Reasoning Over Pre-training: Evaluating LLM Performance and Augmentation in Women's Health
Imprialou, M.; Kaltsas, N.; Oliinyk, V.; Vigrass, T.; Schwarzmann, J.; Rosenthal, R.; Glastonbury, C.; Wigley, C.; Gillam, M.; Kanani, N.; Supramaniam, P.; Granne, I.; Lindgren, C. M.
Show abstract
Recent advances in large language models (LLMs) show promise in clinical applications, but their performance in womens health remains underexamined 1. We evaluated LLMs on 2,337 questions from obstetrics and gynaecology, including 1,392 from the Royal College of Obstetricians and Gynaecologists Part 2 examination (MRCOG Part 2) 2, a UK-based test of advanced clinical decision-making, and 945 from MedQA3, a dataset derived from the United States Medical Licensing Examination (USMLE). The best-performing model--OpenAIs o1-preview4 enhanced with retrieval-augmented generation (RAG)5,6--achieved 72.00% accuracy on MRCOG Part 2 and 92.30% on MedQA, exceeding prior benchmarks by 21.6%1. General-purpose reasoning models outperformed domain-specific fine-tuned models such as MED-LM7. We also analyse performance by clinical subdomain and discover lower accuracy in areas like fetal medicine and postpartum care. These findings highlight the importance of reasoning capabilities over domain-specific fine-tuning and demonstrate the value of augmentation methods like RAG for improving accuracy and interpretability8.
Matching journals
The top 2 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Evaluating large language model workflows in clinical decision support: referral, triage, and diagnosis 94%
- A Framework to Assess Clinical Safety and Hallucination Rates of LLMs for Medical Text Summarisation 94%
- Interpretable Fine-tuned Large Language Models Facilitate Making Genetic Test Decisions for Rare Diseases 93%
Similar papers in this journal
- Evaluation of the performance of GPT-3.5 and GPT-4 on the Medical Final Examination 93%
- CONSORT-TM: Text classification models for assessing the completeness of randomized controlled trial publications 93%
- Large Language Models Improve the Identification of Emergency Department Visits for Symptomatic Kidney Stones 93%
Similar papers in this journal
- Performance of Generative Pretrained Transformer on the National Medical Licensing Examination in Japan 93%
- Modular Clinical Decision Support Networks (MoDN)—Updatable, Interpretable, and Portable Predictions for Evolving Clinical Environments 93%
- Cardiology Knowledge Assessment of Retrieval-Augmented Open versus Proprietary Large Language Models 93%
Similar papers in this journal
- Benchmarking transformer-based models for medical record deidentification: A single centre, multi-specialty evaluation 96%
- EYE-Llama, an in-domain large language model for ophthalmology 95%
- A multi-task domain-adapted model to predict chemotherapy response from mutations in recurrently altered cancer genes 91%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.