As stated by Andrews [ 48 ], the figure is not intended to find the perfect level of associations between clusters. However, it aims to estimate the approximate number of clusters to facilitate further discussion. The research stream of AI in healthcare is divided into two main strands. The blue strand focuses on medical information systems and the internet. Some papers are related to healthcare organisations, such as the Internet of Things, meaning that healthcare organisations use AI to support health services management and data analysis.
AI applications are also used to improve diagnostic and therapeutic accuracy and the overall clinical treatment process [ 2 ]. If we consider the second block, the red one, three different clusters highlight separate aspects of the topic.
The first could be explained as AI and ML predictive algorithms. Through AI applications, it is possible to obtain a predictive approach that can ensure that patients are better monitored. This also allows a better understanding of risk perception for doctors and medical researchers. In the second cluster, the most frequent words are decisions , information system , and support system.
This means that AI applications can support doctors and medical researchers in decision-making. Information coming from AI technologies can be used to consider difficult problems and support a more straightforward and rapid decision-making process. In the third cluster, it is vital to highlight that the ML model can deal with vast amounts of data. From those inputs, it can return outcomes that can optimise the work of healthcare organisations and scheduling of medical activities. Furthermore, the word cloud in Fig.
The figure depicts how AI is linked to healthcare and how it is used in medicine. Figure 9 represents the search trends based on the keywords analysed. The research started in First, it identified research topics related to clinical decision support systems. This topic was recurrent during the following years. Interestingly, in , studies investigated AI and natural language processes as possible tools to manage patients and administrative elements. Table 9 represents the number of citations from other articles within the top 20 rankings.
The analysis allows the benchmark studies in the field to be identified [ 48 ]. For instance, Burke et al. The paper critically evaluates tangible interdisciplinary solutions that also include AI. Immediately thereafter, Ahmed M. Finally, the third most cited article lays the groundwork for developing deep learning by considering diverse health and administrative information [ 51 ].
This section analyses the diffusion of AI in healthcare around the world. It highlights countries to show the geographies of this research. It includes all published articles, the total number of citations, and the collaboration network.
The following sub-sections start with an analysis of the total number of published articles. Figure 9 and Table 10 display the countries where AI in healthcare has been considered. The USA tops the list of countries with the maximum number of articles on the topic It is immediately evident that the theme has developed on different continents, highlighting a growing interest in AI in healthcare.
The figure shows that many areas, such as Russia, Eastern Europe and Africa except for Algeria, Egypt, and Morocco, have still not engaged in this scientific debate. This section discusses articles on AI in healthcare in terms of single or multiple publications in each country.
It also aims to observe collaboration and networking between countries. Table 11 and Fig. Italy, Spain and New Zealand have the most significant number of citations. Figure 11 depicts global collaborations. The blue colour on the map represents research cooperation among nations. Additionally, the pink border linking states indicates the extent of collaboration between authors. The primary cooperation between nations is between the USA and China, with two collaborative articles. Other collaborations among nations are limited to a few papers.
This section aims to strengthen the research scope by answering RQ3: What are the research applications of artificial intelligence for healthcare? Benefiting from the topical dendrogram, researchers will provide a development model based on four relevant variables [ 69 , 70 ]. AI has been a disruptive innovation in healthcare [ 4 ]. With its sophisticated algorithms and several applications, AI has assisted doctors and medical professionals in the domains of health information systems, geocoding health data, epidemic and syndromic surveillance, predictive modelling and decision support, and medical imaging [ 2 , 9 , 10 , 64 ].
Furthermore, the researchers considered the bibliometric analysis to identify four macro-variables dominant in the field and used them as authors' keywords. Therefore, the following sub-sections aim to explain the debate on applications in healthcare for AI techniques.
These elements are shown in Fig. One of the notable aspects of AI techniques is potential support for comprehensive health services management. These applications can support doctors, nurses and administrators in their work. For instance, an AI system can provide health professionals with constant, possibly real-time medical information updates from various sources, including journals, textbooks, and clinical practices [ 2 , 10 ]. These applications' strength is becoming even more critical in the COVID period, during which information exchange is continually needed to properly manage the pandemic worldwide [ 71 ].
Other applications involve coordinating information tools for patients and enabling appropriate inferences for health risk alerts and health outcome prediction [ 72 ]. AI applications allow, for example, hospitals and all health services to work more efficiently for the following reasons:. Patients can stay informed and engaged in their care by communicating with their medical teams during hospital stays.
Additionally, AI can contribute to optimising logistics processes, for instance, realising drugs and equipment in a just-in-time supply system based totally on predictive algorithms [ 73 , 74 ]. Interesting applications can also support the training of personnel working in health services. This evidence could be helpful in bridging the gap between urban and rural health services [ 75 ].
Finally, health services management could benefit from AI to leverage the multiplicity of data in electronic health records by predicting data heterogeneity across hospitals and outpatient clinics, checking for outliers, performing clinical tests on the data, unifying patient representation, improving future models that can predict diagnostic tests and analyses, and creating transparency with benchmark data for analysing services delivered [ 51 , 76 ].
Another relevant topic is AI applications for disease prediction and diagnosis treatment, outcome prediction and prognosis evaluation [ 72 , 77 ]. Because AI can identify meaningful relationships in raw data, it can support diagnostic, treatment and prediction outcomes in many medical situations [ 64 ].
It allows medical professionals to embrace the proactive management of disease onset. Additionally, predictions are possible for identifying risk factors and drivers for each patient to help target healthcare interventions for better outcomes [ 3 ]. AI techniques can also help design and develop new drugs, monitor patients and personalise patient treatment plans [ 78 ].
Doctors benefit from having more time and concise data to make better patient decisions. Automatic learning through AI could disrupt medicine, allowing prediction models to be created for drugs and exams that monitor patients over their whole lives [ 79 ]. One of the keyword analysis main topics is that AI applications could support doctors and medical researchers in the clinical decision-making process. According to Jiang et al. According to Bennett and Hauser [ 80 ], algorithms can benefit clinical decisions by accelerating the process and the amount of care provided, positively impacting the cost of health services.
Therefore, AI technologies can support medical professionals in their activities and simplify their jobs [ 4 ]. Finally, as Redondo and Sandoval [ 81 ] find, algorithmic platforms can provide virtual assistance to help doctors understand the semantics of language and learning to solve business process queries as a human being would.
Another challenging topic related to AI applications is patient data and diagnostics. AI techniques can help medical researchers deal with the vast amount of data from patients i. AI systems can manage data generated from clinical activities, such as screening, diagnosis, and treatment assignment. In this way, health personnel can learn similar subjects and associations between subject features and outcomes of interest [ 64 ].
These technologies can analyse raw data and provide helpful insights that can be used in patient treatments. They can help doctors in the diagnostic process; for example, to realise a high-speed body scan, it will be simpler to have an overall patient condition image.
In terms of data, interesting research perspectives are emerging. For instance, we observed the emergence of a stream of research on patient data management and protection related to AI applications [ 82 ]. For diagnostics, AI techniques can make a difference in rehabilitation therapy and surgery.
Numerous robots have been designed to support and manage such tasks. For surgery, AI has a vast opportunity to transform surgical robotics through devices that can perform semi-automated surgical tasks with increasing efficiency. The final aim of this technology is to automate procedures to negate human error while maintaining a high level of accuracy and precision [ 84 ].
Finally, the period has led to increased remote patient diagnostics through telemedicine that enables remote observation of patients and provides physicians and nurses with support tools [ 66 , 85 , 86 ]. This study aims to provide a bibliometric analysis of publications on AI in healthcare, focusing on accounting, business and management, decision sciences and health profession studies. Using the SLR method of Massaro et al.
Additionally, we investigate the trend of scientific publications on the subject, unexplored information, future directions, and implications using the science mapping workflow. Our analysis provides interesting insights. These journals deal mainly with healthcare, medical information systems, and applications such as cloud computing, machine learning, and AI. Additionally, in terms of h-index, Bushko R. Burke et al.
Finally, in terms of keywords, co-occurrence reveals some interesting insights. For instance, researchers have found that AI has a role in diagnostic accuracy and helps in the analysis of health data by comparing thousands of medical records, experiencing automatic learning with clinical alerts, efficient management of health services and places of care, and the possibility of reconstructing patient history using these data.
Second, this paper finds five cluster analyses in healthcare applications: health services management, predictive medicine, patient data, diagnostics, and finally, clinical decision-making. These technologies can also contribute to optimising logistics processes in health services and allowing a better allocation of resources.
Third, the authors analysing the research findings and the issues under discussion strongly support AI's role in decision support. These applications, however, are demonstrated by creating a direct link to data quality management and the technology awareness of health personnel [ 87 ].
Several authors have analysed AI in the healthcare research stream, but in this case, the authors focus on other literature that includes business and decision-making processes. In this regard, the analysis of the search flow reveals a double view of the literature.
On the one hand, some contributions belong to the positivist literature and embrace future applications and implications of technology for health service management, data analysis and diagnostics [ 6 , 80 , 88 ]. On the other hand, some investigations also aim to understand the darker sides of technology and its impact.
For example, as Carter [ 89 ] states, the impact of AI is multi-sectoral; its development, however, calls for action to protect personal data. Similarly, Davenport and Kalakota [ 77 ] focus on the ethical implications of using AI in healthcare. According to the authors, intelligent machines raise issues of accountability, transparency, and permission, especially in automated communication with patients.
Our analysis does not indicate a marked strand of the literature; therefore, we argue that the discussion of elements such as the transparency of technology for patients is essential for the development of AI applications. A large part of our results shows that, at the application level, AI can be used to improve medical support for patients Fig.
However, we believe that, as indicated by Kalis et al. The potential of algorithms includes data analysis. There is an immense quantity of data accessible now, which carries the possibility of providing information about a wide variety of medical and healthcare activities [ 91 ]. With the advent of modern computational methods, computer learning and AI techniques, there are numerous possibilities [ 79 , 83 , 84 ]. For example, AI makes it easier to turn data into concrete and actionable observations to improve decision-making, deliver high-quality patient treatment, adapt to real-time emergencies, and save more lives on the clinical front.
In addition, AI makes it easier to leverage capital to develop systems and facilities and reduce expenses at the organisational level [ 78 ]. Studying contributions to the topic, we noticed that data accuracy was included in the debate, indicating that a high standard of data will benefit decision-making practitioners [ 38 , 77 ].
AI techniques are an essential instrument for studying data and the extraction of medical insight, and they may assist medical researchers in their practices. Using computational tools, healthcare stakeholders may leverage the power of data not only to evaluate past data descriptive analytics but also to forecast potential outcomes predictive analytics and to define the best actions for the present scenario prescriptive analytics [ 78 ].
The current abundance of evidence makes it easier to provide a broad view of patient health; doctors should have access to the correct details at the right time and location to provide the proper treatment [ 92 ]. Further reflection concerns the skills of doctors.
Studies have shown that healthcare personnel are progressively being exposed to technology for different purposes, such as collecting patient records or diagnosis [ 71 ]. This is demonstrated by the keywords Fig. Among the main issues arising from the literature is the possible de-skilling of healthcare staff due to reduced autonomy in decision-making concerning patients [ 94 ].
Therefore, the challenges and discussion we uncovered in Fig. In terms of theoretical contribution, this paper extends the previous results of Connelly et al. In terms of practical implications, this paper aims to create a fruitful discussion with healthcare professionals and administrative staff on how AI can be at their service to increase work quality.
Furthermore, this investigation offers a broad comprehension of bibliometric variables of AI techniques in healthcare. It can contribute to advancing scientific research in this field. Like any other, our study has some limitations that could be addressed by more in-depth future studies. For example, using only one research database, such as Scopus, could be limiting.
Moreover, although we analysed peer-reviewed scientific papers, because the new research topic is new, the analysis of conference papers could return interesting results for future researchers. Additionally, as this is a young research area, the analysis will be subject to recurrent obsolescence as multiple new research investigations are published. Finally, although bibliometric analysis has limited the subjectivity of the analysis [ 15 ], the verification of recurring themes could lead to different results by indicating areas of significant interest not listed here.
Concerning future research perspectives, researchers believe that an analysis of the overall amount that a healthcare organisation should pay for AI technologies could be helpful. If these technologies are essential for health services management and patient treatment, governments should invest and contribute to healthcare organisations' modernisation. New investment funds could be made available in the healthcare world, as in the European case with the Next Generation EU programme or national investment programmes [ 95 ].
Additionally, this should happen especially in the poorest countries around the world, where there is a lack of infrastructure and services related to health and medicine [ 96 ]. On the other hand, it might be interesting to evaluate additional profits generated by healthcare organisations with AI technologies compared to those that do not use such technologies.
Further analysis could also identify why some parts of the world have not conducted studies in this area. It would be helpful to carry out a comparative analysis between countries active in this research field and countries that are not currently involved. It would make it possible to identify variables affecting AI technologies' presence or absence in healthcare organisations.
The results of collaboration between countries also present future researchers with the challenge of greater exchanges between researchers and professionals. Therefore, further research could investigate the difference in vision between professionals and academics. In the accounting, business, and management research area, there is currently a lack of quantitative analysis of the costs and profits generated by healthcare organisations that use AI technologies.
Therefore, research in this direction could further increase our understanding of the topic and the number of healthcare organisations that can access technologies based on AI. Finally, as suggested in the discussion section, more interdisciplinary studies are needed to strengthen AI links with data quality management and AI and ethics considerations in healthcare. In pursuing the philosophy of Massaro et al. We performed this study with a bibliometric analysis aimed at discovering authors, countries of publication and collaboration, and keywords and themes.
We found a fast-growing, multi-disciplinary stream of research that is attracting an increasing number of authors. The research, therefore, adopts a quantitative approach to the analysis of bibliometric variables and a qualitative approach to the study of recurring keywords, which has allowed us to demonstrate strands of literature that are not purely positive.
There are currently some limitations that will affect future research potential, especially in ethics, data governance and the competencies of the health workforce. Artificial intelligence to improve back pain outcomes and lessons learnt from clinical classification approaches: three systematic reviews.
NPJ Digit Med. Article Google Scholar. Global evolution of research in artificial intelligence in health and medicine: a bibliometric study. J Clin Med. Hamid S. The opportunities and risks of artificial intelligence in medicine and healthcare [Internet].
Artificial intelligence, machine learning and health systems. J Glob Health. Concepts of artificial intelligence for computer-assisted drug discovery chemical reviews. Chem Rev. Using artificial intelligence to reduce diagnostic workload without compromising detection of urinary tract infections. Digital health is a cultural transformation of traditional healthcare.
Classification of cervical neoplasms on colposcopic photography using deep learning. Sci Rep. Finding undiagnosed patients with hepatitis C infection: an application of artificial intelligence to patient claims data. Clinical decision support in the era of artificial intelligence. On the shoulders of giants: undertaking a structured literature review in accounting. Account Auditing Account J. Junquera B, Mitre M.
Value of bibliometric analysis for research policy: a case study of Spanish research into innovation and technology management. How to design a winning business model. Harvard Business Review [Internet]. Aria M, Cuccurullo C. J Informetr. Bibliometric methods in management and organization. Organ Res Methods. Secinaro S, Calandra D. Halal food: structured literature review and research agenda.
Br Food J. Twenty-five years — of the international business review: a bibliometric overview. Int Bus Rev. Application of artificial intelligence to wastewater treatment: a bibliometric analysis and systematic review of technology, economy, management, and wastewater reuse. Process Saf Environ Prot. Rehabilitation using virtual reality technology: a bibliometric analysis, — A bibliometric analysis of text mining in medical research.
Soft Comput. Data mining and machine learning techniques applied to public health problems: a bibliometric analysis from to Comput Ind Eng. A bibliometric analysis and visualization of medical big data research. Use of machine learning in geriatric clinical care for chronic diseases: a systematic literature review.
The most influential manuscripts in robotic surgery: a bibliometric analysis. J Robot Surg. Artificial intelligence in health care: bibliometric analysis. J Med Internet Res. Choudhury A, Asan O. Role of artificial intelligence in patient safety outcomes: systematic literature review.
Entrepreneurial universities: a bibliometric analysis within the business and management domains. Technol Forecast Soc Change. Social media for entrepreneurship: myth or reality? A structured literature review and a future research agenda. Int J Entrep Behav Res. From output to outcome measures in the public sector: a structured literature review.
Int J Organ Anal. Google Scholar. Intellectual capital and business model: a systematic literature review to explore their linkages. J Intellect Cap. IC and public sector: a structured literature review. Knowledge translation in the healthcare sector. A structured literature review. Electron J Knowl Manag. La performance nel settore pubblico tra misure di out-put e di outcome. Una revisione strutturata della letteratura ejvcbp. Dumay J, Cai L.
A review and critique of content analysis as a methodology for inquiring into IC disclosure. Current status and applications of Artificial Intelligence AI in medical field: an overview. Curr Med Res Pract. Paul J, Criado AR. The art of writing literature review: what do we know and what do we need to know? The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. Data quality methods and applications in health care system: a systematic literature review.
Int J Bus Manag. The use of micafungin in neonates and children: a systematic review. HIV screening in pregnant women: a systematic review of cost-effectiveness studies. Int J Health Plann Manag. Levy Y, Ellis TJ. A systems approach to conduct an effective literature review in support of information systems research.
Chen G, Xiao L. Selecting publication keywords for domain analysis in bibliometrics: a comparison of three methods. J Informet. Community curation in open dataset repositories: insights from zenodo. Procedia Comput Sci. Artificial Intelligence for healthcare with a business, management and accounting, decision sciences, and health professions focus [Internet].
Zenodo; [cited Mar 7]. Elango B, Rajendran D. Authorship trends and collaboration pattern in the marine sciences literature: a scientometric Study. Int J Inf Dissem Technol. Jacoby WG. Electoral inquiry section Loess: a nonparametric, graphical tool for depicting relationships between variables q. In Andrews JE. An author co-citation analysis of medical informatics.
J Med Libr Assoc. Author cocitation: a literature measure of intellectual structure. J Am Soc Inf Sci. Santosh KC. J Med Syst. A systematic review of wearable patient monitoring systems—current challenges and opportunities for clinical adoption. Kumar S, Kumar S. Collaboration in research productivity in oil seed research institutes of India. In: Proceedings of fourth international conference on webometrics, informetrics and scientometrics. Gatto A, Drago C. A taxonomy of energy resilience.
Energy Policy. Levitt JM, Thelwall M. Alphabetization and the skewing of first authorship towards last names early in the alphabet. Saad G. Exploring the h-index at the author and journal levels using bibliometric data of productive consumer scholars and business-related journals respectively.
Egghe L. Theory and practise of the g-index. Schreiber M. A modification of the h-index: the hm-index accounts for multi-authored manuscripts. Engqvist L, Frommen JG. The h-index and self-citations. Trends Ecol Evol. London School of Economics. Impact of social sciences. Lotka A. The frequency distribution of scientific productivity. J Wash Acad Sci. Khan G, Wood J. Information technology management domain: emerging themes and keyword analysis. Oxford University Press. Oxford English Dictionary [Internet].
Artificial intelligence in healthcare: past, present and future. Stroke Vasc Neurol. Calandra D, Favareto M. Bokolo Anthony Jnr. The state of the art of nurse rostering. J Sched. Simulation optimization for an emergency department healthcare unit in Kuwait. Eur J Oper Res. Clustering with dendrograms on interpretation variables. Anal Chim Acta. Wartena C, Brussee R. Topic detection by clustering keywords. In: 19th international workshop on database and expert systems applications.
IEEE Access. Exploring the impact of artificial intelligence: prediction versus judgment. Inf Econ Policy. Chakradhar S. Citation Type. Has PDF. Publication Type. More Filters. Highly Influenced. View 10 excerpts, cites background and methods. Research Feed. View 1 excerpt, cites background. Mining open source text documents for intelligence gathering. Advancing artificial intelligence research and dissemination through conference series: Benchmark, scientific impact and the MICAI experience.
A Siberian reality check on open source information. View 1 excerpt, references methods. Open Source Intelligence. A systematic literature review and meta-analysis on artificial intelligence in penetration testing and vulnerability assessment. View 2 excerpts, references background. Artificial Intelligence AI : Multidisciplinary perspectives on emerging challenges, opportunities, and agenda for research, practice and policy.
View 1 excerpt, references background. Open-source intelligence for risk assessment.