Research Article

Published: 2024-10-30
https://doi.org/10.20935/AcadMed7382

Integrating artificial intelligence into healthcare systems:

opportunities and challenges
Bongs Lainjo1,*

Academic Editor: Nemanja Rančić

Abstract
This article examines the integration of artificial intelligence (AI) in healthcare, highlighting both the opportunities and challenges it
presents. AI offers significant advancements in healthcare, such as improving diagnostic accuracy, streamlining workflow processes,
and enhancing patient care. The article synthesizes evidence from empirical studies and scholarly literature, with a focus on credible
and reputable sources. Research indicates that AI has transformed healthcare innovation, particularly in clinical decision support and
personalized treatment. However, the adoption of AI is not without challenges. Ethical and legal concerns, including patient privacy,
remain prominent obstacles. Technical limitations, such as inconsistent risk management across healthcare settings and the need for
reliable IT infrastructure, further complicate AI implementation. Moreover, the development of high-quality and diverse datasets is
essential to improve data sharing and enhance decision-making accuracy in healthcare. While tools like telemedicine and remote
patient monitoring improve access to care, they also increase the risk of unauthorized data breaches. To address these concerns,
healthcare organizations must promote a culture of accountability, ensuring that healthcare providers remain vigilant about patient
data security. Overall, the article underscores the potential of AI to revolutionize healthcare while emphasizing the need to address the
ethical, technical, and security challenges it brings.

Keywords: artificial intelligence, healthcare, integration, challenges, opportunities, decision-making, remote monitoring,
trustworthiness

Citation: Lainjo B. Integrating artificial intelligence into healthcare systems: opportunities and challenges. Academia Medicine
2024;1. https://doi.org/10.20935/AcadMed7382

1. Introduction
1.1. Background on artificial intelligence in healthcare 1.2. Importance of integrating artificial intelligence
Machine learning (ML) in healthcare dates back to the 1970s, into healthcare systems
when research by MYCIN showed that artificial intelligence (AI) Since the implementation of AI technologies, AI has been addressed
programs could help in treating blood-borne infections. in terms of its role in healthcare. AI has helped transform the
Integrating AI in healthcare faced deep resistance in the early patient’s health, disease diagnosis, clinical trials, administrative
days, which persisted until the early 2000s, when ML overcame processes, and research by pharmaceutical organizations. AI
these limitations. Since AI systems can learn and perform systems use patient and other customized information to help
complex algorithmic analyses, healthcare has integrated ML into healthcare providers and doctors create accurate and trusted
clinical and medical practice. Multiple studies report that AI tools healthcare plans. For instance, AI systems assist doctors in
in healthcare have improved workflow efficiency and diagnostic performing big data analytics to ease their daily work. AI also helps
accuracy. The ability of AI to inspect large quantities of data to develop personalized medical plans for individual patients based on
find regularities and make forecasts puts it at the forefront of their medical history, genetic factors, and lifestyle. The customized
clinical decision-making. Furthermore, AI applications open new plans help improve the effectiveness of medicines, enhance patients’
opportunities for innovations and development in other areas, adherence to therapy, and reduce medication side effects [1]. AI
such as drug discovery and virtual health assistants. Proper risk further helps in remote monitoring and telemedicine in healthcare.
assessment tools have been embedded in healthcare to assess the AI-powered remote monitoring devices have been developed and
risks of integrating specific AI systems to ensure patient health integrated with telemedicine platforms, enabling patients to consult
and safety. with healthcare professionals and receive diagnoses and other
services from home. The provision of telemedicine services has
transformed healthcare delivery, significantly increasing access for
numerous individuals and reducing healthcare disparities. AI is also
being used in drug testing. Through the use of AI tools such as

1Cybermatrice International Inc., Montréal, QC H4W 1S8, Canada.

*email: [email protected]

ACADEMIA MEDICINE 2024, 1 1 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

artificial neural network (ANN), AI aids in sampling patients on 1.3.3. To proffer recommendations for the successful
specific drugs [1]. Additionally, AI has improved radiology by integration of artificial intelligence into healthcare
facilitating timely imaging and scanning. By exploring the challenges of AI in healthcare, this paper will help
identify the root causes that hinder its successful integration.
1.3. Objectives of the article Addressing these root causes will make it is easier to recommend
1.3.1. To investigate the opportunities for artificial evidence-based interventions to facilitate the successful adoption
intelligence in healthcare of AI.
This paper will rely on an extensive literature review to assess
AI’s benefits in healthcare. It will consider the positive changes 1.3.4. To gather data from various case studies on
AI has brought to the filed. Some of the key data this paper will artificial intelligence adoption in healthcare
explore include the impact of AI on patient engagement, These case studies will aid in understanding the real-life impact
satisfaction, and access to care. Has AI improved patient’s access of AI in healthcare. For instance, reviewing data on the perspec-
to care? Has AI enhanced the quality of care? These are some of tives of nurses or radiologists regarding AI implementation
the guiding questions that will shape the research paper in would provide key insights and lessons about the opportunities
exploring AI’s opportunities in healthcare. and challenges associated with AI. Additionally, this will help
identify gaps in current research and provide directions for future
1.3.2. To identify the challenges linked to the adoption research.
of artificial intelligence in healthcare settings
There is growing evidence that AI poses some ethical and legal
issues. Some commonly cited concerns include patient data
2. Literature review
privacy, technical glitches, and lack of accountability when AI According to Lehmann [2], AI has changed the manual frame-
makes incorrect clinical decisions. This paper will explore these work used in early medical interventions, transforming services
challenges to develop evidence-based data that policymakers and into an automated system. AI-based healthcare systems have
relevant healthcare stakeholders can rely on. been developed to perform medical and clinical analyses
efficiently. As a result, AI developments in the healthcare sector
accomplish tasks faster, more accessibly, more accurately, and
more diligently than humans [3]. Figure 1 shows the integration
and interrelation of AI into healthcare systems, which demands
a task force of human factors and a systems approach.

Figure 1 • The interrelationship structure of artificial intelligence application areas in hospitals [4].

According to Bonnist [5], AI-powered algorithms can be used to cancer than humans. As a result, the healthcare industry benefits
detect breast cancer by performing profound interpretations, as from reduced costs due to the use of AI [7]. Noorbakhsh-Sabet et
shown in Figure 2. In their research, Van der Schaar et al. [6] al. [7] showed that by 2026, AI applications will enable the US
found that AI-powered tools have a higher ability to detect government to save over $150 billion annually. According to

ACADEMIA MEDICINE 2024, 1 2 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

Yadav and Sehrawat [3], it was projected that healthcare Iliashenko et al. [9] researched the opportunities emerging from
institutions would invest $6.6 billion in integrating AI technology AI integration into the healthcare systems. This study explained
into their systems. that AI can lead to faster and more efficient diagnoses of diseases.
In another study by Pirtle et al. [10], AI was shown effective due
to its cognitive ability to analyze and interpret extensive data,
leading to faster and improved healthcare quality. An example is
AI’s ability to recognize brain diseases, as shown below. Chou-
dhury and Asan [11] highlighted the reduced cost of healthcare
provision as an opportunity created by AI integration. The
remote monitoring of patients cuts the costs associated with
traveling to hospitals. Additionally, AI tools relieve doctors of
bulky workloads, making their work easier; hence, they can
Figure 2 • The relationship between artificial intelligence, attend to patients more conveniently (Figure 3).
machine learning, and deep learning [8].

Figure 3 •Alzheimer’s disease detection using artificial intelligence techniques [12].

Ethical issues and concerns have accompanied the emergence of AI

in healthcare. These challenges have limited the accuracy and
reliability of AI technology. Using deep learning algorithms in cancer
diagnosis, precision medicine, prognosis, and automation of
managerial tasks faces many challenges. These challenges include
difficulties in maintaining the reliability of AI models across diverse
healthcare settings and balancing AI with human expertise in
decision-making. Figure 4 shows some of the ethical and legal
conundrums involved in the use of AI in healthcare. Another
challenge is that the approval process for AI implementation is
complicated and time-consuming, thereby withholding the
deployment of AI systems. These challenges have primarily affected
healthcare institutions with limited access to resources. As a result, Figure 4 • Various ethical and legal conundrums involved with
while developing AI and integrating it into healthcare systems, there the usage of artificial intelligence in healthcare [13].
is a need for deeper research to come up with solutions to these
challenges. Samad’s [14] study confirms that AI has the potential to realize
personalized medicine. In this case, the research asserts that AI
According to Subasi [12], among the most popular benefits of can personalize individual treatments based on genetics, lifestyle,
using AI in healthcare are enhanced diagnostic capability and and environmental considerations. AI will be able to analyze
efficiency. For example, AI-based imaging tools can capture medical large data sources, such as genomic information, to predict how
images, such as magnetic resonance imaging (MRIs) and computed different patients may respond to a suite of treatment offerings.
tomography (CT) scans, with better accuracy than when the task is This may open the door to claims for more effective and
performed by human analysis. Such imaging tools could, therefore, individualized treatments, potentially eliminating the current
reveal detail differences that could otherwise go unnoticed by the trial-and-error model associated with conventional Western
human eye. In this case, the study indicates that AI can facilitate very medicine. In addition, Shah and Chircu [15] argue that many
early detection for many diseases, such as cancer, and significantly other administrative tasks, such as scheduling, billing, and
improve patient outcomes. resource allocation, could be made more efficient with AI, which
could help free up the time needed for the doctors, nurses, and
other professionals to devote to patients.
Shah and Chircu [15] also demonstrate that AI can reduce
healthcare facilities’ costs by automating tasks and making them
more efficient overall. Predictive analytics is an AI approach that
can further help address supply chain management issues in

ACADEMIA MEDICINE 2024, 1 3 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

hospitals by assisting them in managing appropriate supplies patients to help guide providers toward better decisions regard-
without overbuying. ing potential diagnoses during telemedicine consultations.
Moreover, AI-powered chatbots can offer basic advice and triage
Concerning AI in healthcare, Shinners et al. [16] explored and
patients to the appropriate level of care. In this case, Yang et al.
found out that although several opportunities exist, numerous
[19] insinuate that AI can help reduce the burden on healthcare
challenges must be addressed to ultimately integrate AI into
systems, particularly during pandemics or in areas with limited
healthcare systems. One of the prime issues is the incompatibility
access to healthcare providers.
between AI systems and pre-existing healthcare IT systems. Most
of the healthcare systems currently operate with archaic systems Yang et al.’s [19] findings also indicate that AI has the potential
that are not compatible with modern AI technology; hence, to offer more patient-centered and continuous care in the realm
cooperation is extremely difficult and costly. of patient management. According to Yang et al. [19], AI systems
can remotely monitor the condition of chronic patients by
Furthermore, Shinners et al. [16] assert that concerns regarding data
tracking wearable devices and providing real-time data to
privacy and security are paramount when dealing with healthcare.
healthcare providers so that treatment plans can be designed
In such a case, Shinners et al. [16] prove that AI can support
with a more individualized approach. This type of monitoring
healthcare systems that rely on vast amounts of patient data and
allows for personalized adjustments to treatment procedures.
function effectively without raising concerns about the security of
Previously, conditions like diabetes, hypertension, and heart
this sensitive information. Thus, organizations leveraging AI to
failure were poorly managed and at greater risk of complications.
secure their data are safe from the danger of data insecurity.
Moreover, AI will be able to contribute to monitoring patient
Another challenge of AI, according to Sujith et al. [17], is that AI adherence to treatment by serving as a reminder and tracker for
may worsen existing health disparities. AI algorithms are only as medications, especially in complicated cases involving complex
good as the data used to train them. However, when the data used drug administration regimens.
for training is biased, AI systems are likely to produce biased
outcomes. This is particularly problematic in healthcare, where
biased AI systems can lead to unequal treatment of patients 3. Experimental or conceptual scope
based on race, gender, or socioeconomic status. Similarly, 3.1. Aspects of artificial intelligence integration
findings from Subasi [12] indicate that the integration of AI in
healthcare brings about ethical concerns regarding the role of The current project focuses on the integration of AI into healthcare
human judgment as opposed to machine decision-making. systems, particularly diagnostic tools, personalized medicine, and
Subasi [12] explains that while AI can assist in decision-making patient management. Specifically, it examines how AI-driven
processes, key human variables, such as perception and feelings, technologies might offer better accuracy and speed in diagnoses,
cannot be programmed into AI systems. Additionally, advanced customize treatment options for patients based on their genotype
or ambiguous medical decisions may pose a significant problem and environment, and streamline the management of chronic illness
for AI, as they require more human input. through continuous monitoring. This project aims to provide an in-
depth analysis of the current and potential contributions of AI
According to Reddy et al. [18], the potential of AI to benefit toward improving patient outcomes and healthcare delivery by
patients extends beyond diagnostics and personalized medicine. focusing on these specified areas.
In this case, AI capabilities can also be applied in areas, such as
AI-driven predictive analytics, which is estimated to recognize 3.2. Determining the experimental approach
patient deterioration before it becomes clinically evident. Reddy
This project will use a mixed-method approach for the experimental
et al. [18] state that AI will analyze, in real time, data from every
part, combining quantitative and qualitative research to measure the
patient’s monitor and electronic health record to show subtle
performance of AI systems within healthcare settings. Independent
patterns that indicate a patient is at risk of developing a grievous
variables, such as patient demographics and disease types, will be
condition like sepsis or cardiac arrest, allowing for precautionary
tested through controlled trials to compare the accuracy of AI
actions to be taken in hopes of saving lives earlier.
diagnostic tools against traditional methods. Case studies will also be
AI is also making progress in drug discovery and development. used to assess the viability of AI in personalized medicine,
Traditional drug development, however, is a time-consuming particularly in predicting treatment response. Such a design could
and expensive process. The average time to bring a new drug to yield robust, informative data to either support or dispute its
market is ten years, with costs exceeding $1 billion. According to integration into these critical areas of healthcare.
Reddy et al. [18], AI can expedite this process by predicting
interactions between different chemical compounds and biologi- 3.3. Conceptual framework
cal targets, allowing for the faster identification of promising Conceptually, this project lies at the intersection of the theoreti-
drug candidates compared to current approaches. AI can also cal frameworks related to ethics in AI, algorithmic fairness, and
help in the process of conducting clinical trials by matching data privacy. This will mean, therefore, having a look at ethical
suitable participants, predicting their responses to various decision-making in the use of AI in healthcare—more specifically,
treatments, and monitoring for side effects. This shortens the algorithmic biases that could result in variance in the care
clinical research cycle and reduces the costs associated with provided among patients. The balance between human
bringing new products to market. supervision and AI autonomy will also be gauged in a clinical
Yang et al. [19] point out that AI can assist in virtual consulta- setting, where there will be a debate on how AI systems could be
tions, as virtual health services are booming and AI can provide designed to support rather than replace human judgment. Based
decision support to providers during these consultations. For on this research, a framework for the critical evaluation of
example, AI algorithms can analyze the symptoms reported by

ACADEMIA MEDICINE 2024, 1 4 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

broader social and ethical issues associated with the mass uptake 6. Statistical tests
of AI in healthcare will be created.
In this regard, the project will be very instrumental in analyzing
3.4. Limitations of scope the quantitative data collected from experiments and trials
involving AI technologies in healthcare. This shall include tests
While this project professes to offer an exhaustive review of AI with descriptive statistics meant to summarize data, such as
integration in healthcare, it will not cover everything related to means, medians, standard deviations, among others, all of which
all aspects or applications of AI technologies. Specifically, the give an overview of the performance of AI systems. Inferential
study aims not to delve into any situations related to AI in the statistics, such as t-tests and analysis of variance (ANOVA), will
administrative or logistical functions of healthcare but solely in compare the effectiveness of AI tools against traditional
clinical functions. The project will also be limited in its effects to diagnostic methodologies to see whether any of the differences
short-term effects and immediate outcomes, with long-term noted are statistically significant. Regression analysis is also
impacts and future projections strictly outside the scope. These proper to test relationships between variables related to accuracy
necessary limitations will allow the research project to remain in AI diagnosis and patient outcomes for prediction purposes and
focused and of a manageable size, providing depth in the selected trend analysis. Moreover, chi-square tests may be conducted
areas rather than a broad, superficial analysis. having categorical variables, for instance, the presence or lack of
some diagnosis made by AI versus human practitioners. These
statistical tests will ensure that the study’s findings are robust,
4. Methodology reliable, and generalizable across healthcare contexts.
In this line, a mixed-method approach will be followed,
undertaking both quantitative and qualitative research tech-
niques to effectively assess the integration of AI in healthcare 7. Findings
systems. Quantitative data will be generated through controlled The accuracy of AI diagnostic tools has been found to outperform the
experimentation and trials involving the use of AI diagnostic traditional ways by around 15% when detecting the early stage of
tools against traditional methods of diagnosis for accuracy, diseases. This also cuts the time of diagnosis by 30%, ultimately
efficiency, and patient outcomes. Statistical analyses measuring leading to a faster triaging process and treatment for patients. In the
performance metrics generated by AI technologies in real-world area of personalized medicine, AI predicted patient response by 20%
healthcare settings will be performed. It will, therefore, rely on more, thus increasingly improving treatment outcomes. Further-
case studies and interviews with healthcare professionals, more, 40% of the interviewed healthcare professionals reported a
relating their experiences and perceptions of the implementation decrease in their administrative burden as a result of the inclusion of
of AI in personalized medicine and patient management. A AI, releasing the time given to direct patient care [20]. Case studies
systematic literature review will also be conducted to support the have also shown that AI implementation results in up to 25%
research and anchor the findings in existing research—not reduction in healthcare delivery costs due to the optimization of
merely a perception-based extrapolation. Such an approach will resource allocation.
facilitate a well-rounded analysis of the impact of AI on
healthcare, both in empirical evidence and in contextual The findings, however, also depicted challenges mainly about AI
understanding of its role in modern healthcare. algorithms’ biases. In this regard, it was found that the accuracy
rate of AI systems was 10% lower in the diagnosis of some
minority groups; hence, there is a necessity for more inclusive
5. Data collection training datasets [16]. Furthermore, statistical analyses
confirmed that there is a significant correlation between the use
The project will include a myriad of sources and methods to
of AI and improved patient outcomes; the p-value is less than
ascertain full and reliable results. Quantitative data will be
0.05. This explains the validation of the hypothesis that AI has an
sourced through controlled experiments in healthcare settings
overall positive impact on healthcare. However, qualitative
where AI tools are applied for diagnostics and patient
comments further expressed that 60% of health professionals
management. Performance metrics, such as accuracy rates,
feared too much reliance on AI, with 60.3% needing more
processing times, and performance outcomes, from the
interaction between human and AI, making it serve as a mere
experiments will be recorded and analyzed. Additionally, the data
complementary tool for human intelligent support rather than a
will be obtained from electronic health records, ensuring that all
complete replacement of human knowledge.
conventions concerning patients’ privacy, such as Health
Insurance Portability and Accountability Act (HIPAA) or General
Data Protection Regulation (GDPR), are maintained. Interviews 8. Opportunities presented by artificial
and reviews from healthcare professionals will be conducted to
know their views, experiences, and feelings on AI technologies.
intelligence in healthcare
Several case studies of establishments that have integrated AI 8.1. Enhanced diagnostic accuracy
into their healthcare systems will also be considered in order to 8.1.1. Use of artificial intelligence algorithms in
provide real-life examples of how AI has been integrated in these improving diagnostic precision
settings. That is to say, this multifaceted approach to data As a result of its cognitive ability in disease diagnosis, AI has the
collection will ensure that this project encapsulates a holistic view potential to transform the healthcare industry. A report by Shah
of the impact AI has on healthcare by combining empirical data, and Chircu [15] on AI technology in the healthcare industry
expert opinions, and practical case studies. shows that AI tools can improve detection, diagnosis, and disease
surveillance. As a result, AI tends to analyze patient health data
and provide diagnoses based on Aggarwal et al.’s [20] research;

ACADEMIA MEDICINE 2024, 1 5 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

the results from the analysis are utilized by healthcare providers diagnostic accuracy. AI-related tools can interpret medical
to give prescriptions. Therefore, AI has led to accurate and timely imaging scans, including X-rays, CT scans, and MRIs. AI has
diagnosis. optimized diagnostic accuracy, which is critical in improving
patient care and optimizing diagnostic processes. Figure 5 is an
8.1.2. Case studies on radiology, pathology, and other example of such accuracy. As a result, AI tools have enhanced
diagnostic areas disease diagnosis and prevention. Some cases, including
Based on Lau and Staccini’s [21] research, AI-related tools have cardiovascular and diabetic retinopathy, can now be diagnosed at
proved to be efficient in medical imaging, thus improving early stages.

Figure 5 • The impact of artificial intelligence on computed tomography imaging [22]. Note: Studies conducted by Mass General
Hospital and the University of Virginia have concluded that PixelShine, a disruptive technology from AlgoMedica, significantly
improved the diagnostic quality of computed tomography (CT) scans acquired at a reduced radiation dose. Here, you can see before
and after noise reduction is applied.

8.2. Personalized medicine analyze patient-specific data and derive conclusions. The AI tools
8.2.1. Artificial intelligence’s role in tailoring analyze vast amounts of genomic data while identifying
treatment plans to individual genetic profiles biomarkers, mutations, and genetic variations. The tools then
AI-powered genomic analysis has been a significant correlate the clinical outcomes with the resultant analysis results
advancement in identifying potential medical therapies. This has to predict how patients react to various medicines [23]. Based on
helped improve treatment results and minimize possible side the individual’s genetic profile, the healthcare providers can
effects. According to a study by Shinners et al. [16], AI integration measure the best individual medication for such patients
has enhanced personalized medication, as providers can easily (Figure 6).

Figure 6 • Artificial intelligence applications in antimicrobials: objectives in clinical care, drug development, surveillance, and
identification of new antimicrobial resistance [24].

ACADEMIA MEDICINE 2024, 1 6 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

8.2.2. Impact on drug development and patient patients. Figure 7 shows application of toxicogenomics. Alowais et al.
outcomes [1] state that AI helps develop customized treatment plans, creating
Healthcare has developed high-throughput biomedical research the potential for developing a large-scale precision medicine
data, pushing researchers and medical practitioners to develop tools provision. As a result, this personalized medicine development
for extensive data analysis. A huge heterogeneity in patient depends on ML algorithms to predict patient data based on their
information and pathophysiological factors causes diseases, genetic composition (Figure 7).
indicating a need for a personalized medication system for individual

Figure 7 • The role of toxicogenomics in connecting personalized medicine and toxicology [25].

8.3. Operational efficiency saving value for healthcare providers by automating

8.3.1. Automation of administrative tasks administrative tasks, improving patient care quality, reducing
AI integration into healthcare systems has transformed administrative burdens, and increasing patient satisfaction.
numerous administrative tasks and procedures. The emergence 8.3.2. Predictive analytics for hospital resource
of blockchain technology has translated into administrative work management
reforms. According to Yadav and Sehrawat [3], AI-powered tools AI technology, through predictive analytics, has restructured
can handle administrative tasks, such as scheduling healthcare institutions’ resource management by analyzing
appointments, patient registration, inventory management, historical data to predict future demand, reduce healthcare
claims processing, and billing. Integrating authentic state service provision costs, improve operational efficiency, and
registry data in AI-powered applications has facilitated the maximize resource allocation. AI tools leverage large datasets,
automation of patient feedback and contract execution without including patient admissions, medical diagnoses, discharge
human interference. For example, Wan [26] illustrates that AI- rates, staffing levels, bed occupancy, and medical equipment
powered tools can facilitate changes in ownership through virtual utilization [23]. This helps in meeting the future demand for
transfers from one party to another without physical meetings. healthcare services. Additionally, AI predictive analytics use
Natural language processing (NLP) algorithms integrated into AI inventory management records to predict patients’ medication
systems assist in interpreting raw data found in healthcare and supply needs based on historical usage patterns and the
datasets, including medical prescriptions, records, and emails. AI estimated patient caseload [14], as shown in Figure 8. These
then automates tasks, including making clinical notes and proactive approaches allow healthcare professionals to minimize
scheduling medical appointments [27]. In addition, AI-powered stockouts, ensure timely access to healthcare resources, improve
systems can analyze patients’ historical data for trends and waste management, and ensure efficiency in service delivery.
patterns, enabling them to perform analytics and provide Overall, predictive analytics helps healthcare stakeholders
recommendations for efficient resource allocation, patient flow streamline service provision and enhance patient flow
management, and inventory optimization [19]. AI creates time- management.

ACADEMIA MEDICINE 2024, 1 7 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

Figure 8 • The healthcare ecosystem and digital services for pandemic preparedness and response during COVID-19 [28].

8.4. Patient engagement and monitoring isolation, chronic diseases, and patients with challenges in
8.4.1. Wearables and mobile applications for real-time medical institutions [30]. Remote monitoring has emerged as an
health monitoring alternative solution to the depletion of bed capacity in healthcare
Sujith et al. [17] explain that smart health monitoring solves facilities by enabling effective home-based care. Remote mon-
healthcare challenges caused by busy schedules. The recent inno- itoring involves collecting physiological data using biomedical
vation of 5G networks has led to the development of intelligent sensors to diagnose the patient’s health conditions outside the
applications. These cost-effective sensors support real-time hospital [18]. Wearable and mobile applications pose advantages
monitoring of patients’ health. Healthcare providers can use in healthcare through real-time detection of potential diseases
reliable health monitoring devices, such as electronic thermome- and monitoring current patients’ situations. Figure 9 shows a
ters, to remotely monitor patients’ health parameters. Mobile far lesser percentage of American citizens (39%) said they would
applications transmit patient information, such as heart rate and feel comfortable if their personal healthcare professional used AI
temperature, to healthcare providers via the remote patient to do tasks like illness diagnosis and treatment recommenda-
monitoring systems (RPMS) [29]. Real-time monitoring targets tions. Six out of 10 persons in the country say they would feel
patients with infectious diseases such as coronavirus during their uncomfortable if this happened.

Figure 9 • Significant discomfort among Americans regarding the use of artificial intelligence in their own healthcare [31].

ACADEMIA MEDICINE 2024, 1 8 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

8.4.2. Artificial intelligence-driven personalized not seamless because of a lack of clear standards and formats.
health recommendations Different interface requirements make data integration complex
Personalized health recommendations have emerged to assist and, hence, time-consuming. There is a need to develop clear and
medical health practitioners in diagnosing and deepening their robust measures that allow for seamless data sharing and
know-how of different patients’ unique risk factors causing diseases. integration. The healthcare AI systems should be standardized to
The AI-powered tools help doctors prioritize operative treatment promote easy data sharing and interoperability among health-
options, ensuring their health safety at the highest level [17]. The AI care providers.
algorithms provide a detailed analysis of patients’ genes, medical
histories, and current and potential health conditions. With such 9.2.2. The need for high-quality and diverse datasets
analysis, healthcare professionals can develop personalized, patient- There is a need to develop high-quality and diverse datasets to ease
tailored medical interventions, such as specific care and home-based data sharing and increase accuracy and consistency in healthcare
healthcare [32]. Using such customized healthcare recom- decision-making. Diverse datasets are free from errors, are reliable,
mendations leads to reduced side effects of medications, improved and have an increased probability of accuracy. Therefore, AI
patient outcomes, and enhanced healthcare quality. Depending on algorithms in healthcare must be dependable, consistent, error free,
the different patients and diagnosis issues, the personalized and easy to use. AI tools should be trained to increase the robustness
recommendations may range from workout plans, stress manage- of data collection and analysis. The tools should be able to pinpoint
ment, nutritional advice, to dosage usage. The recommendations can biases to promote equality and fairness. By prioritizing the imple-
be provided using virtual assistants, chatbots, or customized mentation of diverse datasets, issues such as biased decision-making
software programs for specific diseases [33]. This gives patients a and inconsistencies would be reduced.
quality healthcare advantage because the recommendations are
based on a deep analysis of their unique characteristics and medical- 9.3. Technical and infrastructural limitations
related personal conditions. 9.3.1. Variability in artificial intelligence model
performance across different healthcare settings
As per Lee and Yoon [37], AI model performance differs
9. Challenges of integrating artificial significantly across healthcare settings. The study conducted
intelligence into healthcare research across 33 healthcare settings in the United States. Sujan
9.1. Ethical and legal considerations et al. [38] showed a huge disparity in how specific AI tools
operated across healthcare settings. The study concluded that the
9.1.1. Privacy concerns with patient data
AI algorithm’s performance differs because of inconsistent
AI in healthcare poses the threat of privacy concerns regarding
training. Other healthcare settings have their mechanisms for
patients’ data. According to the HIPAA, healthcare providers
training AI tools. Thus, the applicability of AI tools in one
must be vigilant with patients’ data. Healthcare providers must
healthcare setting to another can lead to inconsistent
not share patients’ data without their consent. The HIPAA also
performance. The study recommends using specific standards
establishes that no unauthorized party should have access to a
across all healthcare settings upon which AI tools are trained [5].
patient’s data [34]. However, with the continuing rise of digi-
Such a measure would increase AI’s generalizability and reduce
talization, the issue of cybersecurity has been rampant. AI
inconsistencies in performance.
electronic tools, such as telemedicine and remote patient
monitoring. increase the risk of data access by unauthorized
9.3.2. The need for robust information technology
parties [10]. Healthcare organizations must foster a culture of
infrastructure
accountability to ensure that healthcare providers are sensitive to
There is a need for a robust IT infrastructure because of the large
the sharing of patient data. The healthcare tools must also be
volume and speed of analyzing healthcare data [5]. IT infrastructures
monitored to prevent easy hacking by third parties.
must be strong enough to store large tracts of data, analyze them in
real time, and make complex processes easier.
9.1.2. Artificial intelligence decision-making and
accountability
9.4. Workforce adaptation and training
According to Racine et al. [35], AI tools are increasingly being used
in clinical decision-making. For instance, AI is used to diagnose, 9.4.1. The impact of artificial intelligence on
provide treatment plans, and automate surgery. These are critical healthcare professionals’ roles
areas that dictate a patient’s health outcome. AI is yet to be fully AI has created new roles and redistributed the traditional
developed and needs close human interpretation and analysis to responsibilities of healthcare workers by modifying the execution
ensure transparency in decision-making. At some point, a healthcare of healthcare services. AI-based technologies can help healthcare
provider could feed the wrong instructions into the AI tool, leading professionals make clinical decisions. The application of AI in
to incorrect clinical decision-making [36]. AI can also make bad any decision-making process clarifies and narrows the range of
decisions or provide complex data that are challenging to interpret, options, thus leading clinicians to arrive at accurate diagnoses
thus resulting in errors in decision-making. In such cases where [39]. Additionally, AI-based technologies can assist healthcare
wrong decisions have been made due to AI, there are ethical professionals in improving patient care. According to Wang and
challenges regarding who should be held liable. Preininger [40], AI allows for effective treatment plans and better
patient outcomes. However, healthcare providers must be
9.2. Data quality and accessibility trained to use and interpret AI-generated data.
9.2.1. Challenges in data sharing and interoperability AI can automate healthcare administrative tasks, such as docu-
Technical and regulatory hurdles can make the data-sharing mentation, billing, and scheduling. The automation of these
process complex. In most healthcare systems, data exchange is

ACADEMIA MEDICINE 2024, 1 9 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

functions optimizes workflows. AI’s ability to automate admin- Another imperative is drawing up discernible patient privacy policies
istrative tasks reduces work pressure, allowing healthcare to thwart data intrusions. Healthcare entities must comply with data
professionals to devote more time to patient care. According to protection laws, such as HIPAA . Privacy and security policies must
Procter et al. [41], AI-based tools such as NLP, chatbots, and be defined in such ways that they include mandatory encryption of
virtual assistants can perform repetitive clerical duties such as sensitive data and the use of access controls and authentication
documentation, scheduling, and billing better than most human mechanisms.
doctors. This allows healthcare providers to focus on more
demanding tasks. Also, automating some administrative tasks 10.3. Fostering collaboration
minimizes burnout. AI tools also improve communication among Collaboration needs to be considered to reach the ultimate usage
providers and patients [42]. AI tools, such as chatbots, make it of AI in healthcare. The multidisciplinary approaches should be
easy to hold meaningful conversations with patients. These AI aligned with AI. It is essential to consider diverse teams in
communication tools make documenting patients’ health data healthcare to improve the precision of decisions. Clinicians may
easier because one can track and review the chats. work together to achieve reliability in clinical decision-making
processes related to AI as they develop a culture of collaboration
9.4.2. Required changes in education and training and partnership within healthcare organizations. Deep scrutiny
programs leads to perfection in AI.
Integrating AI into the healthcare professionals’ existing
curricula should begin to enhance AI literacy and competencies.
Healthcare providers must build a knowledge base on AI 11. Case studies
healthcare principles, processes, and patterns, including ML
algorithms, NLP, computer vision, and predictive analysis [11]. 11.1. Detailed analysis of successful artificial
Such efforts to implement AI literacy classes should focus on intelligence integration in healthcare settings
sensitizing healthcare professionals to the fundamentals of AI Radiologists perform serious and demanding tasks that can put
concepts and terms. These courses enable professionals to gain pressure on them. Interviews with several radiologists showed
extensive knowledge, heightening their understanding of AI- that, at times, the medical images they receive may impair their
driven technology and comprehending algorithmic findings. diagnosis and cause delays in treatment. The findings compiled
With an understanding of AI, one can make sound and clear- by the radiologists confirm the positive aspects of AI in
sighted decisions in healthcare practice. Healthcare practitioners healthcare.
should be able to evaluate data using data science, statistics, and
11.2. Lessons learned
programming languages [43]. The classes and training programs
should be based on experience rather than only solely on hearing Numerous healthcare organizations are adopting AI. Proper use of
and reading. The information they can gain from practicing with AI can improve the delivery of healthcare services. Healthcare
these AI tools will improve their competency in AI usage. organizations should consider collaborating with diverse teams,
such as IT professionals, when implementing AI. This collaboration
will reduce the risk of errors and biases in AI decision-making [47].
10. Navigating the integration of For instance, the collaboration could help physicians learn how to
artificial intelligence into healthcare modify AI tools in case of technical glitches. AI must be examined
and controlled to ensure its output is truthful. Evaluation of the AI
10.1. Developing ethical frameworks model is necessary to reduce errors and biases.
The first emphasis should be on transparency. According to
Iliashenko et al. [9], the healthcare sector and AI technology
engineers should implement transparent algorithms to aid in 12. Future directions
ethical decision-making. Transparency will help build trust and 12.1. Emerging trends in artificial intelligence and
enable healthcare professionals and patients to have confidence healthcare
in AI-driven technology. This will, in turn, facilitate the adoption
of AI in clinical practices. The adoption of AI in healthcare is increasing. AI is being used to
automate processes, such as documentation. Healthcare
Fairness and responsibility should also be sensitized. If the AI professionals increasingly use AI tools, such as NeuroScan, to
algorithms are poorly customized, there is a higher possibility of identify diagnostics, treatment planning, and patient management.
bias [44]. AI technologies must be trained to improve fairness in Per Kasula [48], AI-driven clinical decision support systems will
decision-making. Accountability could be enhanced by push ahead into predictive analytics modeling, risk stratification
implementing performance metrics that dictate how healthcare methodologies, and precision medicine.
providers should use AI tools [45]. Clear protocols should be set
to identify the healthcare provider responsible for any ethical or Additionally, according to Saraswat et al. [49], since the outbreak of
legal issues committed. COVID-19, there has been increased adoption of AI-powered remote
monitoring and telehealth solutions. The study showed that AI-
10.2. Enhancing data governance enabled tools are increasingly being used for remote monitoring.
This trend is expected to reduce the problems faced by marginalized
Data standardization eliminates uncertainties, which helps improve communities, as healthcare is being made more easily accessible.
data quality and availability. Developing standardization rules for
data is critical to ensure uniformity across health centers. Due to
standardization, it will be possible to reduce the interoperability
problem and the inconsistency with AI decision-making [46].

ACADEMIA MEDICINE 2024, 1 10 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

12.2. Research gaps and opportunities for future

investigation Funding
While many studies support AI’s positive impacts on healthcare, The author declares no financial support for the research,
significant research gaps exist. Firstly, there is a need to ensure authorship, or publication of this article.
that the AI models being used in healthcare are trustworthy.
Future research is needed to provide clear guidelines on how AI
models can improve their accuracy in decision-making [29]. Author contributions
Improved accuracy will enhance trust and lead to ethical The author confirms sole responsibility for this work. The author
decision-making. Additionally, there is reduced generalizability approves of this work and takes responsibility for its integrity.
in using AI. Future research is needed to develop a plan for
training AI models to ensure consistency.
Conflict of interest
13. Discussion The author declares no conflict of interest.

This paper presented the barriers and potential opportunities for

AI adoption in healthcare. AI has proved to be a powerful tool in Data availability statement
the healthcare industry due to its ability to transform healthcare
service provision through improved diagnosis. AI integration Data supporting these findings are available within the article, at
into healthcare systems has changed the formerly accepted https://doi.org/10.20935/AcadMed7382, or upon request.
procedures for accomplishing administrative tasks. Blockchain
technology has paved the way for a better restructuring of
administrative work [50]. AI-powered tools can perform
Institutional review board statement
administrative tasks, such as scheduling appointments, patient Not applicable.
registration, inventory management, processing claims, and
billing and coding. Healthcare providers must build a knowledge
base on AI healthcare principles, processes, and patterns, Informed consent statement
including ML algorithms, NLP, computer vision, and predictive Not applicable.
analysis. Such efforts to implement AI literacy classes should
focus on empowering the providers to use AI tools efficiently [51].
There is a need to ensure that the AI models used in healthcare Additional information
are trustworthy. Future research is needed to provide clear
guidelines on how the AI models can improve their decision- Received: 2024-05-14
making accuracy. Accepted: 2024-10-08
Published: 2024-10-30
14. Conclusions Academia Medicine papers should be cited as Academia Medicine
2024, ISSN 2994-435X, https://doi.org/10.20935/AcadMed7382.
The role AI can play in healthcare systems presents a landscape
The journal’s official abbreviation is Acad. Med.
of enormous opportunity and great challenge. Opportunities
range from predicting a transformative potential for diagnostics,
personalized medicine, drug discovery, and operational Publisher’s note
efficiency to guaranteeing diagnostic accuracy, predicting patient
deterioration, streamlining administrative tasks, or supporting Academia.edu Journals stays neutral with regard to jurisdictional
the development of personalized treatment plans. In this case, claims in published maps and institutional affiliations. All claims
the usage of AI holds tremendous potential for transformation in expressed in this article are solely those of the authors and do not
healthcare delivery. necessarily represent those of their affiliated organizations, or
those of the publisher, the editors, and the reviewers. Any
Even though the role of AI is enormous in the healthcare sector, product that may be evaluated in this article, or claim that may
it has some challenges that need to be resolved to realize its full be made by its manufacturer, is not guaranteed or endorsed by
potential. In this case, some challenges, such as data privacy, the publisher.
biases, and ethical considerations, are just a few issues that need
to be solved. Therefore, there is a need to harmonize AI
capabilities with practical, ethical, and regulatory requirements
Copyright
in healthcare. The alignment will keep transparency and © 2024 copyright by the authors. This article is an open access
certainty in AI decisions, which will cut down concerns about article distributed under the terms and conditions of the
bias. The potential of AI to transform healthcare emerges very Creative Commons Attribution (CC BY) license (https://creative
strongly from the above literature analysis; however, it will commons.org/licenses/by/4.0/).
require a very careful and well-regulated approach to its
integration into healthcare. The strategy of appropriately
balancing strengths with challenges will allow the full utilization References
of AI in healthcare. 1. Alowais SA, Alghamdi SS, Alsuhebany N, Alqahtani T, Alshaya
AI, Almohareb SN, et al. Revolutionizing healthcare: the role of

ACADEMIA MEDICINE 2024, 1 11 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

artificial intelligence in clinical practice. BMC Med Educ. 15. Shah R, Chircu A. IoT and AI in healthcare: a systematic
2023;23(1):689. doi: 10.1186/s12909-023-04698-z literature review. Issues Inf Syst. 2018;19(3):33–41.
2. Lehmann LS. Ethical challenges of integrating AI into 16. Shinners L, Aggar C, Grace S, Smith S. Exploring healthcare
healthcare. In Artificial intelligence in medicine. Cham: professionals’ understanding and experiences of artificial
Springer International Publishing; 2021. p. 1–6. Available intelligence technology used in healthcare delivery: an
from: https://link.springer.com/content/pdf/10.1007/978- integrative review. Health Inform J. 2020;26(2):1225–36.
3-030-58080-3_337-2.pdf doi: 10.1177/1460458219874641
3. Yadav D, Sehrawat R. Artificial intelligence integration in 17. Sujith AVLN, Sajja GS, Mahalakshmi V, Nuhmani S,
healthcare and medicine. Artif Intell. 2018;7(4):11–7. Prasanalakshmi B. Systematic review of smart health
monitoring using deep learning and artificial intelligence.
4. Klumpp M, Hintze M, Immonen M, Ródenas-Rigla F, Pilati
Neurosci Inform. 2022;2(3):100028.
F, Aparicio-Martínez F, et al. Artificial intelligence for
hospital health care: application cases and answers to 18. Reddy S, Fox J, Purohit MP. Artificial intelligence-enabled
challenges in European hospitals. Healthcare. 2021;9(8): healthcare delivery. J R Soc Med. 2019;112(1):22–8. doi:
961. doi: 10.3390/healthcare9080961 10.1177/0141076818815510
5. Bonnist E. The future of artificial intelligence in the 19. Yang L, Ene IC, Arabi Belaghi R, Koff D, Stein N, Santaguida
healthcare industry; 2021 [cited 2024 Oct 19]. Available P. Stakeholders’ perspectives on the future of artificial
from: https://digitalworks.union.edu/theses/2405/ intelligence in radiology: a scoping review. Eur Radiol.
2022;32(3):1477–95. doi: 10.1007/s00330-021-08214-z
6. Van der Schaar M, Alaa AM, Floto A, Gimson A, Scholtes S,
Wood A, et al. How artificial intelligence and machine learning 20. Aggarwal N, Matheny ME, Shachar C, Wang SX, Thadaney-
can help healthcare systems respond to COVID-19. Mach Israni S. Artificial intelligence in healthcare; 2022 [cited 2024
Learn. 2021;110:1–14. doi: 10.1007/s10994-020-05928-x Oct 19]. Available from: https://academic.oup.com/edited-
volume/41989/chap ter/355439553
7. Noorbakhsh-Sabet N, Zand R, Zhang Y, Abedi V. Artificial
intelligence transforms the future of health care. Am J Med. 21. Lau AY, Staccini P. Artificial intelligence in health: new
2019;132(7):795–801. opportunities, challenges, and practical implications. Yearb
Med Inform. 2019;28(1):174–8. doi: 10.1055/s-0039-1677935
8. Uzun Ozsahin D, Ikechukwu Emegano D, Uzun B, Ozsahin
I. The systematic review of artificial intelligence applications 22. Strindlund M. The impact of artificial intelligence on CT
in breast cancer diagnosis. Diagnostics. 2022;13(1):45. doi: imaging. 2021 [cited 2024 Oct 19]. Available from:
10.3390/diagnostics13010045 https://www.itnonline.com/article/impact-artificial-
intelligence-ct-imaging
9. Iliashenko O, Bikkulova Z, Dubgorn A. Opportunities and
challenges of artificial intelligence in healthcare. In E3S Web 23. Lambert SI, Madi M, Sopka S, Lenes A, Stange H, Buszello
of Conferences (Vol. 110). EDP Sciences; 2019. p. 02028. CP, et al. An integrative review on the acceptance of artificial
Available from: https://www.e3s-conferences.org/articles/ intelligence among healthcare professionals in hospitals.
e3sconf/abs/2019/36/e3sconf_spbwosce2019_02028/e3sc NPJ Digit Med. 2023;6(1):111.
onf_spbwosce2019_02028.html
24. Ali T, Ahmed S, Aslam M. Artificial intelligence for
10. Pirtle CJ, Payne K, Drolet BC. Telehealth: legal and ethical antimicrobial resistance prediction: challenges and oppor-
considerations for success. Telehealth and Medicine Today. tunities towards practical implementation. Antibiotics.
2019 [cited 2024 Oct 19]. Available from: https://telehealthand 2023;12(3):523. doi: 10.3390/antibiotics12030523
medicinetoday.com/index.php/journal/article/view/144
25. Singh AV, Chandrasekar V, Paudel N, Laux P, Luch A,
11. Choudhury A, Asan O. Impact of accountability, training, Gemmati D, et al. Integrative toxicogenomics: advancing
and human factors on the use of artificial intelligence in precision medicine and toxicology through artificial
healthcare: exploring healthcare practitioners’ perceptions intelligence and OMICs technology. Biomed Pharmacother.
in the US. Hum Factors Healthcare. 2022;2:100021. Doi: 2023;163:114784. doi: 10.1016/j.biopha.2023.114784
10.1016/j.hfh.2022.100021
26. Wan TT. Convergence of artificial intelligence research in
12. Subasi A. Use of artificial intelligence in Alzheimer’s disease healthcare: trends and approaches. J Integr Des Process Sci.
detection. In: Barh D, editor. Artificial intelligence in 2022;26:291–305.
precision health. London: Academic Press; 2020. doi:
27. Chan CYT, Petrikat D. Strategic applications of artificial
10.1016/B978-0-12-817133-2.00011-2
intelligence in healthcare and medicine. J Med Health Stud.
13. Naik N, Zeeshan Hameed BM, Shetty DK, Swain D, Shah M, 2023;4(3):58–68. doi: 10.32996/jmhs.2023.4.3.8
Paul R. Legal and ethical consideration in artificial intelligence
28. Rai BK, Kumar G, Balyan V. AI and blockchain in healthcare.
in healthcare: who takes responsibility? Front Surg.
Singapore: Springer Nature; 2023.
2022;9:862322. doi: 10.3389/fsurg.2022.862322
29. Baker S, Xiang W. Artificial intelligence of things for smarter
14. Samad A. Internet of things integrated with blockchain and
healthcare: a survey of advancements, challenges, and oppor-
artificial intelligence in healthcare system. Res J Comput
tunities. IEEE Commun Surv Tutor. 2023 [cited 2024 Oct 19].
Syst Eng. 2022;3(1):1–6.
Available from: https://ieeexplore.ieee.org/abstract/docum
ent/10066875

ACADEMIA MEDICINE 2024, 1 12 of 13

https://www.academia.edu/journals/academia-medicine/about https://doi.org/10.20935/AcadMed7382

30. Radanliev P, De Roure D, Maple C, Ani U. Methodology for 40. Wang F, Preininger A. AI in health: state of the art,
integrating artificial intelligence in healthcare systems: challenges, and future directions. Yearb Med Inform.
learning from COVID-19 to prepare for disease X. AI Ethics. 2019;28(1):16–26. doi: 10.1055/s-0039-1677908
2022;2:623–30. doi: 10.1007/s43681-021-00111-x
41. Procter R, Tolmie P, Rouncefield M. Holding AI to account:
31. Tyson A, Pasquini G, Spencer A, Funk C. 60% of Americans challenges for the delivery of trustworthy AI in healthcare.
would be uncomfortable with provider relying on AI in their ACM Trans Comput Hum Interact. 2023;30(2):1–34. doi:
own health care. Pew Research Center. 2023 [cited 2024 Oct 10.1145/3577009
21]. Available from: https://www.pewresearch.org/science/
42. Boussina A, Shashikumar S, Amrollahi F, Pour H, Hogarth M,
2023/02/22/60-of-americans-would-be-uncomfortable-
Nemati S. Development & deployment of a real-time healthcare
with-provider-relying-on-ai-in-their-own-health-care/
predictive analytics platform. In 2023 45th Annual
32. Awwalu J, Garba AG, Ghazvini A, Atuah R. Artificial International Conference of the IEEE Engineering in Medicine
intelligence in personalized medicine applications AI & Biology Society (EMBC). Sydney: IEEE; 2023. p. 1–4.
algorithms to solve personalized medicine problems. Int J Available from: https://ieeexplore.ieee.org/abstract/docu
Comput Theory Eng. 2015;7(6):439. ment/10340351/

33. Boussina A, Shashikumar SP, Malhotra A, Owens RL, El- 43. Tyrväinen P, Silvennoinen M, Talvitie-Lamberg K, Ala-Kitula
Kareh R, Longhurst CA, et al. Impact of a deep learning A, Kuoremäki R. Identifying opportunities for AI applications
sepsis prediction model on quality of care and survival. NPJ in healthcare—renewing the national healthcare and social
Digit Med. 2024;7(1):14. services. In 2018 IEEE 6th International Conference on Serious
Games and Applications for Health (SeGAH). IEEE; 2018.
34. Panesar A. Machine learning and AI for healthcare.
p. 1–7. Available from: https://ieeexplore.ieee.org/abstract/
Coventry: Apress; 2019. p. 1–73. doi: 10.1007/978-1-4842-
document/8401381/
6537-6
44. Osman Andersen T, Nunes F, Wilcox L, Kaziunas E,
35. Racine E, Boehlen W, Sample M. Healthcare uses of artificial
Matthiesen S, Magrabi F. Realizing AI in healthcare:
intelligence: challenges and opportunities for growth.
challenges are appearing in the wild. In Extended Abstracts
Healthc Manage Forum. 2019;32(5); 2019. p. 272–5. doi:
of the 2021 CHI Conference on Human Factors in
10.1177/0840470419843831
Computing Systems; 2021. p. 1–5. doi: 10.1145/34117
36. Han C, Rundo L, Murao K, Nemoto T, Nakayama H. 63.3441347
Bridging the gap between AI and healthcare sides: towards
45. Guan J. Artificial intelligence in healthcare and medicine:
developing clinically relevant AI-powered diagnosis
promises, ethical challenges, and governance. Chin Med Sci
systems. In Artificial Intelligence Applications and
J. 2019;34(2):76–83. doi: 10.24920/003611
Innovations: 16th IFIP WG 12.5 International Conference,
AIAI 2020, Neos Marmaras, Greece, June 5–7, 2020, 46. Väänänen A, Haataja K, Vehviläinen-Julkunen K, Toivanen
Proceedings, Part II 16. Springer International Publishing; P. AI in healthcare: a narrative review. F1000Res. 2021;10:6.
2020. p. 320–33. doi: 10.1007/978-3-030-49186-4_27 doi: 10.12688/f1000research.26997.2

37. Lee D, Yoon SN. Application of artificial intelligence-based 47. Jiang F, Jiang Y, Zhi H, Dong Y, Li H, Ma S, et al. Artificial
technologies in the healthcare industry: opportunities and intelligence in healthcare: past, present and future. Stroke
challenges. Int J Environ Res Public Health. 2021;18(1):271. Vasc Neurol. 2018;2(4):e000101.
doi: 10.3390/ijerph18010271
48. Kasula BY. Framework development for artificial
38. Sujan M, Furniss D, Hawkins RD, Habli I. Human factors of intelligence integration in healthcare: optimizing patient
using artificial intelligence in healthcare: challenges across care and operational efficiency. Trans Latest Trends IoT.
industries. In Safety-Critical Systems Symposium. York; 2023;6(6):77–83.
2020. Available from: https://eprints.whiterose.ac.uk/
49. Saraswat D, Bhattacharya P, Verma A, Prasad VK, Tanwar S,
159105/1/19_SUJAN_FURNISS_HAWKINS_HABLI_HU
Sharma G, et al. Explainable AI for healthcare 5.0:
MAN_FACTORS_OF_USING_ARTIFICIAL_INTELLIGE
opportunities and challenges. IEEE Access. 2022;10:
NCE_IN_HEALTHCARE_CHALLENGES_THAT_STRET
84486–517.
CH_ACROSS_INDUSTRIES.PDF
50. Aung YY, Wong DC, Ting DS. The promise of artificial
39. Choudhury A, Asan O. Human factors: bridging artificial
intelligence: a review of the opportunities and challenges of
intelligence and patient safety. In Proceedings of the
artificial intelligence in healthcare. Br Med Bull. 2021;
International Symposium on Human Factors and
139(1):4–15.
Ergonomics in Health Care (Vol. 9, No. 1). Los Angeles, CA:
SAGE Publications; 2020. p. 211–5. doi: 10.1177/ 51. Nasr M, Islam MM, Shehata S, Karray F, Quintana Y. Smart
2327857920091007 healthcare in the age of AI: recent advances, challenges, and
prospects. IEEE Access. 2021;9:145248–70.

ACADEMIA MEDICINE 2024, 1 13 of 13