prepared to accept the continuously evolving technology and its application to provide best healthcare facilities, which gave rise to the various studies on the role of the machine learning (ML), AI, deep learning (DL), etc., in the field of healthcare. Similarly, the rise in digitalised hospitals, medical facilities, records and data has resulted in the improvisation in the field of healthcare, which in turn has increased the need of experts, professionals, experienced and digitally literate workforce teams in the field of entire healthcare system. Understanding the roles of these advanced technologies, impacts being created on the health, lifestyle and the entire healthcare system, along with the perception of the patients towards it, will shape the way for the improvements and the applications of AI and its outcomes to be achieved, resulting in healthier world for the patients and the society. The objective of the study is to create a patient satisfaction model and validate it with respect to factors influencing patient satisfaction of several patients undergoing AI treatment factors. In the study, the United States, Canada, Australia, UAE and China were chosen as a place of survey, as these are advanced countries and the use of AI is highest in these countries compared to other countries, and survey was done with the help of structured questionnaire. In our earlier study, exploratory factor analysis (EFA) was performed for initial knowledge development on the construct of patients undergoing AI treatment. Patient satisfaction rests on six broad dimensions: First factor is personal touch (PT), second factor is comprehensive gap (CG), third factor is answerability (AB), fourth factor is nerve racking (NR), fifth factor is wrong reporting (WR) and sixth factor is enlightened (EL). With the help of confirmatory factor analysis (CFA) and structured equation modelling (SEM), it has emerged from the study that patient satisfaction level of the construct suggests that PT will have a greater impact on patient satisfaction, and it is the most significant factor of patient satisfaction compared to other constructs. Thus, we can conclude that PT still remains the most important factor in the minds of patients before undergoing AI treatment.

Keywords

Artificial intelligence patient satisfaction health structural equation modelling,treatment,awareness

Introduction

We are in the era of digital revolution marked with the fusion of various types of technology (Pang et al., 2018; Schwab, 2017). One of the tools that is leading nowadays in almost every field is artificial intelligence (AI) (Simon, 1991). Big data tools like AI and machine learning are one of the emerging tools which are being used to solve various problems being faced in current tech-savvy world (Le & Kong, 2011; Tan & Hu, 2018; Tan & Le, 2015, 2016; Tan et al., 2019; Thanh et al., 2010). AI and deep learning have emerged as an established tool to discover various interpretations from a raw data without relying on the central knowledge.

In spite of relying on the domain knowledge (Hinton et al., 2006; Joachims, 1999; Kingma et al., 2014; Le & Kong, 2011; Plötz et al., 2011; Tan & Hu, 2018; Tan & Le, 2015, 2016; Tan et al., 2019; Thanh et al., 2010), the application of AI can be seen in almost every field nowadays. More research is being done on its application in the field of healthcare and is being explored further to make its use to cure every possible disease and related issues.

AI is being used in the field of medicine and healthcare to cure complex issues and to make the healthcare system strong by improving and widening the services being provided by the healthcare bodies. In spite of its application in healthcare, we are far from its routine application of AI in healthcare as the potential risks of its application are greater than the benefits being provided by it (Hengstler et al., 2016). As per various researches, near about half of the world lacks proper healthcare facilities, and nearly about 100 million people are penniless and cannot afford better or improvised healthcare services (Sallstrom et al., 2019; World Bank & World Health Organization, 2017). Thus, the WHO as well as the countries started concentrating and investing in AI to bring the necessary evolution in the field of healthcare (House of Lords, 2018; National Science and Technology Council, 2016; World Bank & World Health Organization, 2017).

AI is being applied to detect various early diseases, getting idea about the diseases, for optimised medication, to discover innovative and novel treatments, etc. (Beam & Kohane, 2016; Bishnoi & Narayan, 2018; Fogel & Kvedar, 2018; Jiang et al., 2017; Miotto et al., 2017; Reddy et al., 2019; Topol, 2019). Strong point about AI is its ability to analyse and interpret the large and complex data sets. It is helpful in recognising various complex diseases like ophthalmology, cancer-detection, radiology, mental illness, etc. (Brinker et al., 2019; Hosny et al., 2018; Sengupta & Adjeroh, 2018; Vidal-Alaball et al., 2019). Since human capacity to learn and retain is limited, AI-based tools and machines can be used to retain huge amount of healthcare and medicinal data and sources. And these data sets are being used to analyse and understand human behaviour and patterns, which is quite difficult for humans to do (Wang et al., 2016).

Application of AI into healthcare reduces costs, improves healthcare services, decision making, controls diseases and provides required solutions for various health issues, thus improving and increasing the lifespan of the people, especially the older (Bughin et al., 2017; Hamet & Tremblay, 2017). Progress has been seen to the large extent in the field of healthcare due to AI through digitalised medical records, advanced machines and tools and health experts. It consists of strategies to bring out the development in infrastructure, health opportunities, literacy, research methods and opportunities, economic and legal policies (Canadian Institute for Advanced Research, 2018; China Institute for Science and Technology Policy, 2018; European Commission, 2018; Martinho-Truswell et al., 2018; National Institution for Transforming India Aayog, 2018; Strategic Council for AI Technology, 2017; Villiani, 2018).

In spite of making progress, AI is still facing some challenges in healthcare sector, impacting the healthcare system, organisations and the society (Bughin et al., 2017; Hamet & Tremblay, 2017). Lack of information about the data and limited idea about the probable output impact the healthcare system to a great extent. Similarly, application of AI involves huge amount of quality data, which is not possible on limited investment, and it gives rise to limited information or the biased information (Sallstrom et al., 2019; Wahl et al., 2018). Lack of proper infrastructure or expertise affects the governance of the models, data management, quality and safety of the data to be used. It increases the risk of discrimination, safety and privacy of the clinical data (Chen & Asch, 2017; Wahl et al., 2018). Apart from all these, AI is still inaccessible to many countries due to lack of required investments, and such people lack expertise in healthcare services.

It is necessary to make use of AI-based tools to overcome the barriers that can arise in healthcare or other fields, and above that, it should be eligible enough to be applied in order to find out the solutions. That is, it should give the ideal solution with minimum calculative complexities with maximum stability. Further, there is no machine learning or AI which can be applied to every practical problem or to any situation as it relies on the attributes of the collected data to be used. Thus, in such cases, it is required to make some modifications or innovations in the use of AI in different fields and to explore its various applications to make its optimum usage to discover solutions to the problems (Zahin & Hu, 2019).

Review of Literature

Integration of AI into healthcare has been marked with the remarkable revolution in the whole medical and healthcare system by supporting necessary decision making, complex or chronic diseases, supporting lives and the people in many ways (Bughin et al., 2017; Hamet & Tremblay, 2017). With the continuous technological innovations, healthcare system and hospitals are becoming digitalised with digital data, services, etc.; thus, it is necessary for the healthcare system to pace up with this advancement to manage the significant changes (Australian Institute of Health and Welfare, 2016; Khanna et al., 2013; You & Okunade, 2017). These innovations demand high-quality data in large amount to be used for healthcare services. That is why AI and other related tools are being deployed continuously on a large scale in order to process, interpret and use such huge data being available for the healthcare system. Various countries across the globe are adopting the use of AI in healthcare system in order to develop the infrastructure, researches, opportunities, literacy, etc. (Canadian Institute for Advanced Research, 2018; China Institute for Science and Technology Policy, 2018; European Commission, 2018; Government of the Republic of Korea, 2016; House of Lords, 2018; Martinho-Truswell et al., 2018; Ministry of Economic Affairs Employment of Finland, 2017; National Institution for Transforming India Aayog, 2018; National Research Foundation, 2018; National Science and Technology Council, 2016; Nordic Council of Sweden, 2018; Strategic Council for AI Technology, 2017; The Agency for Digital Italy, 2018; The Danish Government, 2018; UAE, 2017; Villiani et al., 2018; Vinnova, 2018). In 2016, the Australian government created a National Digital Health Strategy in order to produce effective healthcare service with minimum expenditure (Australia’s National Digital Health Strategy, 2018). With further application of AI in healthcare, it can deeply potentially reshape the role of healthcare professionals and their beliefs, assumptions and values in the healthcare system (Erikson & Salzmann-Erikson, 2016).

AI is being used widely in healthcare to create and store data related to medicines, health and the patients in the hospitals. These data in turn are being used by the laboratories in order to carry out the tests and researches on patient’s health and medicinal applications. These data are not limited to the hospitals only but are being used by different laboratories, research organisations, institutes or different IT companies to carry out researches on the application of data for various tasks and wide application of AI in this field (Winter & Davidson, 2019). Data available in healthcare sector are used on the basis of their models to be examined, which ascertain the use of the data, purpose behind the use, quantity and quality of the data to be used, anticipated outcomes, user using the data, etc. An information is the set of data to be used by the organisations to carry out the researches on a large scale these days, and these data are kept in various forms for analysing the further applications of AI in healthcare sector with the help of these data (Data Governance Institute, n.d.; Khatri & Brown, 2010; Perkmann & Schildt, 2015; Susha et al., 2017).

AI acts as a saviour for the medical science as it has proved itself to be the blessing in complex cases and diseases in which all other calculative and statistical tools failed to give desired results (Campbell, 2014; Grossi, 2011; Inza et al., 2010). There have been many successful cases, which have opened the way for AI to be explored further to bring evolution in the whole healthcare system. Through proper decision making and careful application, AI has been applied in cancer detection, recognising and understanding incurable diseases, identifying the possible solutions for them, curing methods for breast and other cancers, identifying the reasons and solutions for mental illness, psychological disorders, fatigue, heart and kidney issues, etc. (Bennett & Hauser, 2013; Daoud & Mayo, 2019; Delen et al., 2004; Lamy et al., 2019; Pereira et al., 2019; Shaikhina et al., 2015).

Till date, AI has been applied mainly for cancer and cardiovascular diseases, as these are the major cause behind the rising mortality rate, along with infectious and other chronic diseases. Early treatment of these diseases is now possible and improving due to continuous extraction of clinical understanding from AI and using it in turn in a well-managed system (Bassaganya-Riera & Hontecillas, 2018; Jiang et al., 2017; Kagawa et al., 2016; Leber et al., 2017). According to Abedi et al. (2017), AI models can be used to diagnose the acute cerebral ischemia rather than trained emergency medical respondents.

Though limited experimentation and variability in the data can reduce the quality of utility being derived, which in turn can be controlled with the help of AI and such related tools (Noorbakhsh-Sabet et al., 2019). AI can be used in many complex and chronic diseases with the help of proper infrastructure and tools (Gulshan et al., 2016). Another most important goal of AI is to optimise the medicines or drugs in the health sector to cure various issues with the help of researches on its wide applications and usages (Haverty et al., 2016). AI has been seen as a key element in recognising reliable and useful outcomes in healthcare system (BouAssi et al., 2017; Fergus et al., 2015). Now, it is possible to make use of deep learning as well in order to carry out the researches in the field of healthcare, and this platform can be used in mobile systems as well (Kiral-Kornek et al., 2018; Stacey, 2018). AI can also be used to diagnose the diseases based on biomedical image processing (Stoitsis et al., 2006), and it has been deployed in image segmentation, multidimensional imaging and thermal imaging in order to bring development in the quality and efficiency in image and analysis (Ghafarpour et al., 2016; Jo et al., 2019). AI can also be seen as an emerging tool in portable ultrasound tools, and it can be used by the untrained persons as well to cure many complex diseases, especially in underdeveloped countries or regions (Personal Ultrasound, 2017). Apart from all these, AI can also be used along with standard decision support systems (DSSs) in order to improve the efficiency and decision making in order to avoid burdening a single person (Elkin et al., 2018; Safdar et al., 2018).

Despite of so many successful outcomes and approaches, AI has faced certain limitations and challenges as well. These include those innate to science of machine learning, difficulties being faced in implementing the logistics, social and cultural barriers, etc. (Kelly et al., 2019). Apart from this, high cost of maintenance, need of complex and huge amount of data for reliable and effective outcomes, high-level supervision to reduce the risk of misuse of the data to the biased outcomes, etc., are all the challenges being faced while using AI in healthcare sector.

Till date, AI has been adopted in almost every field in order to explore and understand the scope of the technology and its implications further, especially in healthcare system. AI has the ability to take the healthcare system to a different level if implied properly, thus improving medical practices and roles of healthcare professionals (Gagnon et al., 2016; Hsieh et al., 2012; Rho et al., 2014). Along with technological innovations, AI should also be used to understand human behaviour and the impact of technology on the person using it as much as possible, to understand the behaviour and perception of the people towards healthcare and the extent to which healthcare professionals are connected with the healthcare system, affecting their roles and ability to deliver services. Thus, with limited reach of AI into healthcare and other related fields, there is need to understand the perceptions and experiences of the people in order to explore the use of AI in various fields (Shinners et al., 2020).

Objective

To identify a conceptual model for understanding the factors influencing patient satisfaction of several patients undergoing AI treatment.

Research Methodology

In the present study, the United States, Canada, Australia, UAE and China were chosen as a place of survey as these are advanced countries and the use of AI is highest in these countries compared to other countries. People in these countries prefer AI-based treatments over other methods of treatment. Thus, 249 samples were collected from different locations in the United States, Canada, Australia, UAE and China with the help of structured questionnaire, which were sent to the respondents through email. Initially, around 40 quality parameters were identified for determining customers’ awareness and preferences about AI in healthcare. Five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree) as used by Cronin and Taylor (1992) was introduced for the measurement of each parameter. Along with these parameters, demographic and psychographic variables were also recorded for the respondents. Data after proper cleaning and validation were used for several multivariate analyses to attain the objectives of the study.

Findings and Analysis

Measurement Model

Confirmatory factor analysis (CFA) is commonly used in social research. Both exploratory and CFA are engaged to understand shared variance of variables. The overall fit of the model was assessed by chi square (χ²), goodness of fit index (GFI), adjusted goodness of fit index (AGFI), comparative fit index (CFI) and root mean square error of approximation (RMSEA). Hu and Bentler (1999) stated that GFI values over 0.9 and AGFI values over 0.8 indicate good data fitting. RMSEA value lies between 0.05 and 0.10 and is considered as an indication of fair fit, and values above 0.10 indicated poor fit. Composite reliability (CR) measures the consistency of content construct indicators, the higher CR indicates that potential variables are internally consistent; and the values should be greater than 0.5.

From Table 1, we can see that we are getting satisfied results for the CFA.

Table 1.

Goodness of Fit Indices (CFA).

CMIN/DF	P	GFI	AGFI	NFI	TLI	CFI	PCLOSE	RMSEA
2.116	0.000	0.831	0.759	0.829	0.888	0.901	0.000	0.067

Note: AGFI, adjusted goodness-of-fit index; CFI, comparative fit index; GFI, goodness-of-fit index; NFI, normed fit index; RMSEA, root mean square error of approximation; TLI, Tucker–Lewis index.

Thus, from Table 2, we can conclude that congruence validity is accepted for this model, except for the construct nerve racking (NR), which has a very low value. From Table 3 too, we can see that average variance extracted (AVE) and CR are very less for the construct NR. From Figure 1, we can also conclude that convergent validity is also accepted as the variables associated with the factors are greater than 0.7.

Table 2.

Discriminant Validity.

	CG	AB	NR	WR	PT
CG	0.74
AB	0.266	0.89
NR	0.211	0.685	0.23
WR	0.382	0.583	0.619	0.88
PT	0.698	0.698	0.723	0.576	0.93

Note: AB, answerability; CG, comprehensive gap; NR, nerve racking; PT, personal touch; WR, wrong reporting.

Bold values indicate the square root of the AVE (main diagonal) is in all cases superior to the correlations among the constructs, which shows discriminant validity.

Table 3.

Results of the Measurement Model.

Measurement Items			AVE	CR
V20	←	CG
V19	←	CG
V18	←	CG	0.74	0.86
V17	←	CG
V21	←	CG
V22	←	CG
V14	←	AB
V13	←	AB	0.89	0.78
V12	←	AB
V11	←	AB
V3	←	NR
V4	←	NR	0.23	0.36
V5	←	NR
V6	←	NR
V7	←	WR
V9	←	WR	0.88	0.80
V15	←	WR
V16	←	PT
V23	←	PT
V24	←	PT
V25	←	PT
V26	←	PT	0.93	0.74
V27	←	PT
V28	←	PT
V30	←	PT

Note: AB, answerability; AVE, average variance extracted; CG, comprehensive gap; CR, composite reliability; NR, nerve racking; PT, personal touch; WR, wrong reporting.

Figure 1.

Note: AB, answerability; AVE, average variance extracted; CG, comprehensive gap; CR, composite reliability; NR, nerve racking; PT, personal touch; WR, wrong reporting.

Figure 2.

Note: AB, answerability; AVE, average variance extracted; CG, comprehensive gap; CR, composite reliability; NR, nerve racking; PT, personal touch; SATF: patient satisfaction; WR, wrong reporting.

From Table 4, we can see that we are getting satisfied results for structured equation modelling. The structural model of AI suggests that personal touch (PT) will have a greater impact on patient satisfaction, and it is the most significant factor of patient satisfaction as the regression weight is 0.44. The study has shown that to a certain extent, there is also a positive relationship between patient satisfaction and the construct answerability (AB), construct wrong reporting (WR) and also construct comprehensive gap (CG), as the regression weights are 0.17, 0.09 and 0.03. Thus, it may be hypothesised that patients in general consider PT plays a very important role in determining patient satisfaction.

Table 4.

Goodness of Fitness (SEM).

CMIN/DF	P	GFI	AGFI	NFI	TLI	CFI	PCLOSE	RMSEA
3.661	.000	0.80	0.79	0.822	0.82	0.90	.000	0.101

Note: AGFI, adjusted goodness-of-fit index; CFI, comparative fit index; GFI, goodness-of-fit index; NFI, normed fit index; RMSEA, root mean square error of approximation; TLI, Tucker–Lewis index.

Conclusion

AI streamlines and eases the lives of patients, doctors, etc., by doing tasks that are done by human beings, with less time investment and very low cost. The AI sector is projected to reach $150 billion by 2026, which shall be empowered as high-growth industry. Modern healthcare, with the help of machines, shall predict, embrace, grasp and proceed. Though PT still remains the most important factor because at the end of day, doctors give a greater understanding about the entirety of their life and how to help patients make decisions. Patient means going through the diagnoses, treatments, operations, etc., but each patient also has different life story that goes far beyond the medical treatment. AI seriously making inroads, but it seems AI in healthcare still remains to be challenging in critical circumstances.

Footnotes

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

References

Abedi

, Goyal

, Tsivgoulis

, Hosseinichimeh

, Hontecillas

, Bassaganya-Riera

, Elijovich

, Metter

J. E.

, Alexandrov

A. W.

, Liebeskind

D. S.

, Alexandrov

A. V.

, & Zand

(2017). Novel screening tool for stroke using artificial neural network. Stroke , 48(6), 1678–1681. https://doi.org/10.1161/STROKEAHA.117.017033

Australia’s National Digital Health Strategy. (2018). Safe, seamless and secure: Evolving health and care to meet the needs of modern Australia.

Australian Institute of Health and Welfare. (2016). Australia’s health 2016 (Australia’s health no. 15. Cat. no. AUS 199). AIHW.

Bassaganya-Riera

, & Hontecillas

(2018). Introduction to accelerated path to cures and precision medicine in inflammatory bowel disease. In Accelerated path to cures (pp. 1–6). Springer. https://doi.org/10.1007/978-3-319-73238-1_1

Beam

A. L.

, & Kohane

I. S.

(2016). Translating artificial intelligence into clinical care. JAMA , 316(22), 2368–2369. https://doi.org/10.1001/jama.2016.17217

Bennett

C. C.

, & Hauser

(2013). Artificial intelligence framework for simulating clinical decision-making: A Markov decision process approach. Artificial Intelligence in Medicine , 57, 9–19. https://doi.org/10.1016/j.artmed.2012.12.003

Bishnoi

, & Singh

N. S.

(2018). Artificial intelligence techniques used in medical sciences: A review. In 8th international conference on cloud computing, data science & engineering (confluence) (pp. 106–113). https://doi.org/10.1109/CONFLUENCE.2018.8442729

Brinker

T. J.

, Hekler

, Hauschild

, Berking

, Schilling

, Enk

A. H.

, Haferkamp

, Karoglan

, Kalle

C. V.

, Weichenthal

, Sattler

, Schadendorf

, Gaiser

M. R.

, Klode

, & Utikal

J. S.

(2019). Comparing artificial intelligence algorithms to 157 German dermatologists: The melanoma classification benchmark. European Journal of Cancer , 111, 30–37. https://doi.org/10.1016/j.ejca.2018.12.016

BouAssi

, Nguyen

D. K.

, Rihana

, & Sawan

(2017). Towards accurate prediction of epileptic seizures: A review. Biomed Signal Process Control , 34, 144–157. https://doi.org/10.1016/j.bspc.2017.02.001

10.

Bughin

, Hazan

, Ramaswamy

, Allas

, Dahlström

, Henke

, & Trench

(2017). Artificial intelligence: The next digital frontier . McKinsey Global Institute.

11.

Campbell

(2014). Machine learning methodology in bioinformatics. In Springer handbook of bio-/neuroinformatics (pp. 185–206). Springer. https://doi.org/10.1007/978-3-642-30574-0_12

12.

Canadian Institute for Advanced Research. (2018). CIFAR Pan-Canadian artificial intelligence strategy . https://www.cifar.ca/ai/pan-canadian-artificial-intelligence-strategy

13.

Chen

J. H.

, & Asch

S. M.

(2017). Machine learning and prediction in medicine—beyond the peak of inflated expectations. The New England Journal of Medicine , 376(26), 2507. https://doi.org/10.1056%2FNEJMp1702071

14.

China Institute for Science and Technology Policy. (2018). China AI development report . Tsinghua University.

15.

Cronin

J. J.

Jr , & Taylor

S. A.

(1992). Measuring service quality: A reexamination and extension. Journal of Marketing , 56(3), 55–68.

16.

Daoud

, & Mayo

(2019). A survey of neural network-based cancer prediction models from microarray data. Artificial Intelligence in Medicine , 97, 204–214. https://doi.org/10.1016/j.artmed.2019.01.006

17.

Data Governance Institute. (n.d.). Data governance definition .

18.

, Walker

, & Kadam

(2005). Predicting breast cancer survivability: A comparison of three data mining methods. Artificial Intelligence in Medicine , 34(2), 113–127. https://doi.org/10.1016/j.artmed.2004.07.002

19.

Elkin

P. L.

, Schlegel

D. R.

, Anderson

, Komm

, Ficheur

, & Bisson

(2018). Artificial intelligence: Bayesian versus heuristic method for diagnostic decision support. Applied Clinical Informatics , 9(2), 432. https://dx-doi-org-s.web.bisu.edu.cn/10.1055%2Fs-0038-1656547

20.

Erikson

, & Salzmann-Erikson

(2016). Future challenges of robotics and artificial intelligence in nursing: What can we learn from monsters in popular culture? The Permanente Journal , 20(3), 15–243. https://dx-doi-org-s.web.bisu.edu.cn/10.7812%2FTPP%2F15-243

21.

European Commission. (2018). Artificial intelligence for Europe . European Commission.

22.

Fergus

, Hignett

, Hussain

, Al-Jumeily

, & Abdel-Aziz

(2015). Automatic epileptic seizure detection using scalp EEG and advanced artificial intelligence techniques. BioMed Research International , 2015. https://doi.org/10.1155/2015/986736

23.

Fogel

A. L.

, & Kvedar

J. C.

(2018). Artificial intelligence powers digital medicine. NPJ Digital Medicine , 1(1), 1–4. https://doi.org/10.1038/s41746-017-0012-2

24.

Gagnon

M. P.

, Simonyan

, Godin

, Labrecque

, Ouimet

, & Rousseau

(2016). Factors influencing electronic health record adoption by physicians: A multilevel analysis. International Journal of Information Management , 36(3), 258–270. https://doi.org/10.1016/j.ijinfomgt.2015.12.002

25.

Ghafarpour

, Zare

, Zadeh

H. G.

, Haddadnia

, Zadeh

F. J. S.

, Zadeh

Z. E.

, Kianersi

, Masoumzadeh

, & Nour

(2016). A review of the dedicated studies to breast cancer diagnosis by thermal imaging in the fields of medical and artificial intelligence sciences. Biomedical Research , 27(2).

26.

Government of the Republic of Korea. (2016). Mid-to-long term master plan in preparation for the intelligent information society: Managing the fourth industrial revolution . Government Republic of Korea.

27.

Grossi

(2011). Artificial neural networks and predictive medicine: A revolutionary paradigm shift. In Artificial neural networks—methodological advances and biomedical applications (pp. 139–150). InTech.

28.

Gulshan

, Peng

, Coram

, Stumpe

M. C.

, Wu

, Narayanaswamy

, Venugopalan

, Winder

, Madams

, Cuadros

, Kim

, Raman

, Nelson

P. C.

, Mega

J. L.

, & Webster

D. R.

(2016). Development and validation of a deep learning algorithm for detection of diabetic retinopathy in retinal fundus photographs. JAMA , 316(22), 2402–2410. https://doi.org/10.1001/jama.2016.17216

29.

Hamet

, & Tremblay

(2017). Artificial intelligence in medicine. Metabolism , 69, S36–S40. https://doi.org/10.1016/j.metabol.2017.01.011

30.

Haverty

P. M.

, Lin

, Tan

, Yu

, Lam

, Lianoglou

, Neve

R. M.

, Martin

, Settleman

, Yauch

R. L.

, & Bourgon

(2016). Reproducible pharmacogenomic profiling of cancer cell line panels. Nature , 533(7603), 333–337. https://doi.org/10.1038/nature17987

31.

Hengstler

, Enkel

, & Duelli

(2016). Applied artificial intelligence and trust—the case of autonomous vehicles and medical assistance devices. Technological Forecasting and Social Change , 105, 105–120. https://doi.org/10.1016/j.techfore.2015.12.014

32.

Hinton

, Osindero

, & Teh

Y. W.

(2006). A fast learning algorithm for deep belief nets. Neural Computing , 18(7), 1527–1554. https://doi.org/10.1162/neco.2006.18.7.1527

33.

Hosny

, Parmar

, Quackenbush

, Schwartz

L. H.

, & Aerts

H. J.

(2018). Artificial intelligence in radiology. Nature Reviews Cancer , 18(8), 500–510. https://doi.org/10.1038/s41568-018-0016-5

34.

House of Lords. (2018). AI in the UK: Ready, willing and able? House of Lords.

35.

Hsieh

S. T.

, Chen

C. D.

, Chen

J. H.

, Lai

C. L.

, & Syu

Y. L.

(2012). An expert system application for respiratory infection diagnostic. In 2012 9th international conference on ubiquitous intelligence and computing and 9th international conference on autonomic and trusted computing (pp. 673–678). IEEE. https://doi.org/10.1109/UIC-ATC.2012.87

36.

Inza

, Calvo

, Armañanzas

, Bengoetxea

, Larrañaga

, & Lozano

J. A.

(2010). Machine learning: An indispensable tool in bioinformatics. In Bioinformatics methods in clinical research (pp. 25–48). Humana Press. https://doi.org/10.1007/978-1-60327-194-3_2

37.

Jiang

, Jiang

, Zhi

, Dong

, Li

, Ma

, Wang

, Dong

, Shen

, & Wang

(2017). Artificial intelligence in healthcare: Past, present and future. Stroke and Vascular Neurology , 2(4), 230–243. https://doi.org/10.1136/svn-2017-000101

38.

Joachims

(1999). Transductive inference for text classification using support vector machines. In ICML: Proceedings of the sixteenth international conference on machine learning (Vol. 99, pp. 200–209).

39.

, Cho

, Lee

S. Y.

, Choi

, Kim

, Min

H. S.

, & Park

(2018). Quantitative phase imaging and artificial intelligence: A review. IEEE Journal of Selected Topics in Quantum Electronics , 25(1), 1–14. https://doi.org/10.1109/JSTQE.2018.2859234

40.

Kagawa

, Kawazoe

, Ida

, Shinohara

, Tanaka

, Imai

, & Ohe

(2016). Development of type 2 diabetes mellitus phenotyping framework using expert knowledge and machine learning approach. Journal of Diabetes Science and Technology , 11, 791–799. https://doi.org/10.1177%2F1932296816681584

41.

Kelly

C. J.

, Karthikesalingam

, Suleyman

, Corrado

, & King

(2019). Key challenges for delivering clinical impact with artificial intelligence. BMC Medicine , 17(1), 195. https://doi.org/10.1186/s12916-019-1426-2

42.

Khanna

, Sattar

, & Hansen

(2013). Artificial intelligence in health–the three big challenges. The Australasian Medical Journal , 6(5), 315. https://dx-doi-org-s.web.bisu.edu.cn/10.4066%2FAMJ.2013.1758

43.

Khatri

, & Brown

C. V.

(2010). Designing data governance. Communications of the ACM , 53(1), 148–152. https://doi.org/10.1145/1629175.1629210

44.

Kingma

D. P.

, Mohamed

, Jimenez Rezende

, & Welling

(2014). Semi-supervised learning with deep generative models. Advances in Neural Information Processing Systems , 27, 3581–3589.

45.

Kiral-Kornek

, Roy

, Nurse

, Mashford

, Karoly

, Carroll

, Payne

, Saha

, Baldassano

, O’Brien

, Grayden

, Cook

, Freestone

, & Harrer

(2018). Epileptic seizure prediction using big data and deep learning: Toward a mobile system. EBioMedicine , 27, 103–111. https://doi.org/10.1016/j.ebiom.2017.11.032

46.

Lamy

, Sekarb

, Guezennec

, Bouaud

, & Séroussia

(2019). Explainable artificial intelligence for breast cancer: A visual case-based reasoning approach. Artificial Intelligence in Medicine , 94, 42–53. https://doi.org/10.1016/j.artmed.2019.01.001

47.

Leber

, Hontecillas

, Abedi

, Tubau-Juni

, Zoccoli-Rodriguez

, Stewart

, & Bassaganya-Riera

(2017). Modeling new immunoregulatory therapeutics as antimicrobial alternatives for treating Clostridium difficile infection. Artificial Intelligence in Medicine , 78, 1–13. https://doi.org/10.1016/j.artmed.2017.05.003

48.

T. T.

, & Kong

H. Y.

(2011). Using subspace pursuit algorithm to improve performance of the distributed compressive wide-band spectrum sensing. Journal of the Korean Institute of Electromagnetic and Science , 11(4), 250–256. https://doi.org/10.5515/JKIEES.2011.11.4.250

49.

Martinho-Truswell

, Miller

, Nti

, Petheram

, & Stirling

(2018). Towards an AI strategy in Mexico: Harnessing the AI revolution . British Embassy Mexico City, Oxford Insights and C Minds.

50.

Ministry of Economic Affairs Employment of Finland. (2017). Finland’s age of artificial intelligence: Turning Finland into a leading country in the application of artificial intelligence . Publications of Ministry of Economic Affairs and Employment.

51.

Miotto

, Wang

, Jiang

, & Dudley

J. T.

(2017). Deep learning for healthcare: Review, opportunities and challenges. Briefings in Bioinformatics , 19(6), 1236–1246. https://doi.org/10.1093/bib/bbx044

52.

National Institution for Transforming India Aayog. (2018). National strategy for artificial intelligence . National Institution for Transforming India.

53.

National Research Foundation. (2018). AI Singapore. https://www.aisingapore.org/media/factsheet/

54.

National Science and Technology Council. (2016). Preparing for the future of artificial intelligence . Executive Office of the President.

55.

Noorbakhsh-Sabet

, Zand

, Zhang

, & Abedi

(2019). Artificial intelligence transforms the future of health care. The American Journal of Medicine , 132(7), 795–801. https://doi.org/10.1016/j.amjmed.2019.01.017

56.

Nordic Council of Sweden. (2018). AI in the Nordic-Baltic region . Nordic Council of Ministers.

57.

Pang

, Yuan

, Zhang

Y. T.

, & Packirisamy

(2018). Guest editorial health engineering driven by the industry 4.0 for aging society. IEEE Journal of Biomedical and Health Informatics , 22(6), 1709–1710. https://doi.org/10.1109/JBHI.2018.2874081

58.

Pereira

C. R.

, Pereira

D. R.

, Weber

S. A.

, Hook

, de Albuquerque

V. H. C.

, & Papa

J. P.

(2019). A survey on computer-assisted Parkinson’s disease diagnosis. Artificial Intelligence in Medicine , 95, 48–63. https://doi.org/10.1016/j.artmed.2018.08.007

59.

Perkmann

, & Schildt

(2015). Open data partnerships between firms and universities: The role of boundary organizations. Research Policy , 44(5), 1133–1143. https://doi.org/10.1016/j.respol.2014.12.006

60.

Personal Ultrasound. (2017). Butterfly network; c2019 . [Internet]. https://english.butterflynetwork.com/

61.

Plötz

, Hammerla

N. Y.

, & Olivier

P. L.

(2011). Feature learning for activity recognition in ubiquitous computing. In Twenty-second international joint conference on artificial intelligence .

62.

Reddy

, Fox

, & Purohit

M. P.

(2019). Artificial intelligence-enabled healthcare delivery. Journal of the Royal Society of Medicine , 112(1), 22–28. https://doi.org/10.1177/0141076818815510

63.

Rho

M. J.

, Young Choi

, & Lee

(2014). Predictive factors of telemedicine service acceptance and behavioral intention of physicians. International Journal of Medical Informatics , 83(8), 559–571. https://doi.org/10.1016/j.ijmedinf.2014.05.005

64.

Safdar

, Zafar

, & Khan

N. F.

(2018). Machine learning based decision support systems (DSS) for heart disease diagnosis: A review. Artificial Intelligence Review , 50(4), 597–623. https://doi.org/10.1007/s10462-017-9552-8

65.

Sallstrom

, Morris

, & Mehta

(2019). Ethical considerations: Artificial intelligence in Africa’s healthcare. Observer Research Foundation.

66.

Schwab

(2017). The fourth industrial revolution (p. 192). Penguin.

67.

Sengupta

P. P.

, & Adjeroh

D. A.

(2018). Will artificial intelligence replace the human echocardiographer? Clinical Considerations , 138(16), 1639–1642. https://doi.org/10.1161/CIRCULATIONAHA.118.037095

68.

Simon

H. A.

(1991). Artificial intelligence: Where has it been, and where is it going? IEEE Transaction on Knowledge and Data Engineering , 3(2), 128–136. https://doi.org/10.1109/69.87993

69.

Shaikhina

, Lowe

, Daga

, Briggs

, Higgins

, & Khovanova

(2015). Machine learning for predictive modelling based on small data in biomedical engineering. IFAC-PapersOnLine , 48(20), 469–474. https://doi.org/10.1016/j.ifacol.2015.10.185

70.

Shinners

, Aggar

, Grace

, & Smith

(2020). Exploring healthcare professionals’ understanding and experiences of artificial intelligence technology use in the delivery of healthcare: An integrative review. Health Informatics Journal , 26(2), 1225–1236. https://doi.org/10.1177%2F1460458219874641

71.

Stacey

W. C.

(2018). Seizure prediction is possible—now let’s make it practical. EbioMedicine , 27, 3–4. https://doi.org/10.1016/j.ebiom.2018.01.006

72.

Stoitsis

, Valavanis

, Mougiakakou

S. G.

, Golemati

, Nikita

, & Nikita

K. S.

(2006). Computer aided diagnosis based on medical image processing and artificial intelligence methods. Nuclear Instruments and Methods in Physics Research Section A , 569(2), 591–595. https://doi.org/10.1016/j.nima.2006.08.134

73.

Strategic Council for AI Technology. (2017). Artificial intelligence technology strategy . Strategic Council for AI Technology.

74.

Susha

, Janssen

, & Verhulst

(2017). “Data collaboratives as ‘bazaars’? A review of coordination problems and mechanisms to match demand for data with supply”. Transforming Government: People, Process and Policy , 11(1), 157–172. https://doi.org/10.1108/TG-01-2017-0007

75.

Tan

L. T.

, & Le

L. B.

(2015). Compressed sensing based data processing and MAC protocol design for smartgrids. In 2015 IEEE wireless communications and networking conference (WCNC) (pp. 2138–2143). IEEE. https://doi.org/10.1109/WCNC.2015.7127798

76.

Tan

L. T.

, & Le

L. B.

(2016). Joint data compression and MAC protocol design for smartgrids with renewable energy. Wireless Communications and Mobile Computing , 16(16), 2590–2604. https://doi.org/10.1002/wcm.2710

77.

Tan

L. T.

, & Hu

R. Q.

(2018). Mobility-aware edge caching and computing in vehicle networks: A deep reinforcement learning. IEEE Transactions on Vehicular Technology , 67(11), 10190–10203. https://doi.org/10.1109/TVT.2018.2867191

78.

Tan

L. T.

, Hu

R. Q.

, & Hanzo

(2019). Twin-timescale artificial intelligence aided mobility-aware edge caching and computing in vehicular networks. IEEE Transactions on Vehicular Technology , 68(4), 3086–3099. https://doi.org/10.1109/TVT.2019.2893898

79.

Thanh

T. L.

, Yun

K. H.

, & Quoc

B. V.

(2010). Projected Barzilai–Borwein methods applied to distributed compressive spectrum sensing. In 2010 IEEE symposium on new frontiers in dynamic spectrum (DySPAN) (pp. 1–7), IEEE. https://doi.org/10.1109/DYSPAN.2010.5457886

80.

The Agency for Digital Italy. (2018). Artificial intelligence at the service of citizens . AGID.

81.

The Danish Government. (2018). Strategy for Denmark’s digital growth . The Danish Government.

82.

Topol

E. J.

(2019). High-performance medicine: The convergence of human and artificial intelligence. Nature Medicine , 25(1), 44–56. https://doi.org/10.1038/s41591-018-0300-7

83.

UAE. (2017). Artificial intelligence strategy . http://www.uaeai.ae/en/

84.

Vidal-Alaball

, Fibla

R. D.

, Zapata

M. A.

, Marin-Gomez

F. X.

, & Solans

F. O.

(2019). Artificial intelligence for the detection of diabetic retinopathy in primary care: Protocol for algorithm development. JMIR Research Protocols , 8(2), e12539. https://doi.org/10.2196/12539

85.

Villiani

, Bonnet

, Rondepierre

, Schoenauer

, Berthet

, Levin

, & Cornut

A. C.

(2018). For a meaningful artificial intelligence: Towards a French and European strategy . Conseil national du numérique.

86.

Vinnova. (2018). Artificial intelligence in Swedish business and society: Analysis of development and potential . Vinnova.

87.

Wahl

, Cossy-Gantner

, Germann

, & Schwalbe

N. R.

(2018). Artificial intelligence (AI) and global health: How can AI contribute to health in resource-poor settings? BMJ Global Health , 3(4), e000798. https://dx-doi-org.web.bisu.edu.cn/10.1136/bmjgh-2018-000798

88.

Wang

, Kung

L. A.

, & Byrd

T. A.

(2016). Big data analytics: Understanding its capabilities and potential benefits for healthcare organizations. Technological Forecasting and Social Change , 126, 3–13. https://doi.org/10.1016/j.techfore.2015.12.019

89.

Winter

J. S.

, & Davidson

(2019). Governance of artificial intelligence and personal health information. Digital Policy, Regulation and Governance , 21(3), 280–290. https://doi.org/10.1108/DPRG-08-2018-0048

90.

World Bank, & World Health Organization. (2017). Half the world lacks access to essential health services, 100 million still pushed into extreme poverty because of health expenses.

91.

You

, & Okunade

A. A.

(2017). Income and technology as drivers of Australian healthcare expenditures. Health Economics , 26(7), 853–862.

92.

Zahin

, & Hu

R. Q.

(2019). Sensor-based human activity recognition for smart healthcare: A semi-supervised machine learning. In International conference on artificial intelligence for communications and networks (pp. 450–472). Springer. https://doi.org/10.1007/978-3-030-22971-9_39