Content uploaded by Bruno De Lièvre
Author content
All content in this area was uploaded by Bruno De Lièvre on Aug 02, 2023
Content may be subject to copyright.
Jl. of Interactive Learning Research (2023) 34(2), 401-424
Integrating Articial Intelligence into Medical
Education: Lessons Learned From a Belgian
Initiative
ILARIA PIZZOLLA
University of Mons, Belgium
ilariavpizzolla@gmail.com
RANIA ARO
University of Mons, Belgium
rania.aro@umons.ac.be
PIERRE DUEZ
University of Mons, Belgium
pierre.duez@umons.ac.be
BRUNO DE LIÈVRE
University of Mons, Belgium
bruno.delievre@umons.ac.be
GIOVANNI BRIGANTI
University of Mons, Belgium
giovanni.briganti@hotmail.com
In the last decades, the medical practice has been facing note-
worthy transformations driven by the advancement of in-
novative technologies like Articial Intelligence (AI). This
rapid and widespread transition generated the increasing need
for an adequate education curriculum, capable of properly
teaching medical students about the prospects and potentials
of AI in healthcare. In this paper, we aim to present and de-
scribe the elaboration and implementation of a new academic
program at the University of Mons (UMONS) designed to
educate medical students about AI in healthcare. The course
402 Pizzolla, Aro, Duez, De Lièvre, and Briganti
was implemented in the 2022-2023 academic year aiming
to train the next generation of healthcare professionals to ef-
fectively leverage AI in their work, ultimately leading to im-
proved patient outcomes and advances in medical research.
Keywords: Articial intelligence in medical education; Ma-
chine learning; Deep learning models; Educational technolo-
gies; Curriculum reforms, Medical curricula.
INTRODUCTION
AI has been widely recognized as one of the most transformative and
ground-breaking technological advancements in healthcare (Obermeyer &
Emanuel, 2016). From disease diagnosis to drug development, AI has suc-
cessfully proven its ecacy in numerous modern-day medicine (Ayyad et
al., 2021; Briganti & Le Moine, 2020; Dlamini et al., 2022; Erickson et
al., 2017b; Shmatko et al., 2022) , but its adoption has yet to be improved
(Briganti & Le Moine, 2020). The eectiveness of AI-powered technologies
in the eld of healthcare relies heavily on the participation of healthcare
professionals (HCPs) in the creation and testing of these technologies (Cha-
row et al., 2021).
The need for integrating AI in medical education has been abundantly
discussed (Charow et al., 2021; Civaner et al., 2022; Sapci & Sapci, 2020)
and literature showed that medical students seem to achieve little knowl-
edge about AI and its applications in clinical practice during their academic
education (Sapci & Sapci, 2020). To respond to the lack of AI literacy, a
new mandatory AI program has been implemented at the University of
Mons (UMONS) for the bachelor’s degree in medicine.
The scope of this work is to showcase and analyse the results, advan-
tages, and limitations of this program which, by leveraging AI, aims to pro-
vide students with real-time feedback, adaptive learning paths, and interac-
tive simulations that replicate real-world clinical scenarios. The goal is to
create a more engaging and eective learning experience that can ultimately
lead to better patient outcomes.
This work will discuss the potential benets of AI in medical train-
ing and investigate the multiple reasons supporting the implementation of
teaching AI in medical curricula. Subsequently, the paper will rst provide
an overview of the current state of AI in medical education and the chal-
Integrating Articial Intelligence into Medical Education 403
lenges faced by medical students, mostly in relation to the lack of AI educa-
tion and the expanding applications of modern technologies in the clinical
sphere. Furthermore, the paper will present and describe the development
and implementation of the new AI program developed at the University of
Mons, Belgium in 2022. The paper will then discuss the evaluation methods
used to assess the eectiveness of the AI program and its ndings, including
student performance and feedback. Finally, we will interpret the results in
light of existing literature, discussing the future educational applications of
AI in healthcare training and its limitations.
Overall, this paper seeks to contribute to the ongoing eorts to improve
and enhance medical education and integrate AI adequate training which
will make HCPs capable to operate eciently in the era of AI and improv-
ing patient care and personalized healthcare.
LITERATURE REVIEW
Medical students are asked to analyse, interpret, and memorize a vast
amount of medical information during their academic training since clinical
expertise is interconnected to having ecient knowledge (Densen, 2011). A
large part of medical training focuses on consuming as much information as
possible and learning how to apply this knowledge to patient care (Paran-
jape et al., 2019). However, since this process is still mainly memorization-
based, the use of AI in the process of manipulating data using has signi-
cant implications for medical education. The current and largely memori-
zation-based curriculum must transition to one that teaches competence in
the eective integration and employment of data from a developing range of
sources (Wartman & Combs, 2018).
Undeniably, in the last decades, medical progress made medical infor-
mation grow at a breakneck speed and the number of data students need to
memorize continues to develop (Densen, 2011). With the advent of new
technologies and medical discoveries, the expansion of medical informa-
tion resulted in the increasing diculty for medical students, physicians and
HCPs to keep pace (Paranjape et al., 2019).
That is where AI steps in. Indeed, articial intelligence is capable of
organizing and merging considerable volumes of data and perfecting the
decision-making process of HPCs, thus simplifying the path leading to di-
agnosis identication and recommend treatments (Briganti, 2023; Coppola
et al., 2021; Paranjape et al., 2019; Topol, 2019). Recent literature endorsed
the role of AI and machine-learning predictive algorithms as extremely ad-
404 Pizzolla, Aro, Duez, De Lièvre, and Briganti
vantageous catalysers and fast analysers of clinical data; these tools repre-
sent indeed the keys to unlocking the data capable of informing real-time
decisions (Chen & Asch, 2017).
Predictive algorithms are already used to avert hospitalization for pa-
tients with low-risk pulmonary embolisms (PESI), prioritize patients for liv-
er transplantation by means of MELD scores (Chen & Asch, 2017). In the
future deep learning is expected to be utilized by a wide range of clinicians,
from specialty doctors to paramedics (Topol, 2019). The implementation of
AI tools has been discussed, tested and/or adopted in several medical areas,
such as histopathology (Abdelsamea et al., 2022; Acs & Hartman, 2020; At-
tallah, 2021; Ayyad et al., 2021; Kayser et al., 2009; Lui et al., 2020; Sal-
danha et al., 2022; Shmatko et al., 2022), radiology (Brady & Neri, 2020;
Codari et al., 2019; Erickson et al., 2017b; Jin et al., 2021; Pianykh et al.,
2020; Reyes et al., 2020; Tajmir & Alkasab, 2018), cancer prevention (Ayy-
ad et al., 2021; Byng et al., 2022; Dlamini et al., 2022; Hegde et al., 2022;
Saldanha et al., 2022; Shmatko et al., 2022).
HCPs must develop and master the skill to interpret such results (Brig-
anti, 2023; Briganti & Le Moine, 2020; Paranjape et al., 2019). As we
are shifting from the Information Age to the Age of Articial Intelligence
(Wartman & Combs, 2018). This technological upheaval requires medical
workers to learn how to manage the exponential growth of available data.
Hence, the actual medical education model will and needs to undergo sub-
stantial reform. Indeed, the expansion of knowledge is likely to force med-
ical schools to dene those concepts that form the essential core of what
students must learn. Present medical curricula mainly leave to the students
the task to form the connections that integrate discipline-based knowledge
with patient complaint/disease-focused information (Densen, 2011). By
implementing AI training in medical education, medical curricula would be
nally equipping future physicians and HCPs with the ability to interpret
enormous amounts of medical data using AI, thus providing better patient
care, diagnosis identication and treatment, in a tighter timeframe. As a
matter of facts, AI applications not only simplify data collection and inter-
pretation, but it also allows it to be done in a shorter amount of time con-
sequently reducing inaccuracies due to human fatigue and simultaneously
lessening healthcare costs (Coppola et al., 2021; Paranjape et al., 2019). De-
cision management can assist the examination of massive amounts of data
and enable the physician to make an informed and meaningful evaluation
(Paranjape et al., 2019).
The implementation of AI, however, does not have to exist in a logic
of exclusion of the human work. On the contrary, the two need to co-ex-
Integrating Articial Intelligence into Medical Education 405
ist and co-operate. Human interaction is not only essential because of the
social and psychological aspects of medical practice, which could also be
accomplished by dierent professional gures (Masters, 2019) but also be-
cause such technologies need to be wisely mastered. Medical practitioners
must be aware of AI implication in order to master the machine, since the
aim of the implementation of AI in healthcare is not to replace doctors, but
enhance and assist their role, combining machine-learning software with the
nest human clinician “hardware” (Chen & Asch, 2017). Solely the equi-
librium of the two elements will allow medical practitioners to dispense a
quality of care that “outperforms what either can do alone” (Chen & Asch,
2017). Despite the fact that computers may at times appear to detect pat-
terns that are imperceptible to humans , the human element is still essen-
tial and therefore needs to be accurately and adequately trained about the
machine applications and potential (Erickson et al., 2017b). Recent research
support the thesis of a multilateral combination of players in the upcoming
healthcare systems: in the future, medical practice is expected to involve a
deliberate collaboration between physicians, other HCPs, patients, and ma-
chines, comprising both hardware and software (Wartman & Combs, 2018).
Medical curricula may benet of a reboot (Wartman & Combs, 2018)
incorporating AI fundamentals in order to empower future HCPs and as-
sure improved patient-centered care (Charow et al., 2021). The lack of AI
literacy in medical education today represents not only a widely discussed
dilemma in literature, but also “a signicant barrier to the adoption and use
of AI-enabled technologies to their full capacity in various medical special-
ties.” (Charow et al., 2021). Although the National Academy of Medicine’s
Quintuple Aim Model (Cox et al., 2017) encourages the integration of AI
technologies in healthcare to enhance patient care, a shortage of AI educa-
tion may be hindering health care systems from fully embracing and imple-
menting these technologies (Charow et al., 2021).
In order to ensure their safe and eective application, it is crucial to
have a thorough understanding of the properties of machine learning tools.
(Erickson et al., 2017). As we enter the era of articial intelligence, the ac-
quisition of new skills and expertise will become compulsory. These may
include utilizing the insights from cognitive psychology, fostering a closer
relationship between humans and machines in education, and increasing the
use of simulations that focus on integrating machines into healthcare deliv-
ery and empowering patients to actively participate in their care. Addition-
ally, learners must have a fundamental grasp of how data is collected, anal-
ysed, and ultimately personalized through articial intelligence applications
in clinical delivery (Wartman & Combs, 2018).
406 Pizzolla, Aro, Duez, De Lièvre, and Briganti
Overview of recent literature related to teaching AI to medical
students
As a result of the rapid growth of AI applications in numerous elds, in
the latest years the interest in integrating AI education into medical school
curricula has signicantly ourished and has been correspondingly nour-
ished by literature. As AI is increasingly being used in healthcare, it is im-
portant for medical students to understand its principles and potential ap-
plications (Briganti, 2023) . In this section, we review some of the recent
literature related to teaching AI to medical students.
Recent literature has highlighted the potential of AI in medical educa-
tion, particularly in helping medical students master complex medical con-
cepts and improve their clinical reasoning skills. Various studies have insist-
ed on the crucial role of implementing AI education in medical studies and
numerous authors, having analysed the current landscape of medical edu-
cation, assessed the ineectiveness of medical curricula to integrate proper
AI training. Nonetheless, AI literacy is far from being widespread among
healthcare students and professionals. Despite the potential benets of AI
over traditional methods, there is a notable absence of its implementation in
curriculum reviews. One possible explanation for it, is the inadequate level
of digitalization within medical education’s learning management systems,
which is necessary to develop a comprehensive curriculum map (Chan &
Zary, 2019)
Moreover, medical students seem to lack the basic scientic knowledge
correlated to the role of AI in their own professional environment. Although
clinicians are asked to master and develop their knowledge of AI in health
care, medical education fails to provide them with such skills (Sapci & Sap-
ci, 2020). Medical students are particularly aware of the increasing role AI
is taking in healthcare and have expressed the necessity to be more educated
about its potential and applications in medical practice (Civaner et al., 2022;
Frommeyer et al., 2022). In 2022, a study highlighted the lack of familiar-
ity of future physicians with AI proved that they would become better phy-
sicians with the widespread use and knowledge of AI. It shows that they
student do not receive enough adequate training on AI in medicine event if
predominantly favourable to structured training on AI applications during
medical education (Civaner et al., 2022).
Additionally, HCPs are inadequately sensibilized and trained about
AI, thus preventing patients from receiving AI improved care. Recent lit-
erature asses that only a small percentage of future HCPs feels adequately
qualied to inform patients about the features and hazards associated with
Integrating Articial Intelligence into Medical Education 407
AI technologies (Civaner et al., 2022), thus enhancing the limited knowl-
edge of medical professionals when it comes to provide reliable information
about AI to their patients (Frommeyer et al., 2022). Furthermore, HCPs can
benet from very few chances for to receive education and training on AI.
Most realistically, AI will continue to drastically transform healthcare in
the following decades. To respond to the changing landscape of medicine,
and train eective medical practitioner, academic education must adapt and
reform their curriculum, incorporating teaching AI (Wartman & Combs,
2018).
Based on existing literature, we have identied three main issues in
present medical education: rstly, AI not widespread as an educational sub-
ject in any healthcare linked curricula; secondly, medical students lack the
basic scientic knowledge related to the role of AI in their own professional
environment; thirdly, HCPs are inadequately sensibilized and trained about
AI, thus preventing patient from receiving AI improved care. Subsequently,
AI programs in medical education, internationally, should include the fol-
lowing elements: a) solve the lack of AI education in medical eld; b) sensi-
bilize medical students, which will be future HCPs, in the eld of new tech-
nologies applied to medicine; c) consider the lack of students’ prior scien-
tic knowledge, which is necessary to approach AI learning.
Challenges and opportunities identied in the literature
AI is becoming increasingly prevalent in the healthcare industry, and
HCPs must be prepared to understand and work with AI systems. However,
current AI education and training opportunities for HCPs are limited, lead-
ing to a shortage of AI-literate HCPs (Frommeyer et al., 2022). This section
explores the challenges and opportunities of AI education in healthcare, as
identied in the literature.
One of the primary challenges of AI education is the lack of standard-
ized and consistent curricula across healthcare institutions (Chan & Zary,
2019; Charow et al., 2021). This is linked to the relatively recent and rapid
emergence of AI in healthcare. Numerous healthcare-linked curricula are
still exploring the potential of AI and are in the process of developing or
testing their own programs to teach AI concepts to HCPs. As a result, there
is a wide-ranging dissimilarity in the complexity and extensiveness of AI
education across dierent healthcare educational institutions due to various
constraints, such as nancial limitations and inadequate technological infra-
structure (Charow et al., 2021; Foster & Tasnim, 2020).
408 Pizzolla, Aro, Duez, De Lièvre, and Briganti
Additionally, the rapidly evolving nature of AI technologies may make
it dicult for educators to keep the curriculum up to date (Chan & Zary,
2019; Kang et al., 2017). The integration of AI education may require sig-
nicant resources, including expertise from AI professionals, access to rel-
evant data sets, and development of educational materials (Charow et al.,
2021; Densen, 2011; He et al., 2019; Reyes et al., 2020). Furthermore, AI
education may require signicant nancial investments in technology, infra-
structure, and training, which may not be feasible for some healthcare insti-
tutions (Charow et al., 2021).
Moreover, the shortage of qualied AI educators in healthcare makes
it dicult to provide adequate training to HCPs and students, and certain
healthcare organizations may not be equipped with AI well-versed teach-
ing personnel to provide thorough education and training on articial intel-
ligence (Chan & Zary, 2019; Charow et al., 2021; Frommeyer et al., 2022).
Sta members may not have the requisite skills and knowledge to eec-
tively implement AI solutions in healthcare operations or the adequate time
(Charow et al., 2021; Wartman & Combs, 2018). This can lead to a slower
adoption of AI and hinder the potential benets it can bring to the organiza-
tion and patient care.
Furthermore, the growing importance of AI in healthcare necessitates
a thorough understanding of its applications, limitations, and ethical con-
siderations for future medical professionals (Densen, 2011; Paranjape et
al., 2019; Srivastava & Waghmare, 2020). Incorporating AI into medicine
brings forth a range of ethical concerns that must be considered in order to
ensure responsible and equitable applications of the technology. One major
ethical issue is the potential for biased decision-making resulting from bi-
ased training data, which could lead to inequitable healthcare outcomes for
certain populations (Obermeyer et al., 2019). Additionally, the use of AI in
medicine raises concerns about patient privacy and data security, as large
amounts of personal health information are required to train and operate
AI systems eectively (Park et al., 2019). Besides, the “black box” nature
of certain AI algorithms may challenge traditional notions of medical ac-
countability, as it may become dicult to determine responsibility in cases
of misdiagnosis or treatment errors (Castelvecchi, 2016; Nicholson Price II,
2014). Ensuring transparency and explainability in AI systems is crucial for
upholding trust and ethical standards in medicine (Holzinger et al., 2019).
Lastly, the widespread adoption of AI may lead to changes in the doctor-pa-
tient relationship, as AI could alter the nature of medical decision-making,
shifting the balance between human expertise and algorithmic recommenda-
tions (Char et al., 2018). Addressing these ethical concerns is essential to
Integrating Articial Intelligence into Medical Education 409
ensure that AI is implemented in a manner that upholds the core principles
of medical ethics and preserves the trust between healthcare professionals
and their patients (Kostkova, 2015).
Despite the challenges, AI education presents signicant opportuni-
ties for future HCPs. One of the primary opportunities is improving patient
outcomes and quality of care (Charow et al., 2021; Kang et al., 2017). AI
technology can help HCPs to identify potential health issues earlier, predict
future health risks, and personalize treatment plans for patients (Srivastava
& Waghmare, 2020). This can lead to improved patient outcomes, increased
eciency, and reduced healthcare costs (Hu et al., 2022; Kang et al., 2017;
Srivastava & Waghmare, 2020).
Moreover, AI education can improve collaboration, time-management
and communication between HCPs and other stakeholders in the healthcare
system. AI technology can help HCPs to collaborate and share data more ef-
fectively and in a shorter period of time, leading to improved care coordina-
tion, enhanced work-ow and better patient outcomes (Lillehaug & Lajoie,
1998; Long & Magerko, 2020; Makridakis, 2017; Robeznieks, 2018).
Description and rationale of the AI program developed for medi-
cal students
To respond to the need of AI training in healthcare related education,
the University of Mons, a Belgian institution that recently launched a Chair
of AI and Digital Medicine, has taken a proactive approach to integrating
AI education within its medical curriculum. This mandatory educational
program aims to equip medical students with the necessary knowledge and
skills related to AI and its applications in healthcare. The program is struc-
tured in a way that encourages students to not only grasp theoretical con-
cepts but also to apply them in real-life situations through group work and
data analysis.
The AI and Digital Medicine educational program consists of ve main
modules, followed by a practical group assignment. Each module is de-
signed to introduce medical students to dierent aspects of AI and its ap-
plications in healthcare:
1. Introduction to AI as a eld
Time: 2 hours.
Teaching Method: Online recorded lecture and supplementary
reading materials
410 Pizzolla, Aro, Duez, De Lièvre, and Briganti
Process: Students learn about the history and evolution of AI, its
various applications, and its potential impact on society.
Performance Assessment: Short online quiz to assess the under-
standing of the fundamental concepts of AI.
2. Introduction to machine learning and expert systems
Time: 4 hours.
Teaching method: Online recorded lecture, supplementary reading
materials, and hands-on exercises.
Process: Students learn about dierent machine learning algo-
rithms, expert systems, and their applications in various elds.
Performance Assessment: Hands-on exercises where students ap-
ply machine learning algorithms on sample data and an online quiz
to assess their understanding of the concepts.
3. Introduction to machine learning in healthcare
Time: 4 hours.
Teaching Method: Online recorded lecture, supplementary reading
materials, and case studies.
Process: Students learn about the specic applications of machine
learning in healthcare, such as diagnostics, treatment planning, and
patient monitoring.
Performance Assessment: Case study analysis where students
apply machine learning concepts to healthcare scenarios and an
online quiz to assess their understanding of the subject matter.
4. Introduction to machine vision
Time: 3 hours.
Teaching Method: Online recorded lecture, supplementary reading
materials, and practical exercises.
Process: Students learn about the basics of machine vision, includ-
ing image processing, feature extraction, and object recognition.
Performance Assessment: Practical exercises where students apply
machine vision techniques on sample images and an online quiz to
test their understanding of the concepts.
5. Introduction to image recognition in healthcare
Time: 3 hours.
Teaching Method: Online recorded lecture, supplementary reading
materials, and hands-on exercises.
Integrating Articial Intelligence into Medical Education 4 11
Process: Students learn about the applications of image recognition
in healthcare, such as medical imaging, diagnostics, and treatment
planning.
Performance Assessment: Hands-on exercises where students ap-
ply image recognition techniques on medical images and an online
quiz to assess their understanding of the subject matter.
In addition to the modules, students were required to work in groups to
analyze datasets using machine learning techniques and produce a consoli-
dated report. This project served as the primary performance assessment,
where students applied the knowledge and skills gained throughout the pro-
gram. Instructors evaluated the reports based on the accuracy and relevance
of the applied methods, the clarity and coherence of the presentation, and
the critical evaluation of the results. This approach ensured constructive
alignment, as the learning objectives, teaching methods, and assessments
were consistent throughout the course.
Besides the core learning materials, ve supplementary online modules
to acquaint the students with the specialized domain of Digital Medicine
were developed. These modules provided essential context for understand-
ing the evolving landscape of healthcare and the role that technology plays
in it. The topics covered included: (a) an overview of health data and its sig-
nicance; (b) an introduction to Electronic Health Records and their role in
modern healthcare; (c) an exploration of the ways in which digital medicine
is transforming healthcare delivery; (d) a discussion of the impact of digi-
tal medicine on research and innovation within the medical eld; and (e) a
comprehensive analysis of the various applications of AI in healthcare.
METHOD
Sample
In the inaugural edition of the AI and Digital Medicine educational
program during the 2022-2023 academic year, a cohort of 25 students was
recruited to participate. In order to assess the eectiveness of the AI and
Digital Medicine educational program, a survey was conducted among
these 25 medical students. This evaluation aimed to gather their feedback
and opinions on various aspects of the program, including the quality and
relevance of the course content, the teaching methods employed, and their
overall satisfaction with the program. By engaging with the students in this
412 Pizzolla, Aro, Duez, De Lièvre, and Briganti
manner, the University of Mons sought to gain a deeper understanding of
the program’s strengths and areas for improvement, ensuring that the course
remains responsive to the needs and expectations of its participants. The
survey results provided valuable insights into the students’ perspectives on
the program and their experiences, which can be used to rene and enhance
future iterations of the AI and Digital Medicine educational program, ul-
timately improving the quality of AI education for medical students at the
University of Mons. We introduce the survey and the analysis of data in the
methods section.
As the program aimed to integrate AI education within the medical cur-
riculum, it was crucial to have a representative sample that could collective-
ly contribute to the development and evaluation of the program. By includ-
ing these 25 students, the University of Mons was able to gather valuable
insights and feedback on the eectiveness of the AI and Digital Medicine
educational program, informing potential improvements and adjustments
for future iterations. 20 out of the 25 students participated in the survey
(80% participation rate).
These students represented a broad range of backgrounds and interests
within the eld of medicine, thereby ensuring a rich and varied perspective
on the applications of AI in healthcare. Students are enrolled in a medical
school program, currently in the third year of medical school. In Belgium,
medical school is a six-year long program split into a bachelor’s degree (3
years) and master’s degree (3 years). At the University of Mons, like many
other medical schools in Belgium, there is no major or minor system in the
curriculum. In the third year of medical school, students are typically 20-22
years old. We chose not to collect age or gender information because of the
low sample size and to respect anonymity.
Survey
In order to assess the students’ perception of the AI program, we con-
ducted a survey composed of 13 questions. The instructor of the course re-
cruited the participants of our study by sending a Google Form question-
naire to all the medical students that were registered for the course. The
survey aimed to evaluate various aspects of the course, such as the poten-
tial impact of AI on medicine, the eectiveness of medical education in
teaching AI concepts, students’ expectations of the program, and the over-
all quality of the course. Additionally, the survey inquired about students’
satisfaction with the topics covered and instructors, their understanding of
Integrating Articial Intelligence into Medical Education 413
AI’s role in medicine, the balance between real-world examples and theo-
retical content, as well as the presentation of AI’s benets, limitations, and
risks in medical diagnosis and treatment. Furthermore, the survey explored
the participants’ condence in critically evaluating AI-generated diagnoses
and treatment recommendations, their ability to apply the acquired knowl-
edge in their future medical practice, and their understanding of the ethical
and legal issues related to AI in medicine. The full list of questions can be
found in the supplementary materials, along with the data obtained from the
survey as well as a graphical approach to the results (Briganti & Pizzolla,
2023).
While we did not perform a formal reliability and validity study on the
survey, we took several steps to ensure that the survey captured the intended
information and provided valuable insights into the students’ experiences
and perceptions of the AI program. Prior to administering the survey, we
conducted pilot testing with a small group of students to identify potential
issues with question clarity, response options, and survey length. Based on
the feedback received, we revised the survey to ensure that the questions
were clear, concise, and accurately captured the intended information.
Statistical analysis
The statistical analysis was performed with the JASP open-source soft-
ware. To analyse the survey results, we performed a descriptive statistics
analysis to identify the core trends in the responses. Since the answers were
based on the Likert scale, we opted to use percentages to present the nd-
ings. This approach allowed us to eciently summarize the students’ per-
ceptions and opinions on various aspects of the AI program, and to iden-
tify areas of success and potential improvement. By using percentages, we
could eectively illustrate the distribution of responses and gain insights
into the overall satisfaction and perceived eectiveness of the AI program
among the participating medical students.
RESULTS
Our data presents survey results from 20 participants regarding their
views on AI’s impact on medicine and their experience with an AI program
in medical education. The main ndings are summarised below:
Impact of AI on future medicine. According to the survey, the major-
ity of respondents believe that AI will have a signicant impact on the eld
414 Pizzolla, Aro, Duez, De Lièvre, and Briganti
of medicine in the future. Specically, 55% of respondents think it is some-
what likely, and 45% think it is very likely.
Teaching AI in medical education. The survey also asked about the
teaching of AI concepts in medical education. The responses were mixed,
with 45% of respondents feeling that it was done somewhat well, while
30% felt that it was not done very well. Only 10% felt that it was done ex-
tremely well, and 15% were unsure.
Eectiveness and quality of the AI program. Respondents were also
asked about the eectiveness of the AI program. While some felt that it ex-
ceeded their expectations or was extremely well done (35%), others felt that
it fell short (20%). The largest group (35%) thought that the AI program met
their expectations. When asked to rate the overall quality of the AI program,
45% of respondents rated it as good, 40% rated it as fair, and 15% rated it as
poor.
Satisfaction with the AI program and instructor. Satisfaction with
the AI program was mixed. While 60% of respondents were somewhat sat-
ised with the topics covered, 25% were neutral, 10% were somewhat dis-
satised, and only 5% were very satised. In terms of instructor satisfac-
tion, 40% of respondents were very satised, 30% were somewhat satised,
25% were neutral, and 5% were somewhat dissatised.
Understanding the impact of AI in present medicine. The AI pro-
gram’s eectiveness in understanding AI’s role in medicine was rated ex-
tremely well by 20%, not sure by 10%, not very well by 5%, and somewhat
well by 65%.
Real-world examples. The survey also asked about the use of real-
world examples in the AI program. While 45% of respondents felt that the
program provided sucient examples, 30% said it provided only a few ex-
amples, and 20% were unsure. Only 5% felt that it was not very well done.
Demonstrating and understanding potential benets and limita-
tions of AI in medicine. When asked about the program’s eectiveness in
demonstrating the potential benets and limitations of AI in medicine, 35%
of respondents rated it extremely well, 45% thought it was somewhat well
done, 15% thought it was not very well done, and 5% said it was not at
all well done. The survey also asked about the eectiveness of the AI pro-
gram in helping respondents understand potential limitations or risks of AI
in medical diagnosis and treatment. 35% of respondents felt the AI program
was extremely eective, while 40% felt it provided a clear understanding.
However, 20% were somewhat uncertain, and 5% were unsure or felt it was
not very well done.
Condence in evaluating AI-generated diagnoses and treatment
recommendations. In terms of condence in evaluating AI-generated diag-
Integrating Articial Intelligence into Medical Education 415
noses and treatment recommendations, 60% felt somewhat condent, while
15% felt neutral and 25% felt somewhat uncondent.
Condence in applying AI program learnings to future medical
practice. In terms of condence in applying AI program learnings to future
medical practice, 40% felt somewhat condent, while 15% were extremely
condent. However, 20% were somewhat uncertain, and 5% were extreme-
ly uncertain.
Understanding ethical and legal issues related to AI in medicine. Fi-
nally, 50% felt that the AI program provided a somewhat clear understand-
ing of ethical and legal issues related to AI in medicine. 20% were neutral,
while 25% were unsure or felt it was not very well done.
DISCUSSION
In this discussion section, we aim to provide a comprehensive examina-
tion of the ndings from our study on the integration of AI in medical edu-
cation at the University of Mons. By assessing the opinions and feedback
from the medical students who participated in the AI program, we can better
understand the eectiveness of this educational approach, its implications
for the future of medical education, and potential areas for further improve-
ment. Additionally, we will address the potential limitations and challenges
associated with incorporating AI into medical education and highlight pos-
sible avenues for future research and development in this eld. Our survey
results provide valuable insights into the perceptions of medical students or
professionals regarding the impact of AI on the medical eld and their expe-
rience with an AI program in medical education.
Specic Survey Components
Impact of AI on medicine. All respondents believe that AI will have an
impact on the eld of medicine, with 55% considering it very likely, thus
supporting the fact that students are aware of the future implications of AI
in health-related practice (Briganti & Le Moine, 2020; Civaner et al., 2022).
This indicates a general consensus on the importance and potential of AI in
shaping the future of healthcare, thus supporting the necessity of improving
AI teaching to provide students with the ability to use and interpret such
technologies (Charow et al., 2021; Kang et al., 2017).
416 Pizzolla, Aro, Duez, De Lièvre, and Briganti
AI in medical education. There was a mixed opinion on how well med-
ical education teaches AI concepts. While a majority (55%) believed it was
taught at least somewhat well, there is still room for improvement as 30%
feel it is not taught very well. This supports the idea of a reboot of medical
curricula in order to better integrate AI concepts to prepare future healthcare
professionals for a rapidly changing landscape (Wartman & Combs, 2018).
AI program evaluation. There was a wide range of opinions on the
AI program’s eectiveness in meeting expectations and its overall quality.
However, more respondents rated the program positively (meeting or ex-
ceeding expectations) than negatively. This could imply that the AI program
generally meets the needs of the participants, but there might be some areas
where it could be improved, particularly in addressing AI legal and ethical
concerns. It is noteworthy that 25% of respondents expressed uncertainty or
dissatisfaction regarding the program’s ability to adequately address ethical
and legal issues related to AI in medicine. Furthermore, 20% of respondents
remained neutral on this question, suggesting a possible lack of AI literacy
in medical curricula and a consequent lack of knowledge regarding ethical
and legal aspects of AI. These ndings highlight the importance of enhanc-
ing the program’s focus on ethical and legal considerations to ensure a com-
prehensive understanding of these critical aspects within the context of AI
in medicine.
Satisfaction with program content and instructors. Most respondents
were at least somewhat satised with the topics covered and the instructors.
This suggests that the program’s curriculum and teaching sta are general-
ly well-received, but there is still potential for further enhancements. Thus,
the results show that if such programs are attributed to qualied AI educa-
tors, students are more likely to be satised with the instructor and the top-
ics covered. However, a huge limitation is represented by the shortage of
AI well-versed teaching personnel (Chan & Zary, 2019; Frommeyer et al.,
2022). As mentioned above, this could result in a delayed adoption of AI,
impeding its potential to oer benets to both the organization and patient
care.
AI program eectiveness. The AI program appears to be eective in
helping participants understand the role of AI in medicine, its benets, limi-
tations, and potential risks. However, the ethical and legal issues associated
with the use of AI in medicine require more attention as participants need
further assistance in comprehending these aspects. To ensure that the imple-
mentation of AI in medicine aligns with the fundamental principles of medi-
cal ethics and sustains the trust between healthcare professionals and their
patients, it is imperative that future programs insist on the ethical concerns
Integrating Articial Intelligence into Medical Education 417
associated with AI in healthcare (Char et al., 2018; Charow et al., 2021;
Kostkova, 2015) .
Condence in AI-related skills. While most respondents are somewhat
condent in their ability to evaluate AI-generated diagnoses and treatment
recommendations, as well as applying AI program learnings to their future
medical practice, there is still a considerable percentage of respondents with
neutral or negative opinions. This could indicate that the AI program might
need to focus more on practical aspects and skill-building to boost partici-
pants’ condence in these areas (Charow et al., 2021). One possible expla-
nation for the variation in student condence could be attributed to the di-
verse educational backgrounds of medical students in Belgium. Although an
entry examination exists for access to the medical curriculum, students may
have dierent levels of exposure to technical subjects, such as mathematics
and statistics. This disparity could lead to a range of views and opinions re-
garding hard sciences, including AI, even when applied to healthcare.
It is important to acknowledge that not all medical students may have a
strong initial interest in AI, which could contribute to the neutral or negative
opinions observed in our survey results. Nevertheless, understanding the
reasons behind this lack of condence is crucial for improving AI educa-
tion in medical programs. To address this issue, future studies could incor-
porate qualitative research methods, such as interviews or focus groups, to
gain a deeper understanding of the factors that inuence student condence
in their AI-related skills. This information could help identify areas for im-
provement in the curriculum, as well as tailor teaching methods to better
accommodate students with diverse educational backgrounds and interests.
Broader implications
By exploring these factors, our ndings can better inform other institu-
tions and curriculum developers as they consider integrating AI education
into their programs. This approach could help to address the needs of stu-
dents with dierent levels of prior knowledge and interest in AI, ultimately
enhancing their condence in using AI in their future medical practice.
The survey results indicate that the medical community generally rec-
ognizes the potential impact of AI on the eld of medicine and that AI pro-
grams in medical education are seen as valuable. However, there are areas
where improvements can be made, such as better integration of AI concepts
in medical education, enhancing the AI program’s content and delivery, and
focusing on practical skill-building to increase participants’ condence in
418 Pizzolla, Aro, Duez, De Lièvre, and Briganti
using AI in their future medical practice (Chan & Zary, 2019). To provide
practical implications for the areas of improvement identied, we propose
the following measures: rst, medical curricula should be designed to in-
terweave AI concepts throughout the various stages of learning, rather than
treating AI as an isolated topic. This could involve embedding AI-related
case studies in clinical and diagnostic courses or incorporating AI-driven
tools in medical simulations and practice settings, allowing students to de-
velop a more holistic understanding of AI’s role in medicine.
Second, to ensure that the AI program remains relevant and engag-
ing for students, the content should be regularly updated to reect the lat-
est advancements in the eld. Additionally, the program should employ di-
verse teaching methods, such as interactive workshops, online tutorials, and
hands-on training with AI-driven tools, to cater to dierent learning styles
and help students better grasp the material.
Third the AI program should place a greater emphasis on developing
students’ practical skills in applying AI to medical practice. This could in-
volve incorporating more real-world examples, case studies, and exercises
that require students to work with AI tools and systems directly. By pro-
viding students with opportunities to apply their AI knowledge in realistic
clinical scenarios, they can build the condence and competence necessary
to eectively utilize AI in their future medical practice.
Despite these challenges, our ndings underscore the importance of
incorporating AI into medical education. In the broader context of medical
education and digital medicine, the signicance of our ndings lies in the
potential to inform other institutions and curriculum developers as they con-
sider integrating AI education into their programs. The collected results pro-
vide valuable insights into the aspects of AI education that are most eec-
tive in helping students grasp the complexities of AI applications in health-
care, as well as the potential areas for improvement. Our ndings indicate
that the AI program at the University of Mons successfully provided medi-
cal students with a foundational knowledge of AI and its use in the medical
eld. The positive feedback from students in our survey suggests that the
integration of AI education in the medical curriculum is both timely and rel-
evant.
The results also suggest several avenues for future research and devel-
opment. One such avenue could be the investigation of optimal teaching
methods and strategies for AI education, ensuring that students are well-
equipped to navigate the complexities of AI applications in healthcare. Fur-
thermore, research could focus on identifying the specic AI competencies
that are most relevant to medical professionals, allowing for the develop-
Integrating Articial Intelligence into Medical Education 419
ment of targeted and eective curricula (Chan & Zary, 2019; Char et al.,
2018; Charow et al., 2021).
In conclusion, our study highlights the importance of integrating AI
into medical education to prepare future healthcare professionals for the
increasingly data-driven and technologically advanced landscape of health-
care. By addressing potential limitations and challenges and exploring av-
enues for future research, we can continue to rene AI education and ensure
that medical students are well-equipped to make the most of AI’s potential
in improving patient care and outcomes.
LIMITATIONS
The results of our study should be met with a number of limitations. We
outline three of them. First, because of the low sample size and the limited
outreach of the program (only one class, in one university, in Belgium), the
percentages reected in our results section on the student’s appreciation of
the AI course might not replicate with a similar program in another univer-
sity. Future studies may endeavor to reproduce our eorts to look for im-
proved outcomes.
Second, although we conducted a large review of the literature, we did
not perform a systematic review, as it was beyond the scope of this study.
Future studies may endeavor to perform a systematic review of active AI
programs in healthcare curricula. Third, we did not perform any statistical
inference on our sample, which largely limits the information we can gath-
er from the questionnaire: this was chosen beforehand because of the low
sample size, which would limit the generalizability of the inferential results.
Future studies may endeavor to perform more in-depth questionnaire sub-
mitted to a larger sample of students, as to be able to perform classical sta-
tistical tests and infer more on the educational processes accompanying the
learning of AI in healthcare curricula.
CONCLUSION
In conclusion, the medical practice has undergone signicant changes
due to the advancements of innovative technologies and AI in recent years.
However, the current landscape of medical education does not seem to ade-
quately prepare medical students for the potential of AI in healthcare. To ad-
dress this issue, the University of Mons (UMONS) has implemented a new
420 Pizzolla, Aro, Duez, De Lièvre, and Briganti
academic program designed to educate medical students about AI in health-
care. While the AI program developed at UMONS has shown great potential
in assisting medical students in their training, it has some limitations that
need to be considered. These limitations suggest that further research and
development are needed to make the program more eective and useful in a
broader context.
REFERENCES
Abdelsamea, M. M., Zidan, U., Senousy, Z., Gaber, M. M., Rakha, E., & Ilyas,
M. (2022). A survey on articial intelligence in histopathology image anal-
ysis. WIREs Data Mining and Knowledge Discovery, 12(6), e1474. https://
doi.org/10.1002/widm.1474
Acs, B., & Hartman, J. (2020). Next generation pathology: Articial intelligence
enhances histopathology practice. The Journal of Pathology, 250(1), 7–8.
https://doi.org/10.1002/path.5343
Attallah, O. (2021). MB-AI-His: Histopathological Diagnosis of Pediatric Me-
dulloblastoma and its Subtypes via AI. Diagnostics, 11(2), Article 2. https://
doi.org/10.3390/diagnostics11020359
Ayyad, S. M., Shehata, M., Shalaby, A., Abou El-Ghar, M., Ghazal, M., El-Mel-
egy, M., Abdel-Hamid, N. B., Labib, L. M., Ali, H. A., & El-Baz, A. (2021).
Role of AI and histopathological images in detecting prostate cancer: A sur-
vey. Sensors, 21(8), Article 8. https://doi.org/10.3390/s21082586
Beck, J.P. (2020, June 29). AI: Giving healthcare professionals ‘The Gift of
Time’. The Parliament Magazine. https://www.theparliamentmagazine.eu/
news/article/ai-giving-healthcare-professionals-the-gift-of-time
Brady, A. P., & Neri, E. (2020). Articial intelligence in radiology—Ethical con-
siderations. Diagnostics, 10(4), Article 4. https://doi.org/10.3390/diagnos-
tics10040231
Briganti, G. (2023). Intelligence articielle: Une introduction pour les clin-
iciens. Revue des Maladies Respiratoires. https://doi.org/10.1016/j.
rmr.2023.02.005
Briganti, G., & Le Moine, O. (2020). Articial intelligence in medicine: To-
day and tomorrow. Frontiers in Medicine, 7, 27. https://doi.org/10.3389/
fmed.2020.00027
Briganti, G., & Pizzolla, I. (2023). Data from: ‘Integrating Articial Intelligence
into Medical Education: Lessons Learned from a Belgian Initiative’. https://
doi.org/10.17605/OSF.IO/5YCJW
Byng, D., Strauch, B., Gnas, L., Leibig, C., Stephan, O., Bunk, S., & Hecht, G.
(2022). AI-based prevention of interval cancers in a national mammography
screening program. European Journal of Radiology, 152, 110321. https://
doi.org/10.1016/j.ejrad.2022.110321
Integrating Articial Intelligence into Medical Education 421
Castelvecchi, D. (2016). Can we open the black box of AI? Nature News,
538(7623), 20. https://doi.org/10.1038/538020a
Chan, K. S., & Zary, N. (2019). Applications and challenges of implementing ar-
ticial intelligence in medical education: Integrative review. JMIR Medical
Education, 5(1), e13930. https://doi.org/10.2196/13930
Char, D. S., Shah, N. H., & Magnus, D. (2018). Implementing machine learning
in health care—Addressing ethical challenges. The New England Journal of
Medicine, 378(11), 981. https://doi.org/10.1056/NEJMp1714229
Charow, R., Jeyakumar, T., Younus, S., Dolatabadi, E., Salhia, M., Al-Mouas-
was, D., Anderson, M., Balakumar, S., Clare, M., Dhalla, A., Gillan, C.,
Haghzare, S., Jackson, E., Lalani, N., Mattson, J., Peteanu, W., Tripp, T.,
Waldorf, J., Williams, S., … Wiljer, D. (2021). Articial intelligence educa-
tion programs for health care professionals: Scoping review. JMIR Medical
Education, 7(4), e31043. https://doi.org/10.2196/31043
Chen, J. H., & Asch, S. M. (2017). Machine learning and prediction in medi-
cine—Beyond the peak of inated expectations. The New England Journal
of Medicine, 376(26), 2507–2509. https://doi.org/10.1056/NEJMp1702071
Civaner, M. M., Uncu, Y., Bulut, F., Chalil, E. G., & Tatli, A. (2022). Articial
intelligence in medical education: A cross-sectional needs assessment. BMC
Medical Education, 22(1), 772. https://doi.org/10.1186/s12909-022-03852-3
Codari, M., Melazzini, L., Morozov, S. P., van Kuijk, C. C., Sconenza, L. M.,
Sardanelli, F., & European Society of Radiology (ESR). (2019). Impact of
articial intelligence on radiology: A EuroAIM survey among members
of the European Society of Radiology. Insights into Imaging, 10(1), 105.
https://doi.org/10.1186/s13244-019-0798-3
Coppola, F., Faggioni, L., Gabelloni, M., De Vietro, F., Mendola, V., Cattabriga,
A., Cocozza, M. A., Vara, G., Piccinino, A., Lo Monaco, S., Pastore, L. V.,
Mottola, M., Malavasi, S., Bevilacqua, A., Neri, E., & Goleri, R. (2021).
Human, all too human? An all-around appraisal of the “articial intelli-
gence revolution” in medical imaging. Frontiers in Psychology, 12. https://
www.frontiersin.org/articles/10.3389/fpsyg.2021.710982
Cox, M., Blouin, A. S., Cu, P., Paniagua, M., Phillips, S., & Vlasses, P. H.
(2017). The role of accreditation in achieving the quadruple aim. NAM Per-
spectives. https://doi.org/10.31478/201710a
Densen, P. (2011). Challenges and opportunities facing medical education.
Transactions of the American Clinical and Climatological Association, 122,
48–58.
Dlamini, Z., Skepu, A., Kim, N., Mkhabele, M., Khanyile, R., Mole, T.,
Mbatha, S., Setlai, B., Mulaudzi, T., Mabongo, M., Bida, M., Kgoebane-
Maseko, M., Mathabe, K., Lockhat, Z., Kgokolo, M., Chauke-Malinga, N.,
Ramagaga, S., & Hull, R. (2022). AI and precision oncology in clinical can-
cer genomics: From prevention to targeted cancer therapies-an outcomes
based patient care. Informatics in Medicine Unlocked, 31, 100965. https://
doi.org/10.1016/j.imu.2022.100965
422 Pizzolla, Aro, Duez, De Lièvre, and Briganti
Erickson, B. J., Koratis, P., Akkus, Z., & Kline, T. L. (2017a). Machine learn-
ing for medical imaging. RadioGraphics, 37(2), 505–515. https://doi.
org/10.1148/rg.2017160130
Foster, M., & Tasnim, Z. (2020). Data science and graduate nursing education:
A critical literature review. Clinical Nurse Specialist, 34(3), 124. https://doi.
org/10.1097/NUR.0000000000000516
Frommeyer, T. C., Fursmidt, R. M., Gilbert, M. M., & Bett, E. S. (2022). The
desire of medical students to integrate articial intelligence into medical
education: An opinion article. Frontiers in Digital Health, 4. https://www.
frontiersin.org/articles/10.3389/fdgth.2022.831123
He, J., Baxter, S. L., Xu, J., Xu, J., Zhou, X., & Zhang, K. (2019). The practical
implementation of articial intelligence technologies in medicine. Nature
Medicine, 25(1), 30–36. https://doi.org/10.1038/s41591-018-0307-0
Hegde, S., Ajila, V., Zhu, W., & Zeng, C. (2022). Articial intelligence in early
diagnosis and prevention of oral cancer. Asia-Pacic Journal of Oncology
Nursing, 9(12), 100133. https://doi.org/10.1016/j.apjon.2022.100133
Holzinger, A., Langs, G., Denk, H., Zatloukal, K., & Müller, H. (2019).
Causability and explainability of articial intelligence in medicine.
WIREs Data Mining and Knowledge Discovery, 9(4), e1312. https://doi.
org/10.1002/widm.1312
Hu, R., Fan, K. Y., Pandey, P., Hu, Z., Yau, O., Teng, M., Wang, P., Li, A.,
Ashraf, M., & Singla, R. (2022). Insights from teaching articial intelli-
gence to medical students in Canada. Communications Medicine, 2(1), Ar-
ticle 1. https://doi.org/10.1038/s43856-022-00125-4
Jin, D., Harrison, A. P., Zhang, L., Yan, K., Wang, Y., Cai, J., Miao, S., & Lu, L.
(2021). Chapter 14—Articial intelligence in radiology. In L. Xing, M. L.
Giger, & J. K. Min (Eds.), Articial Intelligence in Medicine (pp. 265–289).
Academic Press. https://doi.org/10.1016/B978-0-12-821259-2.00014-4
Kang, S. K., Lee, C. I., Pandharipande, P. V., Sanelli, P. C., & Recht, M. P.
(2017). Residents’ introduction to comparative eectiveness research and
big data analytics. Journal of the American College of Radiology, 14(4),
534–536. https://doi.org/10.1016/j.jacr.2016.10.032
Kayser, K., GĂśrtler, Jă., Bogovac, M., Bogovac, A., Goldmann, T., Vollmer, E.,
& Kayser, G. (2009). AI (articial intelligence) in histopathology—From
image analysis to automated diagnosis. Folia Histochemica et Cytobiolog-
ica, 47(3), Article 3. https://doi.org/10.2478/v10042-009-0087-y
Kostkova, P. (2015). Grand challenges in digital health. Frontiers in Public
Health, 3. https://www.frontiersin.org/articles/10.3389/fpubh.2015.00134
Lillehaug, S.-I., & Lajoie, S. P. (1998). AI in medical education—Another grand
challenge for medical informatics. Articial Intelligence in Medicine, 12(3),
197–225. https://doi.org/10.1016/S0933-3657(97)00054-7
Long, D., & Magerko, B. (2020). What is AI literacy? Competencies and design
considerations. Proceedings of the 2020 CHI Conference on Human Fac-
tors in Computing Systems, 1–16. https://doi.org/10.1145/3313831.3376727
Integrating Articial Intelligence into Medical Education 423
Lui, T. K. L., Guo, C.-G., & Leung, W. K. (2020). Accuracy of articial intelli-
gence on histology prediction and detection of colorectal polyps: A system-
atic review and meta-analysis. Gastrointestinal Endoscopy, 92(1), 11-22.e6.
https://doi.org/10.1016/j.gie.2020.02.033
Makridakis, S. (2017). The forthcoming Articial Intelligence (AI) revolu-
tion: Its impact on society and rms. Futures, 90, 46–60. https://doi.
org/10.1016/j.futures.2017.03.006
Masters, K. (2019). Articial intelligence in medical education. Medical Teach-
er, 41(9), 976–980. https://doi.org/10.1080/0142159X.2019.1595557
Nicholson Price II, W. (2014). Black-Box Medicine (SSRN Scholarly Paper No.
2499885). https://papers.ssrn.com/abstract=2499885
Obermeyer, Z., & Emanuel, E. J. (2016). Predicting the future—Big data, ma-
chine learning, and clinical medicine. New England Journal of Medicine,
375(13), 1216–1219. https://doi.org/10.1056/NEJMp1606181
Obermeyer, Z., Powers, B., Vogeli, C., & Mullainathan, S. (2019). Dissecting ra-
cial bias in an algorithm used to manage the health of populations. Science,
366(6464), 447–453. https://doi.org/10.1126/science.aax2342
Paranjape, K., Schinkel, M., Nannan Panday, R., Car, J., & Nanayakkara, P.
(2019). Introducing articial intelligence training in medical education.
JMIR Medical Education, 5(2), e16048. https://doi.org/10.2196/16048
Park, S. H., Kim, Y.-H., Lee, J. Y., Yoo, S., & Kim, C. J. (2019). Ethical chal-
lenges regarding articial intelligence in medicine from the perspective of
scientic editing and peer review. Science Editing, 6(2), 91–98. https://doi.
org/10.6087/kcse.164
Pianykh, O. S., Langs, G., Dewey, M., Enzmann, D. R., Herold, C. J., Schoen-
berg, S. O., & Brink, J. A. (2020). Continuous learning AI in radiology:
Implementation principles and early applications. Radiology, 297(1), 6–14.
https://doi.org/10.1148/radiol.2020200038
Reyes, M., Meier, R., Pereira, S., Silva, C. A., Dahlweid, F.-M., Tengg-Ko-
bligk, H. von, Summers, R. M., & Wiest, R. (2020). On the interpretabil-
ity of articial intelligence in radiology: Challenges and opportunities.
Radiology: Articial Intelligence, 2(3), e190043. https://doi.org/10.1148/
ryai.2020190043
Robeznieks, A. (2018, November 20). 3 ways medical AI can improve workow
for physicians. American Medical Association. https://www.ama-assn.org/
practice-management/digital/3-ways-medical-ai-can-improve-workflow-
physicians
Saldanha, O. L., Quirke, P., West, N. P., James, J. A., Loughrey, M. B., Grabsch,
H. I., Salto-Tellez, M., Alwers, E., Cifci, D., Ghaari Laleh, N., Seibel,
T., Gray, R., Hutchins, G. G. A., Brenner, H., van Treeck, M., Yuan, T.,
Brinker, T. J., Chang-Claude, J., Khader, F., … Kather, J. N. (2022). Swarm
learning for decentralized articial intelligence in cancer histopathology.
Nature Medicine, 28(6), Article 6. https://doi.org/10.1038/s41591-022-
01768-5
424 Pizzolla, Aro, Duez, De Lièvre, and Briganti
Sapci, A. H., & Sapci, H. A. (2020). Articial intelligence education and tools
for medical and health informatics students: Systematic Review. JMIR
Medical Education, 6(1), e19285. https://doi.org/10.2196/19285
Shmatko, A., Ghaari Laleh, N., Gerstung, M., & Kather, J. N. (2022). Articial
intelligence in histopathology: Enhancing cancer research and clinical on-
cology. Nature Cancer, 3(9), Article 9. https://doi.org/10.1038/s43018-022-
00436-4
Singh, R. P., Hom, G. L., Abramo, M. D., Campbell, J. P., Chiang, M. F., &
on behalf of the AAO Task Force on Articial Intelligence. (2020). Current
challenges and barriers to real-world articial intelligence adoption for the
healthcare system, provider, and the patient. Translational Vision Science &
Technology, 9(2), 45. https://doi.org/10.1167/tvst.9.2.45
Srivastava, T., & Waghmare, L. (2020). Implications of articial intelligence
(AI) on dynamics of medical education and care: A perspective. Jour-
nal of Clinical and Diagnostic Research, 14, 1–2. https://doi.org/10.7860/
JCDR/2020/43293.13565
Tajmir, S. H., & Alkasab, T. K. (2018). Toward augmented radiologists: Changes
in radiology education in the era of machine learning and articial intel-
ligence. Academic Radiology, 25(6), 747–750. https://doi.org/10.1016/j.
acra.2018.03.007
Topol, E. J. (2019). High-performance medicine: The convergence of hu-
man and articial intelligence. Nature Medicine, 25(1), 44–56. https://doi.
org/10.1038/s41591-018-0300-7
Wartman, S. A., & Combs, C. D. (2018). Medical education must move from
the information age to the age of articial intelligence. Academic Medicine,
93(8), 1107. https://doi.org/10.1097/ACM.0000000000002044
