A Systematic Review and Meta-Analysis of Artificial Intelligence Tools in Medicine and Healthcare: Applications, Considerations, Limitations, Motivation and Challenges

Artificial intelligence (AI) has emerged as a transformative force in various sectors, including medicine and healthcare. Large language models like ChatGPT showcase AI’s potential by generating human-like text through prompts. ChatGPT’s adaptability holds promise for reshaping medical practices, improving patient care, and enhancing interactions among healthcare professionals, patients, and data. In pandemic management, ChatGPT rapidly disseminates vital information. It serves as a virtual assistant in surgical consultations, aids dental practices, simplifies medical education, and aids in disease diagnosis. A total of 82 papers were categorised into eight major areas, which are G1: treatment and medicine, G2: buildings and equipment, G3: parts of the human body and areas of the disease, G4: patients, G5: citizens, G6: cellular imaging, radiology, pulse and medical images, G7: doctors and nurses, and G8: tools, devices and administration. Balancing AI’s role with human judgment remains a challenge. A systematic literature review using the PRISMA approach explored AI’s transformative potential in healthcare, highlighting ChatGPT’s versatile applications, limitations, motivation, and challenges. In conclusion, ChatGPT’s diverse medical applications demonstrate its potential for innovation, serving as a valuable resource for students, academics, and researchers in healthcare. Additionally, this study serves as a guide, assisting students, academics, and researchers in the field of medicine and healthcare alike.


Introduction
Artificial intelligence (AI) has emerged as a powerful tool with transformative potential across various sectors, and the field of medicine and healthcare is no exception.One remarkable application of AI in this realm is the development of large language models, such as ChatGPT, which have gained significant attention for their ability to generate human-like text based on prompts.ChatGPT's versatile capabilities hold promise for reshaping medical practices, enhancing patient care, and revolutionising the way healthcare professionals interact with both patients and data .
The innovative aspects of ChatGPT's deployment in medicine include a variety of unique applications, such as assisting in the identification of rare diseases and providing support for men's health, demonstrating the model's versatility.Advanced methodologies are at the heart of its training and effectiveness evaluation, which are tailored specifically for healthcare contexts to ensure relevance and dependability.The work is fundamentally interdisciplinary, as it merges AI with expert medical insights in order to effectively tackle significant healthcare challenges.The aim of this synergy is to translate into tangible benefits, which include enhanced patient outcomes and streamlined healthcare processes.This, in turn, marks a substantial real-world impact on the medical field.The areas of ChatGPT's application in medical fields and healthcare are as follows.

Aiding Cosmetic Orthognathic Surgery Consultations
Within the realm of surgical procedures, such as cosmetic orthognathic surgery, Chat-GPT can function as a virtual assistant, offering patients crucial preoperative information.Potential candidates who are considering cosmetic orthognathic surgery may be seeking information regarding the procedure, recovery, risks, and benefits.ChatGPT can offer standardised and accurate responses, which prepare patients for their consultations and help them make informed decisions [5][6][7][8][9][10]42,43].

Support in Disease Diagnosis
The potential of ChatGPT as a diagnostic tool holds promise in the early detection of diseases.ChatGPT can generate potential differential diagnoses by analysing patientreported symptoms and medical history, thereby aiding healthcare providers in narrowing down diagnostic possibilities.However, caution must be exercised as diagnosis requires domain-specific expertise [24].

Cellular Imaging, Revolutionizing Radiology, and Sonar Imaging
Cellular imaging pertains to the utilisation of diverse techniques and technologies for visualising and studying cells at the microscopic level.Scientists and researchers are allowed to examine the structure, function, and behaviour of individual cells or cell populations.Cellular imaging techniques comprise light microscopy, fluorescence microscopy, confocal microscopy, electron microscopy, and various other advanced imaging methods.These techniques are widely used in various fields, such as biology, medicine, and biotechnology, for the purpose of better understanding cellular processes, cell interactions, and disease mechanisms.
The interpretation of medical images, such as radiology and sonar scans, requires precision and accuracy.The potential of ChatGPT lies in its ability to analyse and generate descriptions for medical images, which can enhance the workflow of radiologists.It has the ability to provide initial observations, which can highlight regions of interest and assist radiologists in their analyses [13,[25][26][27][28][29][30][31][32][33].

Pharmaceutical Research and Treatment
In the realm of pharmaceutical research, ChatGPT can contribute by sifting through vast volumes of scientific literature, identifying potential drug candidates, and suggesting innovative research directions.It can assist in drug discovery, research proposal writing, and summarising complex medical research, expediting the research process [11,24,[34][35][36][37][38].

Navigating Limitations and Ethical Considerations
While ChatGPT offers numerous opportunities, it also poses certain limitations.The responses are based on the data on which it was trained, potentially resulting in biased or inaccurate information.Ethical concerns arise when content generated by AI is mistaken for the expertise of a human.The challenge of balancing the role of AI with the need for human judgement and expertise remains [18,19,27,31,35,[38][39][40][41], as depicted in Figure 1.

Cellular Imaging, Revolutionizing Radiology, and Sonar Imaging
Cellular imaging pertains to the utilisation of diverse techniques and technologie visualising and studying cells at the microscopic level.Scientists and researchers ar lowed to examine the structure, function, and behaviour of individual cells or cell po lations.Cellular imaging techniques comprise light microscopy, fluorescence microsc confocal microscopy, electron microscopy, and various other advanced imaging meth These techniques are widely used in various fields, such as biology, medicine, and technology, for the purpose of better understanding cellular processes, cell interacti and disease mechanisms.
The interpretation of medical images, such as radiology and sonar scans, requ precision and accuracy.The potential of ChatGPT lies in its ability to analyse and gene descriptions for medical images, which can enhance the workflow of radiologists.It the ability to provide initial observations, which can highlight regions of interest and sist radiologists in their analyses [13,[25][26][27][28][29][30][31][32][33].

Pharmaceutical Research and Treatment
In the realm of pharmaceutical research, ChatGPT can contribute by sifting thro vast volumes of scientific literature, identifying potential drug candidates, and sugges innovative research directions.It can assist in drug discovery, research proposal wri and summarising complex medical research, expediting the research process [11,24 38].

Navigating Limitations and Ethical Considerations
While ChatGPT offers numerous opportunities, it also poses certain limitations.responses are based on the data on which it was trained, potentially resulting in biase inaccurate information.Ethical concerns arise when content generated by AI is mistaken the expertise of a human.The challenge of balancing the role of AI with the need for hu judgement and expertise remains [18,19,27,31,35,[38][39][40][41], as depicted in Figure 1.Categories and groups covered by all studies of the review.G1: treatment and medi G2: buildings and equipment, G3: parts of the human body and areas of the disease, G4: pati G5: citizens, G6: cellular imaging, radiology, pulse, and medical images, G7: doctors and nu and G8: tools, devices and administration.
Johnson, D et al. [5] utilised the accuracy and completeness of ChatGPT in answe medical queries by academic physician specialists.The results generally show accu information with limitations, which suggests a need for further research and model velopment.Mohammad H et al. [13] conducted a hybrid panel discussion that focuse the integration of ChatGPT, a large language model, in the fields of education, resea and healthcare.The event gathered responses from attendees, both in-person and on Categories and groups covered by all studies of the review.G1: treatment and medicine, G2: buildings and equipment, G3: parts of the human body and areas of the disease, G4: patients, G5: citizens, G6: cellular imaging, radiology, pulse, and medical images, G7: doctors and nurses, and G8: tools, devices and administration.
Johnson, D et al. [5] utilised the accuracy and completeness of ChatGPT in answering medical queries by academic physician specialists.The results generally show accurate information with limitations, which suggests a need for further research and model development.Mohammad H et al. [13] conducted a hybrid panel discussion that focused on the integration of ChatGPT, a large language model, in the fields of education, research, and healthcare.The event gathered responses from attendees, both in-person and online, by utilising an audience interaction platform.According to the study, approximately 40% of the participants had utilised ChatGPT, with a higher number of trainees compared to faculty members having experimented with it.Those individuals who had utilised ChatGPT demonstrated a heightened level of interest in its potential application across a multitude of contexts.Uncertainty was observed regarding its use in education, with pros and cons being discussed for its integration in education, research, and healthcare.The perspectives varied according to role (trainee, faculty, and staff), highlighting the need for further discussion and exploration of the implications and optimal uses of ChatGPT in these sectors.The study emphasises the significance of taking a deliberate and measured approach to adoption in order to mitigate potential risks and challenges.Sallam et al. [44] assessed the advantages and disadvantages of ChatGPT in healthcare education.While it is beneficial for learning, it lacks emotional interaction and poses plagiarism risks.
In conclusion, the integration of ChatGPT into medicine and healthcare holds significant potential to transform various aspects of patient care, education, diagnosis, and research.As technology continues its advancement, it becomes crucial to responsibly harness the capabilities of ChatGPT, while also addressing its limitations and ethical considerations.The exploration of ChatGPT's applications across diverse medical domains emphasises its role as a catalyst for innovation and improvement in the medical landscape.
The main objectives of the study can be succinctly summarised as follows: • To investigate the various applications of ChatGPT in healthcare, such as pandemic management, surgical consultations, dental practices, medical education, disease diagnosis, cellular imaging, sonar imaging, radiology, and pharmaceutical research.• To investigate the potential benefits and risks of integrating ChatGPT in healthcare, assessing its impact on patient care, medical processes, and ethical considerations.• To investigate the ethical issues and the need to balance AI's role in healthcare settings with human judgement.• To offer perspectives and recommendations for the responsible adoption and use of ChatGPT in medicine in general, and cellular imaging in particular, while addressing limitations and ethical concerns.

Materials and Methods
The field of this study is covered by very important query words (keywords), namely, ChatGPT with "Medicine", "Healthcare", "Tackles", "Pandemics/Infectious", "A Cosmetic orthognathic surgery Consultation", "dental Medical/Education", "Disease, Radiology", "sonar", "pharmaceutical", "Treatment", "Diagnosis" and "patient care", "Medical school", "Ophthalmology", "Digital Health", and "operating room".Our study is closed to Englishlanguage studies only.The following digital databases and publishers were selected to search for target papers, and the numbers of articles included the following form: Taylor and Francis (31) 2.
The studies were chosen by conducting literature searches, which were then followed by three rounds of screening and filtering.In the initial iteration, using Mendeley software (1.19.4-win32),only publications that were published in the last eight months were gathered after eliminating duplicate articles.

•
The article should be original or reviewed, written in English, and published in English journals, conferences, case studies, editorials, opinion articles, and letters to the editor.

•
The study specifies that it contains the specified keywords.

•
Any relevant articles are published or in press between November 2022 and August 2023.

•
The main point is on Chat GPT with specific words in the field of healthcare and health.

•
The study includes the study's objectives, design, applications, benefits, risks, concerns, limitations, study field/area, conclusions, and recommended actions.
The focus is on the pros and cons, challenges, as well as all the inclusion criteria shown in Figure 4.

•
The article should be original or reviewed, written in English, and published in English journals, conferences, case studies, editorials, opinion articles, and letters to the editor.

•
The study specifies that it contains the specified keywords.

•
Any relevant articles are published or in press between November 2022 and August 2023.

•
The main point is on Chat GPT with specific words in the field of healthcare and health.

•
The study includes the study's objectives, design, applications, benefits, risks, concerns, limitations, study field/area, conclusions, and recommended actions.
The focus is on the pros and cons, challenges, as well as all the inclusion criteria shown in Figure 4.

Inclusion Criteria
• The article should be original or reviewed, written in English, and published in English journals, conferences, case studies, editorials, opinion articles, and letters to the editor.• The study specifies that it contains the specified keywords.

•
Any relevant articles are published or in press between November 2022 and August 2023.
• The main point is on Chat GPT with specific words in the field of healthcare and health.
• The study includes the study's objectives, design, applications, benefits, risks, concerns, limitations, study field/area, conclusions, and recommended actions.
The focus is on the pros and cons, challenges, as well as all the inclusion criteria shown in Figure 4.

Data Gathering Procedure
All included articles were reviewed, examined, and summarised in accordance with their basic classifications, then saved as Microsoft Word and Excel files to simplify the filtering process.For every piece, the authors read the complete text.We were able to develop the proposed taxonomy using a variety of highlights and comments on the surveyed works, as well as a running classification of all the articles.The remarks were recorded in either paper copy or electronic form, depending on the writing style of each contributor.After this, another procedure was conducted to characterise, describe, tabulate, and draw conclusions about the key findings.

Research Questions
The selection of research questions is a critical step in shaping the purpose of the study and the anticipated results.Consequently, we have formulated the following research questions to align with the primary objective of our systematic literature review:

Data Gathering Procedure
All included articles were reviewed, examined, and summarised in accordance with their basic classifications, then saved as Microsoft Word and Excel files to simplify the filtering process.For every piece, the authors read the complete text.We were able to develop the proposed taxonomy using a variety of highlights and comments on the surveyed works, as well as a running classification of all the articles.The remarks were recorded in either paper copy or electronic form, depending on the writing style of each contributor.After this, another procedure was conducted to characterise, describe, tabulate, and draw conclusions about the key findings.

Research Questions
The selection of research questions is a critical step in shaping the purpose of the study and the anticipated results.Consequently, we have formulated the following research questions to align with the primary objective of our systematic literature review:

Results
The results of the initial query search, which yielded 1273 articles, are as follows: there are 31 articles from Taylor and Francis, 179 articles from Google Scholar, 412 articles from Scopus, 239 articles from Web of Science, 127 articles from Elsevier, 218 articles from Springer, 41 articles from MDPI, 18 articles from IEEE Xplore digital, and 7 articles from Wiley from between November 2022 and August 2023.
The papers were filtered according to the sequence that was adopted in this research and were divided into two categories: 1273 articles were published in the last eight months (November 2022 to August 2023) and 392 papers appeared in all nine databases or publishers, resulting in a total of 1547 papers.Following a comprehensive scan of the titles and abstracts of the papers, an additional 1273 papers were excluded.After the final full-text reading, 92 papers were excluded.The final set consisted of 82 papers, which were divided into eight major categories/groups as follows: G1: treatment and medicine are essential; G2: buildings and equipment play a crucial role; G3: parts of the human body and areas of disease; G4: patients; G5: citizens; G6: the focus is on radiology, pulse, and medical images; G7: revolves around doctors and nurses; and G8: encompasses tools, devices, and administration.
The first category (G1), which comprised 27 articles (32.93%), was focused on treatment and medicine.The second category (G2) consisted of two articles (2.44%) that were focused on buildings and equipment.The third category, known as G3, comprised 14 articles, accounting for 17.07% of the total.These articles focused on parts of the human body and areas affected by disease.The fourth category (G4), which comprised eight articles (9.76%), focused on patients.The fifth category (G5), which comprised three articles (3.66%), was citizens.The sixth category (G6) comprised four articles (4.88%) on cellular imaging, radiology, pulse, and medical images.The seventh category, known as G7, consisted of 17 articles, accounting for 20.73% of the total.This category focused on doctors and nurses.The eighth category, known as G8, consisted of seven articles, accounting for 8.54% of the total.This category focused on tools, devices, and administration.ChatGPT plays a crucial role in pandemic management by aiding in the swift and accurate dissemination of critical information.Its natural language processing capabilities enable it to generate coherent responses, making it a valuable tool for healthcare organisations.Specific advantages include its ability to provide rapid updates, preventive measures, and medical guidelines to the public.ChatGPT assists in addressing queries and concerns, ensuring accurate information flow and timely interventions during health crises [5][6][7][8][9]42,43].

RQ2: How Is ChatGPT Utilised in the Field of Dental Practices, and How Does It Enhance the Overall Patient Experience in This Context?
ChatGPT is integrated into dental practices in order to empower dental assistants and improve the overall patient experience.It assists in the handling of patient inquiries, the provision of oral health tips, and the guidance of patients through postoperative care instructions.ChatGPT ensures standardised and accurate responses, thereby assisting patients in making informed decisions regarding their dental care.This AI integration revolutionises patient communication, appointment scheduling, and postoperative support in dental practices [8,11,12].

RQ3: What Challenges and Ethical Considerations Are Associated with the Integration of ChatGPT into Medical Practices and Healthcare Settings?
The integration of ChatGPT into medical practices and healthcare settings presents challenges and ethical considerations.One challenge is that the responses of ChatGPT are based on the data on which it was trained, which may result in biased or inaccurate information.Ethical concerns arise when content generated by AI is mistaken for the expertise of a human.The challenge lies in striking a balance between the capabilities of AI and the necessity for human judgement and expertise.Ensuring that AI does not replace human healthcare professionals entirely while leveraging its advantages is a significant ethical consideration [18,19,27,31,35,[38][39][40][41].Careful thought and responsible adoption are essential to mitigate potential risks and challenges.The findings of our studies are summarised in Table 1, which includes information about the purpose of the study (in the "aim of study" column), the design and application (in the "design, application(s)" column), the benefits and risks (in the "benefit(s), risks" column), the summary of results for concerns and limitations (in the "concern(s), limitation(s)" column), the main outcomes, the type of study (in the "study field/area" column), and the main outcomes reported by each article in the conclusion section (in the "Suggested Action" column).Table 2 summarises the pros and cons or challenges of some of the papers that were exposed.

•
Privacy, ethics, bias concerns; ChatGPT transforms medicine with potential for innovation.

•
Review of limitations to other studies.
ChatGPT's medical potential is promising but faces challenges.Ethical, safety considerations vital for transforming healthcare with AI.
ChatGPT's applications: text generation, literature analysis, image learning in education, healthcare, and research.

•
Ongoing ethical exploration is needed.

•
Risk of the digital divide widening.
Responsible AI use demands transparency, equity, reliability, and non-harm principles.
Enhances personalised learning, clinical reasoning, and skills development.
• Benefits include interactive learning, but drawbacks like bias, data privacy persist.
ChatGPT's integration in healthcare education offers potential advantages.Further research required to address ethical, bias, and accuracy concerns.
ChatGPT as medical information source with substantial accuracy and potential.

•
ChatGPT's medical accuracy is promising but not entirely reliable.

•
Higher accuracy scores often contrasted with surprising errors.

•
Limited sample size; potential selection and respondent biases.

•
Patient-generated queries, complex cases not well-represented.
ChatGPT shows potential as medical resource but requires validation and improvement.

[15]
Evaluate ChatGPT's performance in medical education, content generation, and deception.Medical education, skill practice, patient interaction simulation, content generation.

•
ChatGPT aids learning but raises ethical concerns.

•
Can deceive professionals and educators.
ChatGPT's potential in education is transformative yet raises ethical concerns.
Improved writing, efficient research, personalised learning, streamlined practice.

•
Promising applications but ethical.• Paraphrased content showed significant matches using plagiarism detection tool.

•
Raising concerns about originality.
ChatGPT generated Digital Twin in healthcare articles.Low plagiarism in author-written text, high similarity in abstract paraphrases.
AI accelerates knowledge expression, academic validity monitored through citations.
ChatGPT has versatile healthcare applications but faces ethical, privacy, and accuracy concerns.
23 [60] Examine AI impact on medical publishing ethics and guidelines.
ChatGPT shows potential for medical education, research, but faces limitations and challenges.
JMIR Medical Education is launching a theme issue on AI.
26 [26] To explore the potential applications of ChatGPT in assisting researchers with tasks such as literature review, data analysis, hypothesis creation, and text generation.
Literature review, data analysis, hypothesis creation, text generation.

•
Lack of context and nuance.
Scientific caution is needed in using ChatGPT due to plagiarism risks, misleading outcomes, lack of context, and AI limitations.
27 [62] To examine the potential impact of large language models (LLM), specifically "ChatGPT", on the nuclear medicine community and its reliability in generating nuclear medicine and molecular imagingrelated content.
Text generation, collaborative tool.

•
Limited learning and improvement.
ChatGPT's multiple-choice answering accuracy was 34%, surpassing random guessing.Improving training and learning capabilities is crucial.
28 [63] To explore applications and limitations of the language model ChatGPT in healthcare.
Investigating ChatGPT's potential in clinical practice, scientific production, reasoning, and education.

•
Real-world performance uncertainty.

•
High-level clinical expertise challenge.
ChatGPT-4 shows potential in patient education for cosmetic surgery, offering accurate and clear information, but its limitations require consideration.

[12]
To explore the capabilities and applications of ChatGPT, a large language model developed by OpenAI.

•
Inability to provide detailed information.

•
Limited scope for complex medical cases.

•
Potential misinformation due to lack of medical expertise.
ChatGPT demonstrates potential in offering valuable medical information to patients, particularly when access to professionals is restricted.However, its limitations and scope should be further investigated for safe and effective use in healthcare.

•
The study acknowledges that, while ChatGPT and other AI-generated content have the potential to improve healthcare communication, they should not be seen as a complete replacement for human interaction.

•
The limitations of AI, such as lack of empathy and context understanding, should be considered when implementing such technology.

•
The study does not delve deeply into the technical challenges and potential biases that AI-generated content may carry.
ChatGPT aids hospitals in enhancing patient care and communication efficiency.
Preliminary medical assessments; quick information for users seeking advice.
• Variable response quality.

•
Inaccuracies in treatment recommendations, ethical concerns.
ChatGPT's case responses need improvement; users require expertise for interpretation.

[73]
To assess ChatGPT's abilities in generating ophthalmic discharge summaries and operative notes.
ChatGPT can potentially aid in creating accurate and rapid ophthalmic discharge summaries.

•
The generated content might include generic text and disclaimers.

•
The quality of responses varies based on the quality of inputs.

•
Detailed operative notes require significant tuning for accuracy.

•
ChatGPT's responses may still contain factual inaccuracies that need correction.
ChatGPT's performance in ophthalmic notes is promising, rapid, and impactful with focused training and human verification.This study highlights ChatGPT's potential benefits in healthcare but concerns about accuracy and reliability persist.Trustworthy implementation requires addressing these issues.
47 [74] To evaluate ChatGPT's performance in comprehending complex surgical clinical data and its implications for surgical education.
Assessing ChatGPT (GPT-3.5 and GPT-4) in understanding surgical data for improved education and training.To explore the utility and accuracy of using ChatGPT-3, an AI language model, in understanding and discussing complex psychiatric diagnoses such as catatonia.
It investigates the potential of ChatGPT-3 as a tool for medical professionals to assist in understanding and discussing complex psychiatric diagnoses.It demonstrates the feasibility of using AI in medical research and education.
• It relies on the accuracy of ChatGPT-3's responses, which may not always be reliable.

•
The study's findings are based on the knowledge and expertise of the author and the neuropsychiatrist, which could introduce subjective biases.

•
The accuracy of ChatGPT-3's responses may vary depending on the quality of the prompt provided.To explore the potential application of ChatGPT as a medical assistant in Mandarin-Chinesespeaking outpatient clinics, aiming to enhance patient satisfaction and communication.
ChatGPT enhances clinic communication, aids patient satisfaction, excels in exams, and improves interactions in non-English medical settings.

•
Clinical trials and approval.

•
Senior physician evaluation may introduce subjectivity.

•
ChatGPT's responses might not fully capture the nuances of real patient interactions.

•
Its generalisability might be limited by the specific cases and physician preferences.
54 [77] To evaluate the accuracy and effectiveness of using ChatGPT for simulating standardised patients (SP) in medical training.
ChatGPT is utilised to simulate patient interactions, saving time, resources, and eliminating complex preparation steps.Offers intelligent, colloquial, and accurate responses, potentially enhancing medical training efficiency.

•
Senior physician evaluation may introduce subjectivity.ChatGPT's responses might not fully capture the nuances of real patient interactions.

•
Generalisability might be limited by the specific cases and physician preferences.

[21]
To evaluate the performance, potential applications, benefits, and limitations of ChatGPT in medical practice, education, and research.

ChatGPT demonstrates proficiency in medical exams
and academic writing, potentially aiding medical education, research, and patient-provider communication.

•
ChatGPT's handling of harmful content, misinformation, and plagiarism is unclear.

•
Complex medical tasks and understanding human anatomy remain challenging.

•
In radiological reports, accuracy and completeness can be concerns.

Medical practice + education + research, G1, G7.
ChatGPT holds potential for medical practice, education, and research but requires refinements before widespread use.Human judgment remains crucial despite its sophistication.• While ChatGPT and LLMs contribute to ethics education.

•
They are not a complete replacement for human-led teaching due to inherent limitations.
Considering strengths and limitations, ChatGPT can be an adjunctive tool for ethics education in healthcare training, accounting for evolving technology.
57 [37] To determine if ChatGPT can support multidisciplinary tumour board in breast cancer therapy planning.
ChatGPT's application in aiding breast cancer therapy decisions; benefits include efficiency and broader information access.
• Lack of Specificity: ChatGPT mostly offers general advice, lacking precision in therapy recommendations.

•
Complex Case: Struggles with complex cases, mislabelling endocrine therapy as hormonal treatment.

•
Inadequate Depth: Inability to provide comprehensive, patient-specific recommendations.
Artificial intelligence aids personalised therapy; ChatGPT's potential in clinical medicine is promising, but it lacks specific recommendations for primary breast cancer patients.
58 [22] The aim of this study was to assess the performance of ChatGPT models in higher specialty training for neurology and neuroscience, particularly in the context of the UK medical education system.
It demonstrates the potential application of ChatGPT models in medical education, specifically in the field of neurology and neuroscience.It highlights their ability to perform at or above passing thresholds in specialised medical examinations, offering a tool for enhancing medical training and practice.

•
Limited specialty focus.ChatGPT assists patients with health inquiries, potentially enhancing information accessibility in healthcare.
ChatGPT's potential in health info provision exists, but development and source quality improvement are necessary.
ChatGPT transforms human-machine interactions, offering innovative solutions, such as enhancing pharmacovigilance processes.

•
Potential for biased or incorrect outputs.

•
Limited understanding of context and nuance.

•
Dependency on input quality and relevance.

•
Ethical concerns in sensitive domains.
61 [38] Evaluate the impact of AI, particularly ChatGPT, in the field of surgery, considering both its benefits and potential harms.
AI, including ChatGPT, can enhance surgical outcomes, diagnostics, and patient experiences.It offers efficiency and precision in surgical treatments.

• Risk of inappropriate interventions by nonmedical individuals. •
Concerns over data privacy and security.

•
Ethical challenges in patient data usage.

•
Possibility of undermining medical professionals' roles.
The growing influence of AI in surgery demands ethical contemplation.To assess ChatGPT's performance in understanding Chinese medical knowledge, its potential as an electronic health infrastructure, and its ability to improve medical tasks and interactions, while acknowledging challenges related to hallucinations and ethical considerations.
• Chinese Medical Knowledge Assessment.
This study aims to evaluate the potential utilisation of ChatGPT in clinical practice and scientific research of infectious diseases, along with discussing relevant social and ethical implications.

• Ethical and Attribution Concerns •
Human Oversight and Responsibility.
To explore the importance of simplifying operations and creating user-friendly interfaces in AI-based medical applications, drawing insights from the success of ChatGPT and its impact on user adoption and clinical practice.

•
Need for Human Oversight and Policies.
To explores AI's impact on dental assistants and nurses in orthodontic practices, examining evolving treatment workflows.
• AI enhances treatment precision and personalisation.

•
Automation of assessments saves time.

•
AI can assist in identifying treatment progress.

•
Potential for improved patient engagement.

•
New roles for dental assistants and nurses.

•
Can streamline administrative tasks.

•
May lead to more efficient and effective care.
• Raises ethical and legal concerns.

•
Dependence on AI may reduce critical thinking.

•
Patient trust in AI may vary.

•
Implementation costs and training needed.

•
Privacy and security risks with patient data.

•
Potential for job displacement.

•
Need for ongoing AI regulation and oversight.

Challenges
In the field of healthcare and the medical field, refer to the difficult or complex issues, obstacles, or problems that healthcare professionals, researchers, organisations, and technologies face while providing medical care, conducting research, and addressing public health concerns.These challenges can arise due to various factors such as scientific advancements, technological limitations, ethical considerations, regulatory frameworks, economic constraints, patient expectations, and more.
In the context of healthcare and the medical field, challenges can encompass a wide range of issues, including but not limited to medical advancements, patient care, resource allocation, healthcare access, disease prevention and control, chronic disease management, healthcare costs, medical ethics, data privacy and security, protecting patient data and ensuring compliance with privacy regulations, interdisciplinary collaboration, patient education, and public health initiatives.In Figure 5, there were numerous challenges that Chatbot models such as GPT-3 could face in the field of healthcare and medicine.

Challenges
In the field of healthcare and the medical field, refer to the difficult or complex issues obstacles, or problems that healthcare professionals, researchers, organisations, and tech nologies face while providing medical care, conducting research, and addressing public health concerns.These challenges can arise due to various factors such as scientific ad vancements, technological limitations, ethical considerations, regulatory frameworks economic constraints, patient expectations, and more.
In the context of healthcare and the medical field, challenges can encompass a wide range of issues, including but not limited to medical advancements, patient care, resource allocation, healthcare access, disease prevention and control, chronic disease management healthcare costs, medical ethics, data privacy and security, protecting patient data and en suring compliance with privacy regulations, interdisciplinary collaboration, patient educa tion, and public health initiatives.In Figure 5, there were numerous challenges that Chatbo models such as GPT-3 could face in the field of healthcare and medicine.

Language Understanding and Medical Terms
Medical professionals frequently use specialised terminology and jargon that may be difficult for a general-purpose chatbot to understand.It is difficult to ensure that the chat bot understands medical terminology correctly.
Importance: this paper can illustrate the sophistication needed in natural language processing for medical contexts, underscoring the potential of advanced AI models to bridge communication gaps in healthcare.
Benefits: readers will discover the critical role of contextually aware AI in compre hending patient interactions, resulting in better patient support and care.

Accuracy and Reliability
Of utmost importance in healthcare is ensuring that the information provided by the chatbot is accurate and reliable.Medical information, which can be complex and critical must be accurate to avoid misinformation and potential harm to patients.
Importance: this paper's contribution to the ongoing conversation about the reliability of AI in clinical settings is highlighted, presenting the chatbot as a tool to augment, rathe than replace, human judgement.
Benefits: readers can derive advantages from comprehending the significance of error checking mechanisms and the ongoing updating of medical databases that AI systems must integrate.

Language Understanding and Medical Terms
Medical professionals frequently use specialised terminology and jargon that may be difficult for a general-purpose chatbot to understand.It is difficult to ensure that the chatbot understands medical terminology correctly.
Importance: this paper can illustrate the sophistication needed in natural language processing for medical contexts, underscoring the potential of advanced AI models to bridge communication gaps in healthcare.
Benefits: readers will discover the critical role of contextually aware AI in comprehending patient interactions, resulting in better patient support and care.

Accuracy and Reliability
Of utmost importance in healthcare is ensuring that the information provided by the chatbot is accurate and reliable.Medical information, which can be complex and critical, must be accurate to avoid misinformation and potential harm to patients.
Importance: this paper's contribution to the ongoing conversation about the reliability of AI in clinical settings is highlighted, presenting the chatbot as a tool to augment, rather than replace, human judgement.
Benefits: readers can derive advantages from comprehending the significance of errorchecking mechanisms and the ongoing updating of medical databases that AI systems must integrate.

Privacy and Security
Healthcare data are highly sensitive and subject to strict privacy regulations (like HIPAA in the United States).Chatbots need to adhere to these regulations and ensure that patient data are handled securely and confidentially.
Importance: the stress is on the advanced security protocols and compliance standards that AI systems must adhere to, as they are a cornerstone of healthcare technology.
Benefits: this paper educates readers on the stringent data protection measures required for AI integration in healthcare, with an emphasis on the technology's potential to maintain confidentiality.

Accountability and Responsibility
Decisions made by chatbots in healthcare settings can indeed have real-life consequences.Determining who is accountable for incorrect advice or recommendations provided by a chatbot can be a complex matter.
Importance: enriching the discourse on the role of AI in healthcare decision making by examining the legal and ethical frameworks that regulate AI.
Benefits: the survey can educate readers on the complex interaction between AI recommendations and human decision making, as well as the legal implications.

Human Oversight and Intervention
Although chatbots can assist in various healthcare tasks, there is a requirement for human oversight and intervention, particularly in critical situations.The challenge lies in balancing automation with the human touch.
Importance: the need for human experts to oversee AI systems is underlined, reinforcing the notion of AI as a supportive tool rather than a replacement.
Benefits: healthcare professionals can learn about the importance of their experience in supervising AI, ensuring patient safety and providing high-quality care.

Medical Training and Regulation
Chatbots that provide medical advice may unintentionally bypass the regulatory frameworks established for traditional healthcare providers.It is crucial to ensure that chatbots are developed and used within established medical guidelines.
Importance: address the incorporation of AI within existing medical regulations, emphasising how the paper advances the conversation about regulatory adaptations for AI tools.
Benefits: this paper can help medical professionals understand the importance of regulatory compliance for AI, as well as encourage proactive engagement with technology.

Patient Empowerment
While chatbots can provide information, they should also encourage patients to consult qualified healthcare professionals for personalised advice.It is critical to strike a balance between empowerment and not substituting professional medical care.
Importance: it highlights the chatbot's contribution to promoting well-informed patient decision making, underscoring the paper's adherence to patient-centred care models.
Benefits: readers will appreciate how AI can provide information to patients while also understanding AI's limitations in providing personalised medical advice.

Limitations and the Motivation
The limitations of these studies include a tendency to focus on theoretical potential rather than practical implementation, which leaves some real-world challenges unad-dressed.Furthermore, significant concerns persist regarding the ethical and legal issues surrounding the roles, accuracy, and originality of AI.There exists a necessity for more comprehensive evaluations of content generated by AI and its impact on tasks performed by humans.Additionally, there is a potential risk of overreliance, which could result in a decrease in human critical thinking and involvement.Current studies frequently lack a thorough exploration of the broader implications, which necessitates adopting a more holistic approach to comprehending the complete scope and consequences of integrating AI, such as ChatGPT, into diverse fields.
The integration of ChatGPT in medicine and healthcare, though promising, does come with certain limitations.One key concern is the potential for inaccuracies in medical information provided by ChatGPT, as it lacks the ability to fully comprehend complex medical contexts.Serious medical errors could occur if patient queries are misinterpreted or incorrect diagnoses are generated.Furthermore, the reliance of ChatGPT on training data may introduce biases, which could potentially impact the quality and fairness of the provided information.Ethical challenges arise regarding patient data privacy and consent, as well as accountability for AI-generated medical content.The sharing of outdated or obsolete medical information could be a potential consequence of the lack of real-time updates and dynamic learning in ChatGPT.Lastly, there is a risk of overdependence on AI, which may diminish the role of healthcare professionals, thereby reducing critical thinking and human interaction in medical care.
The motivation for integrating ChatGPT with medicine and healthcare stems from the potential to improve patient care, medical research, and clinical decision-making processes.ChatGPT's natural language processing capabilities provide a user-friendly interface for patients to seek medical information, resulting in increased patient engagement and empowerment.Its ability to analyse large amounts of medical literature assists healthcare professionals in staying up to date on the latest research and treatment options.Chat-GPT can assist in the generation of accurate and concise medical documentation, thereby streamlining administrative tasks for healthcare providers.In addition, it holds promise in facilitating medical education, supporting remote consultations, and optimising clinical workflows.In general, the integration of ChatGPT with medicine and healthcare is in line with the objective of utilising AI technology to progress medical practices and improve patient outcomes.

Conclusions
As part of this investigation, a PRISMA-based systematic literature review was conducted.AI has the potential to transform many industries, including medicine and healthcare.Large language models, such as ChatGPT, have attracted attention for their ability to generate human-like text, revolutionising the healthcare landscape.This paper investigated ChatGPT's applications in medicine, including pandemic management, surgical consultations such as cosmetic orthognathic surgery, dental practices, medical education, disease diagnosis, radiology and sonar imaging, pharmaceutical research, and treatment.The use of ChatGPT in disseminating critical information during pandemics and infectious diseases has showcased its ability to efficiently communicate essential updates and guidelines to the public.Furthermore, ChatGPT demonstrated its role as a virtual assistant in the context of cosmetic orthognathic surgery consultations.It provided standardised information and facilitated informed patient decisions.The integration of AI into dental practices has enhanced patient experiences by efficiently addressing inquiries and providing guidance.
The ability of ChatGPT to simplify complex medical concepts has enhanced medical education, leading to improved knowledge acquisition.It also displayed potential in disease diagnosis by offering differential diagnoses based on patient-reported symptoms and medical history.In the fields of cellular imaging, radiology, and sonar imaging, the descriptive capabilities of ChatGPT could be helpful for radiologists in the analysis of medical images and cellular imaging.Pharmaceutical research, benefiting from ChatGPT, was another realm that expedited research processes.It accomplished this by sifting

Figure 1 .
Figure 1.Categories and groups covered by all studies of the review.G1: treatment and medi G2: buildings and equipment, G3: parts of the human body and areas of the disease, G4: pati G5: citizens, G6: cellular imaging, radiology, pulse, and medical images, G7: doctors and nu and G8: tools, devices and administration.

Figure 1 .
Figure 1.Categories and groups covered by all studies of the review.G1: treatment and medicine, G2: buildings and equipment, G3: parts of the human body and areas of the disease, G4: patients, G5: citizens, G6: cellular imaging, radiology, pulse, and medical images, G7: doctors and nurses, and G8: tools, devices and administration.

Figure 3 .
Figure 3. Distribution by type of articles to number of articles.

Figure 3 .
Figure 3. Distribution by type of articles to number of articles.

Figure 3 .
Figure 3. Distribution by type of articles to number of articles.

Figure 4 .
Figure 4. Flowchart for selecting studies with specific query and eligibility criteria.

Figure 4 .
Figure 4. Flowchart for selecting studies with specific query and eligibility criteria.

3. 1 .
RQ1: How Does ChatGPT Contribute to Pandemic Management and What Specific Advantages Does It Offer in Disseminating Critical Information during Health Crises?

3. 4 .
RQ4: What Are the Key Components Associated with Work Related to ChatGPT in Medicine and Healthcare and Their Contributions in ChatGPT Applications in the Field?

Figure 5 .
Figure 5. Challenges of healthcare and the medical fields.

Figure 5 .
Figure 5. Challenges of healthcare and the medical fields.

Table 1 .
The summary of related works includes the study's goal, design, application(s), benefit(s), risk(s), concern(s), limitation(s), study field/area, conclusion(s), and suggested action for ChatGPT with medicine and healthcare.

Table 2 .
Summary of associated works includes all fields of study, study aim, pros, and cons or challenges.
• Need for updated ethical frameworks.