JMIR Publications

ABSTRACT

Background:

Early-stage clinical findings often appear only as conference posters circulated on social media. Because posters rarely carry structured metadata, their citations are invisible to bibliometric and altmetric tools, limiting real-time research discovery.

Objective:

To determine whether a large language model (LLM) can accurately extract citation data from clinical conference poster images shared on X (formerly Twitter) and link those data to the Dimensions and Altmetric databases.

Methods:

Poster images associated with the 2024 American Society of Clinical Oncology (ASCO) meeting were searched using the terms “#ASCO24”, “#ASCO2024”, and the conference name. Images ≥100 kB that contained the word “poster” in the post text were retained. A prompt-engineered Gemini 2.0 Flash model classified images, summarised posters, and extracted structured citation elements (eg, authors, title, DOI) in JSON. A hierarchical linkage algorithm matched extracted elements against Dimensions records, prioritising persistent identifiers, then title–journal–author composites. Manual validation was performed on a random 20% sample.

Results:

We searched within 115714 posts and 16 574 images, of which 651 met inclusion criteria and yielded 1117 potential citations. The algorithm linked 708/1117 citations (63%) to 616 unique research outputs (580 journal articles; 36 clinical trial registrations). Manual review of 135 sampled citations confirmed correct linkage in 124 cases (accuracy = 92%). DOI-based matching was mostly flawless; most errors occurred where only partial bibliographic details were available. The linked dataset enabled rapid profiling of topical foci (eg, lung and breast cancer) and identification of highly cited institutions and clinical trials.

Conclusions:

This study presents a novel AI-driven methodology for enhancing research discovery and attention analysis from non-traditional clinical scholarly outputs. ASCO was used as an example but this could be used in any conference and clinical poster. Clinical Trial: Not applicable