Automated Category and Trend Analysis of Scientific Articles on Ophthalmology Using Large Language Models: Development and Usability Study

BACKGROUND

In this paper, we present an automated method for article classification, leveraging the power of large language models (LLMs).

OBJECTIVE

The aim of this study is to evaluate the applicability of various LLMs based on textual content of scientific ophthalmology papers.

METHODS

We developed a model based on natural language processing techniques, including advanced LLMs, to process and analyze the textual content of scientific papers. Specifically, we used zero-shot learning LLMs and compared Bidirectional and Auto-Regressive Transformers (BART) and its variants with Bidirectional Encoder Representations from Transformers (BERT) and its variants, such as distilBERT, SciBERT, PubmedBERT, and BioBERT. To evaluate the LLMs, we compiled a data set (retinal diseases [RenD] ) of 1000 ocular disease-related articles, which were expertly annotated by a panel of 6 specialists into 19 distinct categories. In addition to the classification of articles, we also performed analysis on different classified groups to find the patterns and trends in the field.

RESULTS

The classification results demonstrate the effectiveness of LLMs in categorizing a large number of ophthalmology papers without human intervention. The model achieved a mean accuracy of 0.86 and a mean F1-score of 0.85 based on the RenD data set.

CONCLUSIONS

The proposed framework achieves notable improvements in both accuracy and efficiency. Its application in the domain of ophthalmology showcases its potential for knowledge organization and retrieval. We performed a trend analysis that enables researchers and clinicians to easily categorize and retrieve relevant papers, saving time and effort in literature review and information gathering as well as identification of emerging scientific trends within different disciplines. Moreover, the extendibility of the model to other scientific fields broadens its impact in facilitating research and trend analysis across diverse disciplines.

Pediatric Corneal Structural Development During Childhood Characterized by Ultrasound Biomicroscopy

PURPOSE

To quantitatively describe the structural corneal changes from infancy to early adulthood using ultrasound biomicroscopy.

METHODS

In this prospective study, 168 ultrasound biomicroscopy images were obtained from 24 healthy eyes of 24 patients who consented and enrolled in the Pediatric Anterior Segment Imaging Innovation Study. Their ages ranged from birth to 26 years. An established ultrasound biomicroscopy imaging protocol including seven views of one eye per patient were obtained and measured using ImageJ software (National Institutes of Health). Twelve corneal structural parameters were measured. Means were compared between younger and older groups.

RESULTS

Among the 12 measured structures, 5 demonstrated statistically significant differences (P < .05) between patients younger than 1 year and patients older than 1 year. The mean values for corneal cross-sectional width and length, central corneal thickness, and radii of curvature (anterior and posterior) were significantly different in patients younger than 1 year. Curvature and limbus-to-limbus dimensions changed more dramatically than thickness and tissue density. When comparing the youngest to oldest subgroups, anterior curvature flattened (6.14 to 7.55 radius), posterior curvature flattened (5.53 to 6.72 radius), angle-to-angle distance increased (8.93 to 11.40 mm), and endothelial cross-sectional distance increased (10.63 to 13.61 mm).

CONCLUSIONS

Pediatric corneal structures change with age. The most significant changes occur in the first months of life, with additional changes later in childhood. This study further demonstrates the importance of age in pediatric corneal imaging analysis. [J Pediatr Ophthalmol Strabismus. 2020;57(4):238-245.].

Microcyclosporella mali: a novel fungal keratitis in a post-penetrating keratoplasty patient

An 86-year-old woman underwent penetrating keratoplasty for pseudophakic bullous keratopathy and presented with an acute corneal ulcer thereafter. Examination demonstrated a fluffy white infiltrate and epithelial defect with subsequent endothelial plaque formation and anterior chamber inflammation. The ulcer was cultured, and fortified topical vancomycin and tobramycin were initiated but failed to significantly improve the clinical course. Cultures were ultimately positive for fungus Microcyclosporella mali that responded well to topical natamycin with stabilisation of the ulcer after 6 weeks of topical therapy. This is the first reported case of fungal keratitis due to M. mali.