arXiv subjects

Sally L. Baxter

Publications and source records attributed to Sally L. Baxter.

4 recordsLinked to original sources

arXiv2026

An Agentic AI Framework Overcomes Fundamental Limitations of Large Language Models for Glaucoma Detection from Fundus Photography

Large language models (LLMs) show promise in medical image interpretation but suffer from hallucination, limited accuracy, and run-to-run inconsistency. We developed and validated an agentic AI framework integrating LLMs with specialized deep learning tools for glaucoma detection from fundus photography. The workflow had three steps: (1) LLM initial assessment; (2) function calling to invoke specialized tools for image quality (QAModel, FundaQ-8), glaucoma classification (SwinV2-Tiny), and optic disc/cup segmentation (SegFormer-B0); and (3) LLM reflection integrating the initial impression with tool outputs. Two LLMs (Gemini 2.5 Flash, GPT-5.4 mini) were evaluated on two public datasets (ORIGA, n=100; RIM-ONE-v3, n=100) under uncropped and cropped fields of view; all images were independently graded by a masked fellowship-trained glaucoma specialist. The agentic workflow improved classification accuracy by 16 to 47 percentage points across all conditions, reaching within 6 points of the specialist; on RIM-ONE-v3 the best configurations matched the specialist accuracy of 88%. LLM-alone approaches failed in two ways: GPT-5.4 mini showed positive bias (sensitivity 95-100%, specificity 0-5%), while Gemini 2.5 Flash varied stochastically between runs; the agentic workflow corrected both. Cup-to-disc ratio error fell 15-50% (MAE 0.156-0.228 to 0.104-0.132), and correlation with specialist grading rose from weak (r=0.12-0.39) to moderate-strong (r=0.59-0.84). Run-to-run consistency rose from near-random (kappa as low as -0.01) to near-perfect (kappa up to 0.96). Integrating LLMs with specialized tools addressed key limitations of LLM-alone approaches, including over-diagnosis and run-to-run variability. Gains held for both LLMs, suggesting generalizability across backbones, and may signal a shift from monolithic models toward orchestrated multi-agent systems in medical AI.

cs.AI↗

arXiv2025

Glaucoma Detection and Structured OCT Report Generation via a Fine-tuned Multimodal Large Language Model

Objective: To develop an explainable multimodal large language model (MM-LLM) that (1) screens optic nerve head (ONH) OCT circle scans for quality and (2) generates structured clinical reports that include glaucoma diagnosis and sector-wise retinal nerve fiber layer (RNFL) thinning assessments. Design: Retrospective cohort study of 1,310 subjects contributing 43,849 Spectralis ONH OCT circle scans (1,331 glaucomatous and 867 healthy eyes) from the DIGS and ADAGES cohorts. Methods: A MM-LLM (Llama 3.2 Vision-Instruct model) was fine-tuned to generate clinical descriptions of OCT imaging data. Training data included paired OCT images and automatically generated, structured clinical reports that described global and sectoral RNFL thinning. Poor-quality scans were labeled as unusable and paired with a fixed refusal statement. The model was evaluated on a held-out test set for three tasks: quality assessment, glaucoma detection, and RNFL thinning classification across seven anatomical sectors. Evaluation metrics included accuracy, sensitivity, specificity, precision, and F1-score. Model description quality was also evaluated using standard text evaluation metrics. Results: The model achieved 0.90 accuracy and 0.98 specificity for quality triage. For glaucoma detection, accuracy was 0.86 (sensitivity 0.91, specificity 0.73, F1-score 0.91). RNFL thinning prediction accuracy ranged from 0.83 to 0.94, with highest performance in global and temporal sectors. Text generation scores showed strong alignment with reference reports (BLEU: 0.82; ROUGE-1: 0.94; ROUGE-2: 0.87; ROUGE-L: 0.92; BERTScore-F1: 0.99). Conclusions: The fine-tuned MM-LLM generated accurate clinical descriptions based on OCT imaging. The model achieved high accuracy in identifying image quality issues and detecting glaucoma. The model also provided sectoral descriptions of RNFL thinning to help support clinical OCT evaluation.

q-bio.QM↗

arXiv2025

Open Data Sharing in Clinical Research and Participants Privacy: Challenges and Opportunities in the Era of Artificial Intelligence

Sharing clinical research data is key for increasing the pace of medical discoveries that improve human health. However, concern about study participants' privacy, confidentiality, and safety is a major factor that deters researchers from openly sharing clinical data, even after deidentification. This concern is further heightened by the evolution of artificial intelligence (AI) approaches that pose an ever-increasing threat to the reidentification of study participants. Here, we discuss the challenges AI approaches create that blur the lines between identifiable and non-identifiable data. We present a concept of pseudo-reidentification, and discuss how these challenges provide opportunities for rethinking open data sharing practices in clinical research. We highlight the novel open data sharing approach we have established as part of the Artificial Intelligence Ready and Exploratory Atlas for Diabetes Insights project, one of the four Data Generation Projects funded by the National Institutes of Health Common Fund's Bridge2AI Program.

q-bio.OT↗

arXiv2024

Hidden flaws behind expert-level accuracy of multimodal GPT-4 vision in medicine

Recent studies indicate that Generative Pre-trained Transformer 4 with Vision (GPT-4V) outperforms human physicians in medical challenge tasks. However, these evaluations primarily focused on the accuracy of multi-choice questions alone. Our study extends the current scope by conducting a comprehensive analysis of GPT-4V's rationales of image comprehension, recall of medical knowledge, and step-by-step multimodal reasoning when solving New England Journal of Medicine (NEJM) Image Challenges - an imaging quiz designed to test the knowledge and diagnostic capabilities of medical professionals. Evaluation results confirmed that GPT-4V performs comparatively to human physicians regarding multi-choice accuracy (81.6% vs. 77.8%). GPT-4V also performs well in cases where physicians incorrectly answer, with over 78% accuracy. However, we discovered that GPT-4V frequently presents flawed rationales in cases where it makes the correct final choices (35.5%), most prominent in image comprehension (27.2%). Regardless of GPT-4V's high accuracy in multi-choice questions, our findings emphasize the necessity for further in-depth evaluations of its rationales before integrating such multimodal AI models into clinical workflows.

cs.CV↗