AI evidence extraction

Generative artificial intelligence as a research partner in orthopaedics: State of the art

Ricardo Bastos, Américo Dias, Paulo Amado, Philippe Neyret, João Espregueira-Mendes, Sthefan Gabriel Berwanger · Journal of ISAKOS Joint Disorders & Orthopaedic Sports Medicine · 2026

AI-generated evidence extraction, verified across multiple analytical personas. Not a substitute for the peer-reviewed original.

9/10

Relevance

1/4

Quality (LMQS)

Evidence

Citations

0.00

FWCI

This is an AI analysis. Read the peer-reviewed original at the publisher: https://doi.org/10.1016/j.jisako.2026.101126

Methodology & findings

Study design

Narrative review combining PubMed literature search with expert consultation.

Main result

The paper proposes that "a three-phase workflow is proposed: (1) evidence selection using semantic discovery systems to identify and map relevant literature beyond keyword matching; (2) data extraction and synthesis employing RAG-based systems to anchor AI responses to verified PDF sources, thereby minimizing hallucinations; and (3) drafting and refining using large language models (LLMs) for structured composition, linguistic clarity, and iterative manuscript improvement." The authors found that when applied in controlled, evidence-grounded environments, generative AI systems can automate literature synthesis and expedite data extraction while preserving authorial intent.

Reports effect sizes.

Research paradigm

Interpretive/argumentative; technology assessment and practice-oriented

Author conclusions

The authors conclude that "AI represents a paradigm shift in orthopaedic scholarship, functioning as a cognitive exoskeleton that augments rather than replaces human expertise. With vigilant human oversight and adherence to journal ethics, orthopaedic surgeons can leverage AI to enhance research productivity, reproducibility, and quality while upholding the highest standards of scientific integrity."

Risk of bias

Selection bias in literature search (PubMed-only search); Expert selection bias (composition and expertise of interviewed GenAI experts not detailed); Publication bias (published literature may overrepresent successful AI applications); Potential algorithmic bias in GenAI systems themselves; Hallucination risk in AI-generated outputs; Selection bias in PubMed search strategy (not explicitly detailed); Expert selection bias in interactive review process (criteria for expert selection not specified); Potential publication bias inherent in PubMed database searches; Algorithmic bias in generative AI systems discussed conceptually

Limitations

The authors identify that "Risks include algorithmic bias, 'hallucinations', privacy concerns, and ethical issues related to authorship." Additionally, they note that "challenges remain" in implementing these systems, and that while the framework shows promise, "vigilant human oversight and adherence to journal ethics" are essential safeguards.

Open questions raised

Need for better understanding of how to transition from opaque 'black box' generation to grounded, verifiable research assistance
Gap in practical frameworks for orthopaedists to effectively harness AI tools
Need for solutions to mitigate algorithmic bias and hallucinations in AI-generated research content
Ethical guidance needed for authorship attribution when using GenAI in research
The authors identify a need for structured frameworks to help orthopaedists effectively harness AI tools. They emphasize the gap between opaque 'black box' AI generation and grounded, verifiable research assistance. The paper addresses the challenge that "scientific publication remains a vital marker of academic success but is often constrained by clinical workload."
The paper identifies the need for structured frameworks to implement generative AI in orthopaedic research while maintaining scientific integrity. It highlights gaps in balancing AI efficiency with ethical concerns around algorithmic bias, authorship accountability, and hallucination minimization.

Data: not_statedCode: not_statedExtracted from: pdfAgreement 74%

Explore related topics

AI Manuscript Writing Academic Writing Literature Review