Human in the Loop: Embedding Medical Expert Input in Large Language Models for Clinical Applications
Bibliographic record
Abstract
The state-of-the-art performance of large language models (LLMs) in medical natural language (NLP) tasks, including medical query answering, summarization of clinical notes, and generation of medical reports has led to the development of a large number of application studies. However, many of these studies have also identified the key role of human input in generating accurate results with significant efforts focused on identifying an effective mechanism to elicit, model, and integrate human medical expertise in optimizing LLMs. In this paper, we introduce a new approach based on biomedical ontologies as a knowledge model to significantly improve the performance of LLMs in biomedical natural language processing (NLP) applications. Specifically, we focus on a rare pediatric epilepsy called Dravet syndrome (DS) which there is very limited understanding about the mechanisms that result in seizure and demonstrate the effectiveness of a unique epilepsy ontology in improving the accuracy of results. The results of this study create a new pathway for integrating human expertise in LLMs to support high accuracy and consistent results in medical applications.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.005 | 0.024 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.003 |
| Open science | 0.001 | 0.002 |
| Research integrity | 0.002 | 0.001 |
| Insufficient payload (model declined to judge) | 0.004 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".