Prerpints.org logo
Preprint
Review

The Role of ChatGPT in the Advancement of Diagnosis, Management, and Prognosis of Cardiovascular and Cerebrovascular Disease

Altmetrics

Downloads

164

Views

63

Comments

0

A peer-reviewed article of this preprint also exists.

Submitted:

04 October 2023

Posted:

05 October 2023

You are already at the latest version

Alerts
Abstract
Cardiovascular and cerebrovascular disease incidence has risen mainly due to poor control of preventable risk factors and still constitutes a significant financial and health burden worldwide. ChatGPT is an artificial intelligence language-based model developed by OpenAI. Due to the model’s unique cognitive capabilities beyond data processing and the production of high-quality text, there has been a surge of research interest concerning its role in the scientific community and contemporary clinical practice. To fully exploit ChatGPT's potential benefits and reduce its possi-ble misuse, extreme caution must be taken to ensure its implications ethically and equitably. In this article, we explore the language model's possible applications and limitations while empha-sizing its potential value for diagnosing, managing, and prognosis cardiovascular and cerebro-vascular disease.
Keywords: 
Subject: Medicine and Pharmacology  -   Cardiac and Cardiovascular Systems

1. Introduction

Cardiovascular and cerebrovascular disease are still the leading cause of premature morbidity, mortality, and disease disability on a global scale. (1) Despite significant advancements in the standard of care, particularly in higher-income countries, the incidence of these diseases continues to rise. (2) This finding has been partially attributed to poorly controlled preventable risk factors such as hypertension, hyperlipidemia, smoking, diabetes, and obesity, the latter now being widely considered a pervasive epidemic of the 21st century, despite the significant development of newer pharmaceutic agents for primary and secondary prevention, improved invasive procedures and better diagnostic tools. (3) Moreover, the financial burden of cardiovascular and cerebrovascular diseases remains very high and has increased over decades; from 1996 to 2016, United States adult cardiovascular spending has increased by over 100 billion dollars. (4) Similar findings have been observed in the European Union, where the annual cost is about 282 billion euros, with health and long-term care accounting for 155 billion (55%), equivalent to 11% of EU health expenditure. Coronary heart disease and cerebrovascular disorders are accounted for the 27% (€76 billion) of those costs. (5)
In this era of rapid technological progress, artificial intelligence (AI) has emerged as a central pillar of the fourth industrial revolution. Artificial intelligence encompasses a field of computer science in which machines are trained to perform tasks that typically require human intelligence. AI has already gained ground in radiology, where it helps with more accurate image processing and interpretation. (6) In this context, over ten artificial intelligence models have been developed, each demonstrating remarkable capabilities. (7) Since primary prevention and risk factor modification play a pivotal role in the management of cardiovascular and cerebrovascular disease, there has been a surge of interest in artificial intelligence’s role, especially for language-based models, in the advancement of cardiovascular and cerebrovascular disease management and financial burden reduction. One such language-based model, ChatGPT, is at the forefront of this artificial intelligence revolution.
ChatGPT is an artificial intelligence language-based model developed by OpenAI and was based on the Generative Pre-trained Transformer (GPT) architecture, specifically the GPT-4 version. GPT-4 version is a subtype of machine learning algorithms and more specifically a neural network whose primary functions are to understand, generate human-like text, and perform a variety of natural language processing tasks, depending on the input provided. These include text generation, language translation, text summarization, sentiment analysis, named entity recognition, and question answering. The model first launched on November 30, 2022, and quickly divided the scientific community with its capabilities (8,9)ranging from simple student essay generation to full-blown research articles, with submission even including ChatGPT as a co-author. (10) Less than a year after its release, more than 1,200 results appeared after using ChatGPT as a keyword in searches, showing its significant impact on current medical research. The extent of its abilities prompted essential questions regarding its potential application in clinical practice, specifically its power to revolutionize critical medical fields such as cardiovascular and cerebrovascular disease management. The primary objective of this narrative review article is to explore the potential role of ChatGPT in those fields.

2. Search Methods

A comprehensive literature review was meticulously conducted for published articles focused on using ChatGPT in cardiovascular and cerebrovascular diseases. MEDLINE database via PubMed was used to collect relevant articles. A search set combined the keyword "ChatGPT" to gather all the relevant papers on this topic, and no language restriction was used. Using this approach, all articles related to ChatGPT's application in cerebrovascular and cardiovascular disease were collected. Overall, this study analyzed the potential application of ChatGPT for the prompt diagnosis, prognosis, and management of patients with cardiovascular and cerebrovascular disease by considering various published articles.

3. ChatGPT for Diagnosis of Cardiovascular and Cerebrovascular disease

Cardiovascular and cerebrovascular disease manifest from multifaceted complex pathophysiological etiologies, having been linked with multiple risk factors. (11) Given their extremely high impact on morbidity and mortality associated with their prevalence, particular emphasis is required on risk factor modification as a primary means of reducing their incidence. In this context, we initiated a conversation with ChatGPT to explore its potential as a diagnostic tool. The response of the model elucidated several critical instances through which ChatGPT can enhance the diagnosis of these diseases:
ChatGPT can be used to advancethe diagnosis of cardiovascular disease and stroke by analyzing patient data and providing personalized recommendations to healthcare providers, leveraging its natural language processing and machine learning capabilities. Here are some specific ways in which ChatGPT can be used to improve the diagnosis of cardiovascular disease and cerebrovascular disease:
Symptom analysis: ChatGPT can assist in analyzing patient-reported symptoms related to cardiovascular issues or stroke. Patients often describe their symptoms in plain language, and ChatGPT can understand and interpret these descriptions. For example, if a patient reports symptoms such as chest pain, shortness of breath, or sudden weakness on one side of the body, ChatGPT can analyze this information and provide insights to healthcare providers. This can help identify potential cardiovascular or stroke-related issues early in the diagnostic process.
Differential diagnosis: ChatGPT can aid healthcare providers in generating a comprehensive list of potential diagnoses based on a patient's reported symptoms and medical history. By cross-referencing the patient's information with a vast database of medical knowledge and guidelines, ChatGPT can help rule out non-cardiovascular or non-stroke-related conditions and provide a more accurate differential diagnosis. This supports healthcare professionals in making informed decisions about further diagnostic tests and evaluations.
Risk assessment: In addition to symptom analysis, ChatGPT can evaluate other risk factors associated with cardiovascular disease and stroke. It can consider a patient's age, gender, family history, lifestyle factors (such as smoking and diet), and relevant medical history (such as diabetes or hypertension). ChatGPT can provide insights into a patient's overall risk of developing cardiovascular disease or experiencing a stroke by assessing these risk factors. This information helps in risk stratification and informs the diagnostic process.
Real-time data collection: ChatGPT can collect and analyze patient data during interactions. For example, suppose a patient is experiencing acute symptoms, such as sudden chest pain or neurological deficits. In that case, ChatGPT can prompt the patient to provide additional information about the symptoms, their duration, and any associated factors. This real-time data collection can aid healthcare providers in making prompt and accurate diagnostic decisions, particularly in emergency situations.
Referral recommendations: Based on the analysis of symptoms and risk factors, ChatGPT can suggest appropriate referrals to other healthcare providers or specialists. For instance, if a patient's symptoms indicate a potential heart condition, ChatGPT can recommend a cardiology consultation or additional cardiac tests. These referral recommendations help ensure that patients receive timely and appropriate care from the right specialists.
Education and Patient Empowerment: ChatGPT can educate patients about the importance of early diagnosis and seeking medical attention for symptoms related to cardiovascular disease or stroke. It can explain the significance of diagnostic tests, such as electrocardiograms (ECGs), echocardiograms, or brain imaging, and address any concerns or misconceptions patients may have. By empowering patients with knowledge, ChatGPT can encourage them to proactively engage in the diagnostic process.
Support for Telemedicine: In the era of telemedicine, ChatGPT can serve as a virtual assistant during remote consultations with healthcare providers. It can help patients describe their symptoms accurately and provide healthcare professionals with essential information for remote diagnosis. This enhances the accessibility of healthcare services, especially for patients in remote or underserved areas.
In summary, ChatGPT's role in the advancement of the diagnosis of cardiovascular disease and stroke is multifaceted. It contributes to symptom analysis, supports differential diagnosis, assesses risk factors, collects real-time data, facilitates referral recommendations, educates patients, and enhances telemedicine interactions. By assisting healthcare providers in the diagnostic process, ChatGPT plays a valuable role in the early detection and management of these critical medical conditions, ultimately leading to better patient outcomes.
Regarding diagnosis for cardiovascular and cerebrovascular disease, its main strengths are the system’s capacity to offer instantaneous responses regarding symptom analysis, risk assessment, real-time data collection, differential diagnosis aid, referral recommendations, education and patient empowerment, and support for telemedicine. Indeed, language-based models with a simple chatbox-like interface can analyze the symptoms experienced by the patient or other inserted data as input and, after examining them, return a result based on the patient’s disease risk. Against this background, Rizwan and Sadiq explored the diagnostic accuracy and accessibility of ChatGPT in cardiovascular diseases compared to experts in clinical cardiology and imaging. Eight of the ten clinical scenarios processed by ChatGPT were perfectly diagnosed; however, the other two answers given by ChatGPT were partially incorrect since those diseases were associated with the actual diagnosis. Moreover, the therapeutic protocols that ChatGPT suggested in the same study were in accordance with the literature and current medical knowledge. (12) In addition, Kusunose et al. investigated the ability of ChatGPT to answer clinical questions stated in the Japanese Society of Hypertension Guidelines. ChatGPT showed a high accuracy rate, but it varied by available evidence. (13) This ability could be enhanced by incorporating the 10-year risk prediction algorithms for cardiovascular disease estimates, like the SCORE2 for the European population and SCORE2-OP for the older population in four geographical risk regions(14,15). Thus, it can aid in the early detection of high-risk populations and, by analyzing individualized data, refer them promptly to primary prevention care.This feature can be crucial and lifesaving, especially when prompt decisions are imperative, such as determining the need for reperfusion strategies. In addition, artificial intelligence models have been specifically trained to process large amounts of data in much less time than would otherwise require a person. Thus, it may facilitate not only data extraction from large healthcare datasets for research purposes without the risk of human error but also in generating succinct patient medical records or patient history summaries, which are essential for more effective patient management.
However, this is not the only utilization of ChatGTP in medical research. Recently, Teperikidis and his colleagues performed an umbrella review and searched for systematic reviews and meta-analyses to answer the research question of whether there is a causal relationship between protein pump inhibitor administration and major adverse cardiovascular events. The originality of this project was not the main focus but the utilization of ChatGTP in the automated process of this research project. Two independent reviewers performed the systematic search, while a second pair of investigators performed the same task by exclusively using ChatGPT. Interestingly, the studies utilizing ChatGPT were consistent with the conventional manual approach, adding another artificial intelligence capability in medical research. (16,17) Furthermore, there is great potential in other areas of medicine, like patient education. The interaction of patients with the language-based model could provide the initial motivation to get referred to the appropriate healthcare specialist or emergency department. Moreover, numerous patients seek medical advice in search engines, read inappropriate articles, and adopt symptoms and signs, which misdrives them to wrong clinical self-diagnosis and stress them. In this context, ChatGPT could also play an advisory role to these patients, as its recommendations will be more reliable, taking into consideration that such advice is more personalized and evidence-based. (18) Furthermore, the ability to simplify difficult-to-understand pathophysiological concepts or procedures while also being readily accessible makes it a solid option for patient education armamentarium. In this context, Khan et al. explored the capacity of the model to hold conversations about diabetes in a casual manner. They found that its responses were scientifically sound and easy to interpret, signifying the potential to extrapolate this ability with other diseases. (19) In addition, these responses can even be tailored to patients’ specific educational levels and replace the general written materials that utilize a general text for all. Although tested in a different setting than cardiovascular diseases, ChatGPT’s role in generating patient education materials was examined by Hung et al. in the vicinity of plastic surgery and compared to ones created by experts. (20) The authors found that ChatGPT’s education materials were generated in 30 minutes, while the content generated by experts took approximately one month. However, this came at the cost of lower accuracy (50%) and readability of its content measured using the Flesch-Kincaid score (10.5 grade versus 7.5 grade respectively), stressing that there is still room to grow. Chiwon Ahn et al. explored whether ChatGPT can provide reliable information regarding cardiopulmonary resuscitation, which is a topic of general interest and a life-saving procedure for the general population. The authors noted that although ChatGPT cannot provide visual feedback to help readers comprehend the necessary steps, the answers it offered were descriptive and scientifically correct, adding to its potential for general medical educational purposes for a significant public health issue. (6)

4. ChatGPT as a tool for Secondary Prevention and Management

Moreover, the promising future of ChatGPT in primary prevention strategies has also fueled the interest in its potential value as a cardiovascular and cerebrovascular disease management and prognostic tool (secondary prevention). Again, we began a second conversation with ChatGPT regarding its importance in this domain. Its response was as follows:
Chat GPT can be used to advance the management and prognosis of cardiovascular disease and stroke in several ways, leveraging its natural language processing and machine learning capabilities. Here are some specific ways in which Chat GPT can help healthcare providers to manage and predict outcomes for patients with cardiovascular disease and stroke:
Risk stratification: ChatGPT can analyze patient data, including medical history, lifestyle factors, and diagnostic test results, to provide risk stratification for individuals with CVD or stroke. By assessing various risk factors such as age, gender, family history, smoking status, and cholesterol levels, ChatGPT can help healthcare providers identify patients who are at higher risk of adverse outcomes. This risk stratification enables healthcare teams to allocate resources more effectively and provide more intensive management for high-risk patients.
Treatment optimization: ChatGPT can assist healthcare providers in optimizing treatment plans for patients with CVD or stroke. It achieves this by analyzing patient data and generating personalized treatment recommendations. These recommendations take into account the patient's medical history, comorbidities, and medication preferences. By tailoring treatment plans to individual patients, ChatGPT helps healthcare providers identify the most effective interventions, reducing the risk of adverse events and optimizing outcomes.
Medication management: Ensuring medication adherence is crucial in managing CVD and stroke. ChatGPT can play a vital role in helping patients manage their medication regimens effectively. It can provide patients with reminders for medication adherence, explain the importance of each medication, and address any concerns or questions patients may have about their prescribed medications. By improving medication adherence, ChatGPT contributes to better disease management and outcomes.
Symptom management: Patients with CVD or stroke often experience symptoms that require ongoing management. ChatGPT can assist healthcare providers in managing these symptoms by analyzing patient-reported data and providing personalized recommendations. For instance, if a patient reports symptoms like chest pain or shortness of breath, ChatGPT can offer guidance on symptom management strategies, helping patients improve their quality of life and reduce the risk of complications.
Predictive modeling: Chat GPT can help healthcare providers predict outcomes for cardiovascular or stroke patients. By analyzing patient data and identifying patterns, Chat GPT can help healthcare providers predict the likelihood of adverse effects and take proactive steps to prevent them.
Rehabilitation Support: Patients often require rehabilitation and lifestyle adjustments after experiencing a cardiovascular event like a heart attack or stroke. ChatGPT can assist in providing guidance on post-event rehabilitation programs, exercise routines, dietary changes, and lifestyle modifications. It can offer personalized recommendations to help patients recover and reduce the risk of future cardiovascular events.
Patient Education: Educating patients about their condition, treatment options, and lifestyle modifications is essential for effective disease management. ChatGPT can serve as a valuable educational resource, providing patients with clear and accessible information. It can explain complex medical concepts, clarify treatment plans, and address patient queries, empowering patients to actively participate in their care.
In summary, ChatGPT's role in the management and prognosis of cardiovascular disease and stroke lies in its ability to provide personalized, data-driven insights and recommendations. By leveraging patient data and medical knowledge, ChatGPT enhances risk assessment, treatment optimization, medication management, symptom control, predictive modeling, rehabilitation support, and patient education. This comprehensive approach contributes to better patient outcomes, improved quality of care, and more effective management of these critical medical conditions.
For the management and prognosis of cardiovascular and cerebrovascular disease, the system’s value is, again, in providing instant and personalized questions after analyzing individualized data. The accelerated response time of the “heart or brain team” for reperfusion strategies can be lifesaving. At the same time, the model may also declare if they’re eligible for their inclusion in a trial and manage accordingly. (21)Moreover, since language-based models have already been considered a powerful tool in medical education, (22) they may be replicated in patient education as well. Medication nonadherence has been linked with poor patient and caregiver communication (23,24), and language-based models, with their natural text, can provide knowledge regarding the benefits and reasoning behind treatment plans while also answering patients’ questions at any time, even providing symptomatic treatment recommendations during the rehabilitation period after the event. Lastly, the enhanced data processing may further aid in identifying patients' unique clinical characteristics and thus suggest individualized treatment plans, facilitating the transition from evidence-based medicine to a more personalized one while also recognizing the increased likelihood for drug-to-drug interactions (DDIs) when proposing such a plan.

5. Discussion and Future Directions

An overall representation of the possibilities of ChatGPT in the clinical setting is depicted in Table 1. Even though the future of language-based models in medical care and research looks promising, it is of utmost importance to keep in mind several risks and concerns. Firstly, it is worth noting that the current ChatGPT version has been trained with data available up to 2021. Thus, it may need to be updated to keep up with the ever-changing medical landscape, deeming some of its recommendations not up-to-date. Furthermore, the inherent over-dependence on the training data has raised concerns about bias introduction, its external validity (25), and the potential risk of incomplete or inaccurate results. In this context, one of the main limitations of the language-based system is the hallucination bias, where the system reports a scientifically sound answer that cannot be realistically implemented. (26)Moreover, the model has a non-deterministic structure, meaning that the same input may result in different answers, hindering its classification capabilities. Thus, before implementation in the current evidence-based care system, several universal standards must be secured, like bias minimization, reproducibility, and accuracy of the results. In this domain, cross-validation of the information provided by ChatGPT must always be cross-checked with up-to-date human-annotated references and guidelines. Additionally, institutions and expert healthcare providers must solidify the model's unbiased responses while proving its benefit in the contemporary evidence-based medicine system with carefully planned research studies and legislating the context that the generated text follows a standardized reporting rule and does not misinform its users.
Lastly, its diagnostic deductive capabilities need to be more consistent and reproducible. In a study by Kung et al., the performance of ChatGPT was evaluated on the United States Medical Licensing Examination, and the language model performed at or near the passing threshold. (27) In another study, Skalidis et al. showed a similar performance of ChatGPT in the European Exam in Core Cardiology, as it answered correctly at about 60%, considered a passing score. Nevertheless, it could have been more efficient in the American Heart Association Basic Life Support and Advanced Cardiovascular Life Support exams. In this study, ChatGPT failed to reach the passing threshold for any of the exams performed; however, its answers were promising due to being more detailed than the rationale provided by the Advanced Cardiovascular Support Committee. (28) Additionally, in a cross-sectional study by Sinha et al. (2023) (29), the capabilities of ChatGPT in solving higher-order reasoning were evaluated. One hundred random questions, including cardiovascular disease, were used from different pathology and internal medicine areas. The system’s accuracy reportedly reached 75%, also underlining the room to grow in clinical decision-making. Nonetheless, it is evident that artificial intelligence will play an essential role in the next step of the medical field and decision-making. Although artificial intelligence is in its early stages, there is undoubtedly unexploited potential not only for clinical practice but also for research. (30) Indeed, the role of language-based models in academics has been debated. Even though it helps facilitate the writing efficacy, the time required for manuscript preparation, and assisting with data collection, interpretation, and calculations, the model has also been tested in using tools for systematic reviews. Mahuli et al. (31) tested the efficacy of ChatGPT in using risk-of-bias tools and for Data extraction by sharing complete text and materials that needed to be extracted. The authors found remarkable reproducible efficiency in both areas, underlining its potential to streamline these processes. Nevertheless, ChatGPT could improve the language and terminology used in scientific manuscripts and limit syntactical or grammatical errors without affecting the text's original meaning.
Despite its several limitations, ChatGPT is not far from becoming the “stethoscope” of the 2020 physician. Although the users cannot upload an image in the application yet, it is believed that it will be one of the upcoming futures of a newer version of ChatGPT as with the version 4. By expanding on this ability, users could upload electrocardiograms, and the system could aid in their interpretation. (32) Artificial intelligence, with promising results, is already used for ECG interpretation. At the same time, some studies report that its diagnostic accuracy is better than cardiologists-experts in ECG interpretation (33). In contrast, having been trained using data retrieved from large datasets, prediction modeling has been created to predict the risk of developing various cardiac diseases. (33) Adding to the imaging analysis potential, ChatGPT could even play a role in chest X-ray interpretation. At the same time, it is considered a fundamental imaging exam, with many stressing the need for every physician to acquire the skill of interpreting them. This is particularly useful in emergency departments, where X-ray misinterpretation may further drive clinical decision-making from the actual diagnosis. As such, numerous abnormal findings are still missed or misdiagnosed with high overall inter- and intra-variability between healthcare providers. (34) Thus, ChatGPT and AI-driven applications could improve clinical practice in the acute care setting, especially in primary healthcare centers, where general practitioners should take the first actions and guide the treatment plan, double-crossing the information provided by the language-based model.

6. Conclusion

In conclusion, the practice of medicine in the future will likely be influenced by the large language models like ChatGPT. The expanding potential uses for the diagnosis, management, and patient prognosis with high-incidence diseases like cardiovascular and cerebrovascular disease have definite potential through data extraction, high-risk population identification, and clinical decision-making. Nonetheless, as with every new technology it has limitations, and before implementation in the evidence-based care, responsible and thorough evaluation with universally accepted standards must be conducted. AI procedures should be performed under physicians’ guidance so as to eliminate the possibility of language-model mistakes.

Author Contributions

A short paragraph specifying their individual contributions must be provided for research articles with several authors. The following statements should be used “Conceptualization, D.K.; methodology, D.D.C.; software, D.D.C, S.P. and A.C.; validation, L.P., G.G. and S.X., D.K., A.A; writing—original draft preparation, D.D.C, A.A. and A.C.; writing—review and editing, L.P., G.G. and S.X., S.P, D.K; supervision, D.K.; All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Acknowledgments

None to declare.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Roth GA, Mensah GA, Johnson CO, Addolorato G, Ammirati E, Baddour LM, et al. Global Burden of Cardiovascular Diseases and Risk Factors, 1990-2019: Update From the GBD 2019 Study. J Am Coll Cardiol [Internet]. 2020 Dec 22 [cited 2023 Apr 26];76(25):2982–3021. Available online: https://pubmed.ncbi.nlm.nih.gov/33309175/ (accessed on 26 April 2023).
  2. Benjamin EJ, Virani SS, Callaway CW, Chamberlain AM, Chang AR, Cheng S, et al. Heart Disease and Stroke Statistics-2018 Update: A Report From the American Heart Association. Circulation [Internet]. 2018 Mar 1 [cited 2023 Apr 26];137(12):E67–492. Available online: https://pubmed.ncbi.nlm.nih.gov/29386200/ (accessed on 26 April 2023).
  3. Powell-Wiley TM, Poirier P, Burke LE, Després JP, Gordon-Larsen P, Lavie CJ, et al. Obesity and Cardiovascular Disease: A Scientific Statement From the American Heart Association. Circulation [Internet]. 2021 May 25 [cited 2023 Apr 26];143(21):E984–1010. Available online: https://pubmed.ncbi.nlm.nih.gov/33882682/ (accessed on 26 April 2023).
  4. Birger M, Kaldjian AS, Roth GA, Moran AE, Dieleman JL, Bellows BK. Spending on Cardiovascular Disease and Cardiovascular Risk Factors in the United States: 1996 to 2016. Circulation. 2021, 144, 271–82. [Google Scholar] [CrossRef] [PubMed]
  5. Luengo-Fernandez R, Walli-Attaei M, Gray A, Torbica A, Maggioni AP, Huculeci R, et al. Economic burden of cardiovascular diseases in the European Union: a population-based cost study. Eur Heart J. 2023, 26. [Google Scholar]
  6. Ahn, C. Exploring ChatGPT for information of cardiopulmonary resuscitation. Resuscitation 2023, 185, 109729. [Google Scholar] [CrossRef] [PubMed]
  7. Sarker IH. AI-Based Modeling: Techniques, Applications and Research Issues Towards Automation, Intelligent and Smart Systems. SN Comput Sci [Internet]. 2022 Mar [cited 2023 Apr 26];3(2). Available online: https://pubmed.ncbi.nlm.nih.gov/35194580/ (accessed on 26 April 2023).
  8. Muss Google zittern? Künstliche Intelligenz schreibt wie ein Mensch | Leben & Wissen | BILD.de [Internet]. [cited 2023 Apr 26]. Available online: https://www.bild.de/digital/2022/digital/muss-google-zittern-kuenstliche-intelligenz-schreibt-wie-ein-mensch-82129948.bild.html (accessed on 26 April 2023).
  9. The Brilliance and Weirdness of ChatGPT - The New York Times [Internet]. [cited 2023 Apr 26]. Available online: https://www.nytimes.com/2022/12/05/technology/chatgpt-ai-twitter.html (accessed on 26 April 2023).
  10. Stokel-Walker C. ChatGPT listed as author on research papers: many scientists disapprove. Nature [Internet]. 2023 Jan 1 [cited 2023 Apr 26];613(7945):620–1. Available online: https://pubmed.ncbi.nlm.nih.gov/36653617/ (accessed on 26 April 2023).
  11. Hansson GK. Inflammation, atherosclerosis, and coronary artery disease. N Engl J Med [Internet]. 2005 Apr 21 [cited 2023 Apr 26];352(16):1685–95. Available online: https://pubmed.ncbi.nlm.nih.gov/15843671/ (accessed on 26 April 2023).
  12. Rizwan A, Sadiq T. The Use of AI in Diagnosing Diseases and Providing Management Plans: A Consultation on Cardiovascular Disorders With ChatGPT. Cureus 2023, 7.
  13. Kusunose K, Kashima S, Sata M. Evaluation of the Accuracy of ChatGPT in Answering Clinical Questions on the Japanese Society of Hypertension Guidelines. Circulation Journal 2023, 87, CJ–23. [Google Scholar]
  14. de Vries TI, Cooney MT, Selmer RM, Hageman SHJ, Pennells LA, Wood A, et al. SCORE2-OP risk prediction algorithms: estimating incident cardiovascular event risk in older persons in four geographical risk regions. Eur Heart J. 2021, 42, 2455–67. [Google Scholar] [CrossRef] [PubMed]
  15. Hageman S, Pennells L, Ojeda F, Kaptoge S, Kuulasmaa K, de Vries T, et al. SCORE2 risk prediction algorithms: new models to estimate 10-year risk of cardiovascular disease in Europe. Eur Heart J. 2021, 42, 2439–54. [Google Scholar] [CrossRef] [PubMed]
  16. Teperikidis E, Boulmpou A, Papadopoulos C. Prompting ChatGPT to perform an umbrella review. Acta Cardiol. 2023, 1–2. [Google Scholar]
  17. Teperikidis E, Boulmpou A, Potoupni V, Kundu S, Singh B, Papadopoulos C. Does the long-term administration of proton pump inhibitors increase the risk of adverse cardiovascular outcomes? A ChatGPT powered umbrella review. Acta Cardiol. 2023, 1–9. [Google Scholar]
  18. Yeo YH, Samaan JS, Ng WH, Ting PS, Trivedi H, Vipani A, et al. Assessing the performance of ChatGPT in answering questions regarding cirrhosis and hepatocellular carcinoma. Clin Mol Hepatol. 2023, 29, 721–32. [Google Scholar] [CrossRef] [PubMed]
  19. Khan I, Agarwal R. Can ChatGPT Help in the Awareness of Diabetes? Ann Biomed Eng. 2023. [Google Scholar]
  20. Hung YC, Chaker SC, Sigel M, Saad M, Slater ED. Comparison of Patient Education Materials Generated by Chat Generative Pre-Trained Transformer Versus Experts. Ann Plast Surg. 2023, 91, 409–12. [Google Scholar] [CrossRef] [PubMed]
  21. Ho YC, Tsai TH, Sung PH, Chen YL, Chung SY, Yang CH, et al. Minimizing door-to-balloon time is not the most critical factor in improving clinical outcome of ST-elevation myocardial infarction patients undergoing primary percutaneous coronary intervention. Crit Care Med [Internet]. 2014 [cited 2023 Apr 26];42(8):1788–96. Available online: https://pubmed.ncbi.nlm.nih.gov/24717469/ (accessed on 26 April 2023).
  22. Lee, H.; The rise of ChatGPT: Exploring its potential in medical education. Anat Sci Educ [Internet]. 2023 Mar 28 [cited 2023 Apr 26]. Available online: https://pubmed.ncbi.nlm.nih.gov/36916887/ (accessed on 26 April 2023).
  23. Brown MT, Bussell J, Dutta S, Davis K, Strong S, Mathew S. Medication Adherence: Truth and Consequences. Am J Med Sci [Internet]. 2016 Apr 1 [cited 2023 Apr 26];351(4):387–99. Available online: https://pubmed.ncbi.nlm.nih.gov/27079345/ (accessed on 26 April 2023).
  24. Tsigkas G, Apostolos A, Aznaouridis K, Despotopoulos S, Chrysohoou C, Naka KK, et al. Real-world implementation of guidelines for heart failure management: A systematic review and meta-analysis. Hellenic Journal of Cardiology. 2022, 66, 72–9. [Google Scholar] [CrossRef] [PubMed]
  25. Shen Y, Heacock L, Elias J, Hentel KD, Reig B, Shih G, et al. ChatGPT and Other Large Language Models Are Double-edged Swords. Radiology [Internet]. 2023 Apr 1 [cited 2023 Apr 26];307(2). Available online: https://pubmed.ncbi.nlm.nih.gov/36700838/ (accessed on 26 April 2023).
  26. Ji Z, Lee N, Frieske R, Yu T, Su D, Xu Y, et al.
  27. Kung TH, Cheatham M, Medenilla A, Sillos C, De Leon L, Elepaño C, et al. Performance of ChatGPT on USMLE: Potential for AI-assisted medical education using large language models. PLOS Digital Health. 2023, 2, e0000198. [Google Scholar]
  28. Fijačko N, Gosak L, Štiglic G, Picard CT, John Douma M. Can ChatGPT pass the life support exams without entering the American heart association course? Resuscitation 2023, 185, 109732. [Google Scholar] [CrossRef] [PubMed]
  29. Sinha RK, Deb Roy A, Kumar N, Mondal H. Applicability of ChatGPT in Assisting to Solve Higher Order Problems in Pathology. Cureus [Internet]. 2023 Feb 21 [cited 2023 Apr 26];15(2). Available online: https://pubmed.ncbi.nlm.nih.gov/36968864/ (accessed on 26 April 2023).
  30. Salvagno M, Taccone FS, Gerli AG. Can artificial intelligence help for scientific writing? Crit Care. 2023, 27, 75. [Google Scholar] [CrossRef] [PubMed]
  31. Mahuli SA, Rai A, Mahuli AV, Kumar A. Application ChatGPT in conducting systematic reviews and meta-analyses. Br Dent J. 2023, 235, 90–2. [Google Scholar] [CrossRef] [PubMed]
  32. Martínez-Sellés M, Marina-Breysse M. Current and Future Use of Artificial Intelligence in Electrocardiography. J Cardiovasc Dev Dis. 2023, 10, 175. [Google Scholar]
  33. Bridge J, Fu L, Lin W, Xue Y, Lip GYH, Zheng Y. Artificial intelligence to detect abnormal heart rhythm from scanned electrocardiogram tracings. J Arrhythm. 2022, 38, 425–31. [Google Scholar] [CrossRef] [PubMed]
  34. Brady, AP. Error and discrepancy in radiology: inevitable or avoidable? Insights Imaging. 2017, 8, 171–182. [Google Scholar] [CrossRef] [PubMed]
Table 1. Potential ChatGPT Applications in Cardiovascular Disease Diagnosis and Treatment.
Table 1. Potential ChatGPT Applications in Cardiovascular Disease Diagnosis and Treatment.
Diagnostic Tool Potential Treatment Tool Potential
Patient reported symptom analysis Risk stratification
Comprehensive differential diagnosis Treatment optimization
Risk assessment Medication management
Real-time data collection Symptom management
Referral Recommendations Predictive modeling
Education and Patient Empowerment Rehabilitation Support
Support for Telemedicine Patient Education
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Copyright: This open access article is published under a Creative Commons CC BY 4.0 license, which permit the free download, distribution, and reuse, provided that the author and preprint are cited in any reuse.
Prerpints.org logo

Preprints.org is a free preprint server supported by MDPI in Basel, Switzerland.

facebook logotwitter logolinkedin logo
MDPI Initiatives
Important Links
Subscribe

Disclaimer

Terms of Use

Privacy Policy

© 2024 MDPI (Basel, Switzerland) unless otherwise stated