Introduction: With the rapid proliferation of artificial intelligence (AI) tools, important questions about their applicability to manuscript preparation have been raised. This study explores the methodological challenges of detecting AI-generated content in neurosurgical publications, using existing detection tools to highlight both the presence of AI content and the fundamental limitations of current detection approaches. Methods: We analyzed 100 randomly selected manuscripts published between 2023 and 2024 in high-impact neurosurgery journals using a two-tiered approach to identify potential AI-generated text. The text was classified as AI-generated if both a robustly optimized bidirectional encoder representations from transformers pretraining approach (RoBERTa)-based AI classification tool yielded a positive classification and the text’s perplexity score was less than 100. Chi-square tests were conducted to assess differences in the prevalence of AI-generated text across various manuscript sections, topics, and types. In an effort to eliminate bias introduced by the more structured nature of abstracts, a subgroup analysis was conducted that excluded abstracts as well. Results: Approximately one in five (20%) manuscripts contained sections flagged as AI-generated. Abstracts and methods sections were disproportionately identified. After excluding abstracts, the association between section type and AI-generated content was no longer statistically significant. Conclusion: Our findings highlight both the increasing integration of AI in manuscript preparation and a critical challenge in academic publishing as AI language models become increasingly sophisticated and traditional detection methods become less reliable. This suggests the need to shift focus from detection to transparency, emphasizing the development of clear disclosure policies and ethical guidelines for AI use in academic writing.
Glioblastoma is the most aggressive primary brain tumor, with poor prognosis following initial identification. Current diagnostic methods, including neuroimaging and molecular pathology, face several limitations in tumor delineation, differentiation of progression from treatment effects, and classification of tumor grade. Artificial intelligence (AI) and machine learning (ML) have been increasingly investigated for its potential in addressing such challenges. This narrative review examines existing AI applications in glioblastoma screening and classification, as well as their associated methodological shortcomings. A comprehensive literature search was conducted in MEDLINE for studies published in the past five years applying ML methods to glioblastoma screening and classification. Studies which were not peer reviewed, did not discuss screening or classification, or lacked a clearly defined ML methodology were excluded. Study designs, training dataset type, and model efficacies were reviewed for narrative evidence synthesis. AI-based omics models frequently applied genomic, transcriptomic, methylation status, and Raman spectroscopy data to glioblastoma classification. Non-omics AI applications frequently involved imaging-based methods, in addition to histopathologic and clinical studies. Accuracies exceeding 90% were observed in several studies for the identification and classification of glioblastoma. Despite this, challenges remain in clinical implementation due to dataset heterogeneity, inconsistent model validation, and lack of standardized reporting methodology. While large language models are an emerging area of interest, few studies investigated their uses in the screening or classification of GBM. AI offers the potential for significant advancements in GBM screening and classification, but widespread clinical adoption requires improved application of existing reporting guidelines. Future research should focus on model interpretability, further development of high-quality datasets, and implementation.