Abstract
Introduction/Objectives: Cancer impacts a significant number of individuals, with approximately 20 million cases reported in 2022, with nearly 10 million cancer-related deaths globally. Investing in cancer research therefore carries immense value, with the goal of better preventing, diagnosing, and effectively treating cases of various types of cancer. Given the novelty of AI in research, it is important to understand its effects on scientific research. Journal policies on the use of AI in publishing research remain ambiguous due to the absence of universal guidelines for its implementation. To our knowledge, no studies have investigated journal guidelines on AI within oncology research. Therefore, our study aimed to assess how AI is addressed in policy and reporting guidelines within the field of oncology. The primary objective of this study was to evaluate how AI is addressed in policy and reporting guidelines within the top 100 oncology journals. The secondary outcome focused on reviewing and summarizing the policies across these journals regarding the use of AI-generated content, including text, images, and writing, as well as the acceptance or prohibition of AI authorship.
Methods: Clinical oncology journals were identified using the SCImago Journal and Country Rank (SJR) database. The top 100 peer-reviewed clinical journals in oncology were evaluated. Discontinued journals and those without accessible author instructions or editorial contact were excluded. Three investigators, ZT, TS, CO, independently extracted data from author instructions using a masked, duplicate approach. For journals mentioning AI in their author instructions, policy details on authorship, manuscript writing, content and image generation, and AI-specific reporting guidelines were extracted. Correlational analyses in R (v4.2.1) and RStudio examined links between AI policies and journal characteristics.
Results: AI was mentioned in 68% of the top 100 oncology journals. Of the journals, 67% allowed for authorship, while the remaining 33% did not have any statement for AI in authorship. Additionally, 64% of journals required authors to disclose the use of AI, with 36% of journals having no statement of AI in authorship. Of the top 100 journals, 56% allowed for AI for manuscript writing (grammar, revising), whereas 44% had no statements; 12% of journals allowed for AI to be used in content generation (i.e., data curation), 43% did not mention, and 45% having no statement. For image generation, 11% of journals allowed for the use of AI, 46% did not, and 43% had no statement.
Conclusion: The majority of the top 100 oncology journals mention the use of AI in their instructions for authors, with most requiring its disclosure by authors and allowing for its use in manuscript writing. Though, most journals did not allow the use of AI for content generation and image generation. The concern for the transparency and the correct use of AI remains strong, as almost one third of the journals did not mention AI in their instructions for authors. Future studies should be conducted to allow for a clearer picture on the influences of AI within oncological research.
Methods: Clinical oncology journals were identified using the SCImago Journal and Country Rank (SJR) database. The top 100 peer-reviewed clinical journals in oncology were evaluated. Discontinued journals and those without accessible author instructions or editorial contact were excluded. Three investigators, ZT, TS, CO, independently extracted data from author instructions using a masked, duplicate approach. For journals mentioning AI in their author instructions, policy details on authorship, manuscript writing, content and image generation, and AI-specific reporting guidelines were extracted. Correlational analyses in R (v4.2.1) and RStudio examined links between AI policies and journal characteristics.
Results: AI was mentioned in 68% of the top 100 oncology journals. Of the journals, 67% allowed for authorship, while the remaining 33% did not have any statement for AI in authorship. Additionally, 64% of journals required authors to disclose the use of AI, with 36% of journals having no statement of AI in authorship. Of the top 100 journals, 56% allowed for AI for manuscript writing (grammar, revising), whereas 44% had no statements; 12% of journals allowed for AI to be used in content generation (i.e., data curation), 43% did not mention, and 45% having no statement. For image generation, 11% of journals allowed for the use of AI, 46% did not, and 43% had no statement.
Conclusion: The majority of the top 100 oncology journals mention the use of AI in their instructions for authors, with most requiring its disclosure by authors and allowing for its use in manuscript writing. Though, most journals did not allow the use of AI for content generation and image generation. The concern for the transparency and the correct use of AI remains strong, as almost one third of the journals did not mention AI in their instructions for authors. Future studies should be conducted to allow for a clearer picture on the influences of AI within oncological research.
| Original language | American English |
|---|---|
| State | Published - 14 Feb 2025 |
| Event | Oklahoma State University Center for Health Sciences Research Week 2025 - Oklahoma State University Center for Health Sciences, Tulsa, United States Duration: 10 Feb 2025 → 14 Feb 2025 https://medicine.okstate.edu/research/research_days.html |
Conference
| Conference | Oklahoma State University Center for Health Sciences Research Week 2025 |
|---|---|
| Country/Territory | United States |
| City | Tulsa |
| Period | 10/02/25 → 14/02/25 |
| Internet address |
Keywords
- Artificial Intelligence
- journal guidelines
- oncology