Open research questions in Artificial Intelligence in Healthcare and Education

What the literature leaves open

• A primary limitation of this study is the lack of a ‘live’ patient cohort for validation; however, since the study was conducted in a non-English speaking region (Turkey), testing local participants on English-language AI out- puts would have introduced significant linguistic bias. Our findings are limited to English-language outputs generated by free public versions; they cannot be generalized to other languages or paid/professional versions of these models.

  A comparative analysis of readability, quality, and reliability in large language model outputs pertaining to knee osteoarthritis queries · 2026 · DOI

• Promising areas such as personalisation, teaching AI, and engagement have lost visibility despite their long-term importance, while risks and ethical issues remain underexplored and poorly connected to technical research.

  Mapping Three and a Half Decades of AI and Education Literature: Trends, Gaps, and Future Directions · 2026 · DOI

• AI tools need to be incorporated within the curriculum for nursing undergraduates through an ethical approach. Nursing schools need to incorporate AI literacy courses that help their students understand responsible and ethical uses of AI and critical analysis of AI-derived data. Nursing educators need to give proper guidance and oversight on the use of AI-based learning technologies in nursing education. Moreover, interdisciplinary teamwork among nursing departments, computer science, and healthcare informatics may be helpful for a better understanding of AI technology by nursing students. More research is needed on the long-term effects of AI technology in nursing education.

  The Role of Artificial Intelligence in Shaping Critical Thinking and Decision-Making Skills Among Undergraduate Nursing Students · 2026 · DOI

• intelligence: A artificial developing nation’s context. BMC Medical Education, 1060. https://doi.org/10.1186/s12909-025-07223-6 Tong, X., Hu, Y., Long, Y., et al. (2025). The application of problem-based learning (PBL) guided by ChatGPT in clinical education in the of nephrology. BMC Medical department Education, 1048. https://doi.org/10.1186/s12909-025-07427-w Tucker, F. (2025). Doing philosophy and the future of the “good doctor” paradigm. Medicine, Health Care and Philosophy, 28(4), 669–677. https://doi.org/10.1007/s11019-025-10294-3 25(1), Contemporary Education and Teaching Research Vol. 7 Iss. 2 2026 Turner, L., Kelleher, M., Overla, S., et al. (2025). Harnessing the generative power of AI to move education. closer personalized medical Academic Medicine, 1447–1451. https://doi.org/10.1097/ACM.000000000000618 5 100(12), to Ugoala, O., Ebubechukwu, U., Mares, A. C., et al. (2025). Visual in cardiology: Past, present, and future.

  Transformation and Reinvention: A Comprehensive Analysis of Frontiers and Trends in AI-Empowered Medical Education by 2025 · 2026 · DOI

• imaging diagnostics, novel drug development, and precision medicine (Maity & Saikia, 2025; Yu et al., 2025; Mizna et al., 2025). Transformations at the level of diagnostic and therapeutic technologies inevitably necessitate corresponding adjustments in medical education models (Khakpaki, 2025; Cheng & Zhu, 2025). In the face of a highly intelligent healthcare medical education is under considerable pressure with regard to the pace of knowledge renewal, the specificity of skills training, and the adaptability of its pedagogical philosophy (Tucker, 2025; Miguez-Pinto et al., 2025). environment, traditional To ensure that future healthcare professionals can safely, ethically, and effectively utilize AI tools, arising from technological the current medical education system must undergo systematic reform (Cho Kwan et al., 2025). Such reform requires not only innovation in curriculum design and assessment strategies, but also proactive engagement with the ethical dilemmas and practical barriers integration (Boscardin et al., 2025). Encouragingly, preliminary achievements have emerged in AI-enabled medical education. Applications such as adaptive learning pathway simulation immersive are training, gradually being implemented (Turner et al., 2025; Seneviratne & Manathunga, 2025; Wang et al., 2025). design, and automated feedback systems virtual However, during the process of technological deployment, substantial heterogeneity persists among faculty and students with respect to AI literacy, acceptance, and readiness, warranting objective evaluation (Clement David-Olawade et al., 2025; Yazdi et al., 2025). Moreover, algorithmic bias, risks to academic integrity, data privacy concerns, and regulatory lag constitute major bottlenecks limiting equitable and widespread adoption (Stern et al., 2025; Liu et al., 2025; Sun et al., 2025). Against this issue, 41 Corresponding Author: Mingzhe Li The First Affiliated Hospital (The First Clinical Medical School) of Guangdong Pharmaceutical University, P.R. China ©The Author(s) 2026. Published by BONI FUTURE DIGITAL PUBLISHING CO.,LIMITED. This is an open access article under the CC BY License(https://creativecommons.org/licenses/by/4.0/) Contemporary Education and Teaching Research Vol. 7 Iss. 2 2026 the present review synthesizes studies published throughout 2025 to examine the current state of AI literacy among medical students and educators, analyze the effectiveness of emerging application models, and critically evaluate the associated ethical and implementation challenges. The ultimate aim is to provide theoretical grounding and practical guidance for constructing a future-oriented, human– AI collaborative framework in medical education.

  Transformation and Reinvention: A Comprehensive Analysis of Frontiers and Trends in AI-Empowered Medical Education by 2025 · 2026 · DOI

• By building on the synthesised framework, this review proposes a series of integrated recommendations to advance radiography education. Theoretically, the authors advocate for a pedagogical shift towards curricula in which technology integration (AI, virtual reality, data analytics) is not a standalone module but is embedded in innovative teaching methods such as simulation and problem-based learning, to create a symbiotic relationship between how students learn and what they learn. Practically, this necessitates developing flexible, modular curricular architectures that can rapidly adapt to new technologies and evolving professional scopes. A crucial new approach is the intentional design of ethics by design modules, which directly intertwine lessons on AI algorithm operation with ethical principles of patient data governance and cultural inclusivity, to ensure that ethical practice is an applied skill, not an abstract concept. The ethics by design approach moves beyond teaching ethics as a discrete, abstract topic and instead systematically embeds ethical considerations into the very fabric of technological and clinical education. The goal is to produce graduates who do not merely understand ethical principles but are equipped to apply them proactively in complex, real-world scenarios, particularly those involving advanced technologies. Ethics by design modules can be operationalised by directly intertwining technical instruction with ethical deliberation. For instance: • • In AI and machine learning modules: Lessons on how AI algorithms operate for image analysis should be concurrently paired with critical discussions on algorithmic bias, data fairness, and accountability. Students should evaluate case studies where training data lacks diversity, leading to diagnostic disparities across different patient demographics. In data management and analytics training: While learning to handle and secure medical imaging data, the curriculum must integrate rigorous exercises on patient data governance, informed consent for data use in research, and the implications of data breaches, to ensure that privacy is a default practice. 44 Sioux McKenna and Susan van Schalkwyk, “A Scoping Review of the Changing Landscape of Doctoral Education,” Compare: A Journal of Comparative and International Education 54, no. 6 (August 17, 2024): 984–1001, https://doi.org/10.1080/03057925.2023.2168121.

  Framing a future-ready radiography curriculum – A scoping review · 2026 · DOI

• In addition, future research should examine the interaction between AI-based chatbots and healthcare professionals, with particular attention to accountability, trust, and the long-term implications for clinical workflows. Second, access to detailed institutional-level waiting list data remains limited, restricting the precision of regional comparisons.

  Artificial Intelligence and Chatbots in Patient Management · 2026 · DOI

• Learning Health System frameworks [12] specify AI lifecycle management but do not define how continuous assurance loops should incorporate feedback from adaptive architectural performance (e.g., space utilization, patient flow efficiency, environmental responsiveness) to inform iterative AI and design improvements.

  The <scp>HALO</scp> Model: A Learning Health System Framework for Artificial Intelligence · 2026 · DOI

• Deployment challenges for AI in primary care [11] focus on data, technical, and user dimensions but do not address how adaptive architectural design (e.g., flexible clinic layouts, modular equipment placement) could mitigate these barriers or how agentic systems should coordinate with physical space constraints in resource-limited settings.

  A Review of Artificial Intelligence Applications in Healthcare: Clinical Value and Primary Care Adaptation Challenges · 2026 · DOI

• Hernan Inojosa 1,2, Lars Masanneck3, Isabel Voigt1,2, Dirk Schriefer Nele von Horsten1, Judith Wenk1,2, Iva Gasparovic-Curtini1,2, Rocco Haase1,2, Sven G. Meuth3, Hagen B.

  Exploring attitudes and acceptance of artificial intelligence in multiple sclerosis from the patient perspective · 2026 · DOI

• Artificial intelligence (AI) is increasingly integrated into healthcare delivery, yet patient acceptance in resource constrained settings remains incompletely characterized.

  Patient attitudes toward artificial intelligence in Jordanian Healthcare: A cross-sectional survey study · 2026 · DOI

• In response, health systems are increasingly adopting large language models (LLMs) to generate draft replies to patient messages, yet little is known about how patients interpret and evaluate artificial intelligence (AI) involvement in this communication channel.

  Patient Perspectives on AI-Drafted Electronic Portal Messages · 2026 · DOI

• Studies in the literature have showed mixed results of GenAI's impact on students' performance in HPE, with some reporting improved exam outcomes (Hsu, 2023; Roganović, 2024) while others found better results through traditional methods (Saravia-Rojas et al.

  Using GenAI for Objective Structured Clinical Examination (OSCE) Preparation: A Retrospective Study in Australia and Malaysia · 2026 · DOI

• The future of libraries is likely to involve greater collaboration between AI technologies and human expertise. Future developments may include: • Fully integrated AI-supported library systems • Smart digital repositories • Advanced multilingual assistance • Personalized information recommendation systems • Greater automation in cataloging and indexing Balanced implementation and ethical regulation will remain essential for sustainable AI adoption. © 2026 The Author(s). Published by IJCOPE Journal.

  Chatgpt and its Impact on Library users: A Comprehensive Study · 2026 · DOI

• In this context, there are ongoing debates regarding the application of human milk-based fortifiers (HMF) versus bovine milk-based fortifiers (BMF), but robust evidence is lacking.

  Cross-LLM AI platform meta-research: Non-inferiority of bovine milk-based fortifiers to human milk-based fortifiers · 2026 · DOI

• 71, Holmes, W., Bialik, M., & Fadel, C. (2019). Artificial intelligence in education: Promises and implications for teaching and learning. Center for Curriculum Redesign. OpenAI. (2023). GPT-4 technical report. arXiv. https://doi.org/10.48550/arXiv.2303.08774 Rajkomar, A., Hardt, M., Howell, M. D., Corrado, G., & Chin, M. H. (2023). Ensuring fairness in machine learning to advance health equity. Annals of Internal Medicine, 178(6), 866–872. https://doi.org/10.7326/M23-0039 Rogers, E. M. (2003). Diffusion of innovations (5th ed.). Free Press. Russell, S., & Norvig, P. (2021). Artificial intelligence: A modern approach (4th ed.). Pearson. Schwab, K. (2017). The revolution. Crown Business. fourth industrial Topol, E. (2023). Artificial intelligence and the future of medicine: Current trends and opportunities. The Lancet Digital Health, 5(6), e356–e364. https://doi.org/10.1016/S2589- 7500(23)00096-1 UNESCO. (2021). Recommendation on the ethics of artificial intelligence. UNESCO Publishing.

  Applications of Artificial Intelligence in Solving Real-World Problems Across Disciplines · 2026 · DOI

• Given the expanding role of AI in healthcare, fur- ther exploration of how structured curricular integra- tion of AI literacy impacts student anxiety and their behavioral intentions to adopt AI in future clinical roles is warranted.

  Perceptions; attitudes and anxiety toward artificial intelligence among medical students: A cross-sectional study · 2026 · DOI

• While further acknowledged stratification between web-enabled and static models is warranted, and future studies should re-evaluate LLM performance using updated versions, many of which now include web search features.

  Evaluating AI Chatbots for Pediatric Contact Lenses: A Study on Accuracy, Readability, and Reliability · 2026 · DOI

• This study has several limitations that should be acknowledged. First, artificial intelligence–based language models exhibit dynamic behav- ior due to periodic updates, which may result in different responses to the same question over time. Although each question was posed to the model under zero-state conditions at different time points, the findings reflect the performance of ChatGPT-5.2 only within the spe- cific period during which the study was conducted. Second, the assessment was performed exclusively by physicians, and patients were not included in the evaluation process. Patient edu- cation and communication in healthcare could be further strengthened if assessments of comprehensibility, usability, and applicability were conducted within a patient-centered context. Studies incorporating pa- tient-based evaluations would allow a more accurate determination of the real-world impact of AI-generated medical information. Although the mean scores across all evaluated domains were high, some domains demonstrated low to moderate inter-rater reli- ability. This may be attributed to the inherently subjective nature of certain evaluation criteria. In particular, assessments of accuracy and 76 Sezen. AI-Based Patient Guidance in Influenza Journal of Izmir Chest Hospital 2025;39(3):73–77 relevance depend on clinical judgment and context, allowing variabil- ity in evaluators’ interpretations and rating approaches. In addition, the study was conducted solely using the ChatGPT-5.2 model and was limited to influenza-related questions categorized as frequently asked questions based on WHO guidance. While the use of WHO-based questions enhances consistency, information-seek- ing behaviors, patient attitudes, and expectations may vary across different geographical regions. Evaluating other large language mod- els, expanding the range of questions, or focusing on different medi- cal topics may yield different results. CONCLUSION The findings of this study demonstrate that ChatGPT-5.2 was suc- cessful in generating high-quality responses to frequently asked questions about influenza. The results highlight the potential of the ChatGPT-5.2 model to support educational communication in health- care, particularly in the context of infectious diseases such as influen- za. However, variability in inter-rater reliability suggests the presence of subjective elements in the evaluation of AI-generated medical con- tent. Nevertheless, this study supports the effectiveness of AI-assist- ed tools in patient education. Although AI-based patient education tools show promise, they should be implemented as complementary resources in healthcare settings and not as substitutes for profes- sional medical expertise.

  İnfluenza ile İlgili Sıkça Sorulan Sorulara Yapay Zekâ Tarafından Üretilen Yanıtların Doğruluğu ve Hasta Yönlendirmesi · 2026 · DOI

• Each model was queried once per paragraph. Because large language models employ stochastic generation, outputs may vary across repeated identical prompts. Our results therefore represent a snapshot of model behavior and do not assess test-retest consistency. Future studies using repeated queries could further characterize intra-model variability. This work has several important limitations that frame it as a pilot investigation. Firstly, the scope was intentionally limited to 35 paragraphs on anterior segment diseases, to enable a detailed analysis. This necessarily excludes other major ophthalmic subspecialties (e.g., retina, glaucoma, pediatrics). Future studies should expand to these areas, in order to test the generalizability of these findings and identify subspecialty-specific performance patterns. Secondly, the methodological design tested a single, simple prompt without real-time database access. This was a deliberate choice to assess baseline knowledge, but does not reflect more advanced use cases involving retrieval-augmented generation (RAG). Subsequent research should evaluate how performance changes with optimized prompts, role-playing instructions, or live PubMed integration. The prompt’s instruction was recognized as a simulation of a user query, not a technical capability of the models. Importantly, RAG-enabled systems are increasingly used in practice and may yield different citation accuracy profiles. For this reason, these findings should not be extrapolated to workflows that include live retrieval. Thirdly, the rapid evolution of AI technology itself is a fundamental constraint. The models evaluated (ChatGPT GPT-5.1, Copilot 4.2, DeepSeek-R1, Gemini Ultra 2.5) are subject to continuous updates by their developers. Our findings are strictly tied to the model versions, training data cut-offs, and capabilities as of our access period (November-December 2025). Performance metrics, error rates, and even relative rankings could shift with subsequent releases. This underscores the exploratory nature of our study and highlights the critical need for continuous, longitudinal, and version-specific benchmarking of AI tools in academic settings. Finally, the statistical comparison focused on accuracy proportions and error category distributions. More granular analyses, such as examining agreement between models on specific citations or correlating error rates with paragraph complexity or publication date, were beyond the scope of this initial evaluation, but represent valuable avenues for future research. CZECH AND SLOVAK OPHTHALMOLOGY AOP 2026 CONCLUSION This exploratory pilot study provides a snapshot evaluation of four contemporary AI models (late-2025) for citation generation in anterior segment ophthalmology.

  Evaluation of AI Citation Accuracy in Anterior Segment Research · 2026 · DOI

• Introduction: Generative artificial intelligence (AI) can produce realistic clinical scenarios on demand and deliver immediate, individualized feedback, yet its use to teach ethical reasoning, rather than to address the ethics of AI itself, remains underexplored in interprofessional healthcare education.

  Students' Perceptions of an AI-Enhanced Ethics Learning Platform: A Pilot Study on Interprofessional Healthcare Education · 2026 · DOI

• This systematic review has several methodological strengths. The study adhered to a pre-specified protocol reg- istered in PROSPERO and followed PRISMA guidelines and the PICO framework, ensuring transparency, reproduc- ibility, and a structured approach to study selection. Data extraction was guided by the CHARMS checklist, and two reviewers independently screened titles, abstracts, and full texts, with a third reviewer resolving any discrepancies. The search strategy included multiple major databases, increas- ing the likelihood of capturing all relevant studies. Another key strength of this review is its specific focus on ML mod- els developed solely using administrative health data. By focusing on these models, we provide insights into tools that are inherently easier to scale and integrate into routine clinical workflows, without requiring manual data entry or additional interventions for the patient. Furthermore, the review highlights the emerging use of interpretability meth- ods, such as SHAP values, which are essential for mitigat- ing clinician distrust in “black box” models. The review is also subject to limitations. The small num- ber of eligible models and the substantial heterogeneity in study populations, feature sets and outcome definitions, Page 11 of 14 103 precluded the possibility of a meta-analysis. Direct per- formance comparisons between studies must therefore be interpreted with some caution. Furthermore, while the use of administrative data could enable high scalability, we must also recognize that such data are typically collected for bill- ing purposes and not research [51]. Contrary or inaccurate evidence can easily arise when comparing findings from such databases across countries and time periods, as local coding practices, healthcare policies and diagnostic criteria heavily influence the underlying data structure and thus any models derived from such data [52]. Therefore, any perfor- mance metrics from an automated model that relies on auto- matically extracted administrative data, must be weighed against the risk of significant discrepancies between this data and clinical reality. Despite these limitations, this review provides a com- prehensive overview of current ML-based fracture risk prediction using administrative data and highlights method- ological considerations and evidence that can inform future research and potential clinical implementation of models in this context.

  Artificial Intelligence Approaches for Osteoporotic Fracture Risk Prediction Using Administrative Health Data: A Systematic Review · 2026 · DOI

• Data collection was limited by the convenience sample of students from one institution that agreed to participate in the study, which may limit ability to generalise findings. Self-selection bias is a risk, given it is possible that students who had previous experience with or held stronger views about AI (positively or negatively) may have been more likely to show interest in participating in this study. To mitigate sampling bias, we aimed for a diverse sample in each focus group in terms of gender, age, discipline and year of study. Future comparative research with students from different institutions and countries may add to our findings. In this study, students were not presented with an AI-driven VP to interact with outside of their own experiences. In the future, student sentiments may change with differing levels of exposure, so further research with various prototypes may be beneficial.

  Healthcare students’ perspectives on artificial intelligence-driven virtual patients for learning communication skills · 2026 · DOI

• Ciona Dewan1, Raghu Raja Mehra2 1Invictus International School, Amritsar Email: cionadewan025[at]gmail.com 2Invictus International School, Amritsar Email: raghu[at]invictusschool.edu.in Abstract: Early and accurate detection of disease is one of the most decisive factors in patient survival, treatment cost and quality of life. Artificial intelligence (AI), and in particular machine learning and deep learning, has emerged as a powerful ally in this effort, capable of analysing medical images, electronic health records, laboratory results and wearable-sensor data with remarkable speed and consistency. This paper reviews the role of AI in early disease detection, surveying the principal techniques, the typical detection pipeline, and applications across cancer, cardiovascular, ophthalmic and neurological disorders. A comparison with conventional diagnostic methods shows that AI systems can match or exceed clinician-level accuracy in several screening tasks while operating at scale. The paper then proposes an integrated, privacy-preserving and explainable framework for clinical deployment, and critically examines the advantages, limitations, and ethical and regulatory challenges involved. Finally, it outlines future directions—including federated learning, explainable AI and continuous wearable monitoring—that could make trustworthy, equitable early detection a routine part of care.

  The Role of Artificial Intelligence in Early Disease Detection Techniques, Applications, Challenges and Future Directions · 2026 · DOI

• This editorial is primarily a normative and conceptual analysis rather than an empirical investigation. Consequently, the arguments presented are not supported by prospective clinical outcome data, patient- 2026 Lourdunathan et al. Cureus 18(6): e111430. DOI 10.7759/cureus.111430 2 of 6 reported measures, or formal assessments of clinician attitudes toward algorithmic governance. Future studies should evaluate whether the proposed principles influence patient trust, informed consent quality, shared decision-making, or clinical outcomes. Furthermore, although Magnifica Humanitas provides a coherent ethical framework grounded in Catholic social teaching, alternative secular, human rights-based, and pluralistic bioethical traditions may arrive at similar conclusions regarding patient autonomy, transparency, accountability, and protection from exploitation. The applicability of this framework may therefore vary across healthcare systems, cultural settings, and legal environments [2]. The discussion of health data colonialism intentionally highlights potential risks associated with unequal data governance and commercialization of health information. However, international data sharing has also generated substantial scientific, educational, and public health benefits. Future scholarship should focus on identifying governance structures that preserve innovation while ensuring transparency, fairness, and equitable distribution of benefits. Finally, the rapid evolution of AI technologies, privacy regulations, and data governance frameworks may alter contemporary understandings of algorithmic transparency, data ownership, and digital sovereignty. As a result, the ethical recommendations proposed in this editorial should be viewed as adaptive principles requiring periodic reassessment as technology and regulatory environments continue to evolve [4,5]. Conclusion: embracing the theological standard As the boundaries of surgical innovation continue to expand through robotics, machine learning, and advanced digital platforms, clinicians must ensure that technological progress remains aligned with the fundamental purpose of medicine: the care of persons. Questions surrounding algorithmic influence, data governance, and informed consent are no longer theoretical concerns but practical challenges that increasingly affect everyday surgical practice [1]. Regardless of one's religious or philosophical commitments, the emergence of surgical AI necessitates renewed attention to patient dignity, transparency, accountability, and meaningful informed consent. Magnifica Humanitas contributes a distinctive perspective to this discussion by emphasizing that technological innovation should remain accountable to the human person whom it is intended to serve [3]. By fostering dialogue on digital consent and responsible stewardship of patient data, the surgical community can help ensure that technological advancement enhances, rather than diminishes, the humanity at the center of clinical care.

  The Surgeon and the Algorithm: Why the Vatican’s Bioethical Blueprint Is Essential for Modern Surgical Practice · 2026 · DOI

Most-cited papers in Artificial Intelligence in Healthcare and Education

• TRIPOD+AI statement: updated guidance for reporting clinical prediction models that use regression or machine learning methods · BMJ · 2024 · 2,289 citations
• Resistance to Medical Artificial Intelligence · Journal of Consumer Research · 2019 · 1,546 citations
• The Role of AI in Hospitals and Clinics: Transforming Healthcare in the 21st Century · Bioengineering · 2024 · 742 citations
• The Role of ChatGPT, Generative Language Models, and Artificial Intelligence in Medical Education: A Conversation With ChatGPT and a Call for Papers · JMIR Medical Education · 2023 · 713 citations
• <scp>ChatGPT</scp> and a new academic reality: <scp>Artificial Intelligence‐written</scp> research papers and the ethics of the large language models in scholarly publishing · Journal of the Association for Information Science and Technology · 2023 · 712 citations
• Large Language Models in Medical Education: Opportunities, Challenges, and Future Directions · JMIR Medical Education · 2023 · 589 citations
• Evaluation and mitigation of the limitations of large language models in clinical decision-making · Nature Medicine · 2024 · 572 citations
• The rise of <scp>ChatGPT</scp>: Exploring its potential in medical education · Anatomical Sciences Education · 2023 · 519 citations
• A systematic literature review of artificial intelligence in the healthcare sector: Benefits, challenges, methodologies, and functionalities · Journal of Innovation & Knowledge · 2023 · 464 citations
• Collaborating with ChatGPT in argumentative writing classrooms · Assessing Writing · 2023 · 450 citations

Most recent work

• More Versus Better: Artificial Intelligence, Incentives, and the Emerging Crisis in Peer Review · Organization Science · 2026
• Artificial Intelligence in Medical Education: Transformative Potential, Current Applications, and Future Implications · JMIR Medical Education · 2026
• Ambient Artificial Intelligence Scribe Adoption and Documentation Time in the Emergency Department · Annals of Emergency Medicine · 2026
• Transforming evidence synthesis: A systematic review of the evolution of automated meta-analysis in the age of AI · Research Synthesis Methods · 2026
• ChatGPT Psychosis: A Love Story — Project Prospectus and Formal Design Specification for an AI-Native Glyphic Novel (v0.3) · Zenodo (CERN European Organization for Nuclear Research) · 2026
• Evidence of Unreliable Data and Poor Data Provenance in Clinical Prediction Model Research and Clinical Practice · medRxiv · 2026
• Artificial Intelligence and Big Data for Precision Medicine: A Review of Bioinformatics-Driven Healthcare Applications · Frontiers in Computer Science and Artificial Intelligence · 2026
• SΔϕ-60 — Utilitarian Subject-Splitting Index: Aggregate Utility, Greater-Good Claims, Sacrifice Capture, and Repair Burden (v1.0, AI-Readable Package) · Zenodo (CERN European Organization for Nuclear Research) · 2026
• Performance Evaluation of Large Language Models in Multilingual Medical Multiple-Choice Questions: Mixed Methods Study · JMIR Medical Education · 2026
• Examining the Mediating Roles of Digital Literacy and Critical Thinking in the Relationship between ChatGPT Use and University Students’ Mathematical Problem-Solving Efficacy · Journal of Mathematics, Science and Technology Education · 2026

Related topics in Medicine