Artificial intelligence applications spreading into editorship: A critical conundrum for editors and publishers

Abstract

Artificial intelligence (AI) is a modern technological advancement that is rapidly evolving and spreading around the world. This technology is being introduced into different fields of application, including medicine and healthcare in general, disseminating technological innovations that are transforming medical practice and the lives of patients. Government organizations, public and private healthcare providers (deployers), as well as software companies (developers) are called to generate and govern applications designed to enhance diagnostic performance and patient management, delivering accurate care together with improved clinical outcomes. There are several successful examples that illustrate the positive impact of AI in medicine. In a recent study, a world-renowned hospital (Mount Sinai Health System in New York) implemented AI algorithms to analyze ultrasound images, resulting in faster and more accurate diagnoses of heart diseases. This implementation not only increased diagnostic efficiency but also improved patient outcomes, demonstrating the transformative potential of AI in clinical practice. In this type of implementation, a cost–benefit analysis should take into consideration the process of software and platform validation, and the development process should be overseen in generating AI with the specific purpose that it correctly responds to proven clinical usefulness.1 According to the United Nations Educational, Scientific and Cultural Organization (UNESCO), AI and large multi-modal models (LMMs) should strictly follow and respond to the principles of being transparent, accountable and understandable, safe and secure, of service to human beings, aligned with human values and inclusive.2 The UNESCO guidelines for appropriate development and use of AI, issued in 2021, are in agreement with the World Health Organization (WHO) recommendations, published in the same year and revised in 2024, titled Ethics and governance of artificial intelligence for health.3 Theoretically and ethically, AI should help humans of any race and economic status to enjoy the same rights with respect, which in medicine means receiving a gold standard of care, whether in low- or medium-high-income countries. Unfortunately, this universal right and goal has not been reached to date, and the dissemination of AI applications is still far from being equally distributed in the different countries. The implementation of AI has a significant economic impact, especially on healthcare systems in low- and middle-income countries. Making these technologies accessible globally requires strategies to reduce costs and increase availability. Initiatives that promote collaboration and resource sharing can help democratize access to these innovations and maximize their effectiveness in medical practice. The spread of AI algorithms and platforms in medicine is not the sole field of application. It has been estimated, for example, that the AI-based large-language writing program ChatGPT, only 2 months after its free version was released, had some 100 million active users monthly, making it the fastest-growing software application in human history.4 The ongoing technological revolution, driven by advances in AI and advanced statistical techniques, is reshaping the landscape of academia and publishing and especially the landscape of healthcare, bringing new technologies and ways to optimize processes and improve operational efficiency. Clinical ultrasound, a crucial diagnostic tool, may potentially benefit from these new technologies. In fact, AI algorithms are being developed to interpret images with increasing precision, assisting doctors in the early detection of diseases and in the decision-making process. Additionally, advanced statistical analysis is providing valuable insights into health patterns, enabling a more personalized and effective approach to patient treatment. Despite the benefits, clinicians face significant practical challenges in adopting AI technologies. Many report difficulties in integrating these systems into daily routines, and there is a pressing need for skill enhancement and better connectivity with the technology, necessitating additional training to accurately interpret the results generated by AI. Understanding and overcoming these challenges is crucial for maximizing the benefits of emerging technologies and ensuring successful adoption and a positive impact on clinical practice. However, we are witnessing that, while on one hand AI and its derivatives (machine learning, deep learning, and neural networks) may potentially create health benefits, on the other hand, there is an urgent need for regulatory legislation, as unethical and fraudulent misuse is also on the rise. In this context, we are seeing that government positions are still far from being up-to-date in regulating such AI technologies. Ethical and privacy implications, for example, are crucial considerations in the implementation of AI in medicine. Protecting patient data is essential to maintaining trust and integrity. Today, AI technologies must be developed with stringent data security protocols to ensure that patient information is handled with the utmost care and respect. In this evolving scenario, it may not be surprising that a high-income country like the United Kingdom does not yet have specific legislation on the use of software and AI in medicine rather than a changed program roadmap producing documents that will serve as a consultation to arrive in the future to a secondary legislation. Currently, the United Kingdom government's Medicines and Healthcare products Regulatory Agency (MHRA), which regulates medicines, medical devices and blood components, has only released the UK Medical Devices Regulations 2002 and the Data Protection Act 2018.5 The recently approved European Union Regulation 1689/2024 (EU AI Act) is the first example of international political legislation published with the intention to normalize the use of the AI in the healthcare system.6 Starting in January, my first action as Editor-in-Chief was to launch a new journal section dedicated to AI, for which thanks to the invaluable dedication of our Associate Editor (V.A.L.J.), we can now count on a team with five statisticians acting as reviewers of manuscripts submitted in this field. Our journal's position is at the forefront of this transformation, using a multidisciplinary approach that brings together experts in AI, statistics, and medicine. Our editorial focus is clear: to investigate and highlight the impact of these technologies on the development of new solutions for clinical ultrasound and other imaging modalities, validating their contributions in both medical practice and patient experience. Collaboration among AI professionals, statisticians, and healthcare practitioners is essential for advancing applications in medicine. Establishing regulations, platforms, and initiatives that facilitate this collaboration can accelerate the development of innovative solutions and ensure that they meet real clinical needs and impact patients' lives. The synergy between these disciplines promotes a holistic approach that can maximize the positive impact of AI in healthcare. In our pages, readers will find a variety of articles ranging from case studies to innovative research that describe analyses of trends and challenges in the use of AI and statistics in clinical ultrasound. We are committed not only to highlight the technological advancements but also to rigorously evaluating the impact of these solutions in daily clinical practice, considering the ethical rigor required in medicine. Our goal is to provide a valuable resource for doctors, researchers, and the healthcare community in general (providers and healthcare professionals), helping them to “navigate” and introduce these innovations with confidence and competence in their daily practice. Education and training programs are fundamental for empowering physicians and healthcare professionals to use AI effectively. Integrating AI and data analysis education into medical curricula can prepare the next generation of professionals to use these tools competently, thereby improving patient care and operational efficiency. However, academic, research, and publishing activities aimed at disseminating scientific knowledge is not immune to cyber-attacks and/or fraudulent misconduct. We cannot neglect the duty to alert the scientific community as well as publishers that our journal, JCU, as has happened with others, has faced and managed a dramatic editorial attack. Fortunately, thanks to the highly qualified Wiley Ethics Committee and the watchful attention of all our Editorial Board Members (EBMs), the attack was counteracted by identifying and rejecting almost 170 manuscripts submitted over the last 6 months. This editorial has been written with the specific purpose to address and raise some important and consequential questions; that is, publishers are called upon to deploy robust technologies, and more importantly firm editorial decisions, to regulate the use/misuse of AI platforms and specifically those produced to deliver unrealistic and unproven self-automated generated text and manuscripts. At JCU, we aim to foster a continuous and constructive dialogue among the various disciplines involved, promoting collaborations that will lead to even more significant discoveries, contributing to a future where technology and medicine come together to improve the health and well-being of people. We would like to thank in advance all the contributors, AEs, EBMs, readers, and publishers and their ethics committees for their continuous support and invite each of them to notify the journal's EBMs of any potential case or situation they should come to know or even suspect. Our effort is to actively contribute together to link technological innovation to medical practice, creating solutions that have the potential to transform the lives of human beings. But it is also time to update the old aphorism “publish or perish” to the more correct one for the present day, “publish with ethics.”

Ähnliche Arbeiten