Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2505.23806 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912403472515072 |
|---|---|
| author | Lee, Sihyeon Song, Hyunjoo Lee, Jong-chan Lee, Yoon Jin Lee, Boram Lim, Hee-Eon Kim, Dongyeong Seo, Jinwook Kim, Bohyoung |
| author_facet | Lee, Sihyeon Song, Hyunjoo Lee, Jong-chan Lee, Yoon Jin Lee, Boram Lim, Hee-Eon Kim, Dongyeong Seo, Jinwook Kim, Bohyoung |
| contents | Deploying large language models (LLMs) in clinical settings faces critical trade-offs: cloud LLMs, with their extensive parameters and superior performance, pose risks to sensitive clinical data privacy, while local LLMs preserve privacy but often fail at complex clinical interpretation tasks. We propose MedOrchestra, a hybrid framework where a cloud LLM decomposes complex clinical tasks into manageable subtasks and prompt generation, while a local LLM executes these subtasks in a privacy-preserving manner. Without accessing clinical data, the cloud LLM generates and validates subtask prompts using clinical guidelines and synthetic test cases. The local LLM executes subtasks locally and synthesizes outputs generated by the cloud LLM. We evaluate MedOrchestra on pancreatic cancer staging using 100 radiology reports under NCCN guidelines. On free-text reports, MedOrchestra achieves 70.21% accuracy, outperforming local model baselines (without guideline: 48.94%, with guideline: 56.59%) and board-certified clinicians (gastroenterologists: 59.57%, surgeons: 65.96%, radiologists: 55.32%). On structured reports, MedOrchestra reaches 85.42% accuracy, showing clear superiority across all settings. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_23806 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MedOrchestra: A Hybrid Cloud-Local LLM Approach for Clinical Data Interpretation Lee, Sihyeon Song, Hyunjoo Lee, Jong-chan Lee, Yoon Jin Lee, Boram Lim, Hee-Eon Kim, Dongyeong Seo, Jinwook Kim, Bohyoung Computation and Language Artificial Intelligence Deploying large language models (LLMs) in clinical settings faces critical trade-offs: cloud LLMs, with their extensive parameters and superior performance, pose risks to sensitive clinical data privacy, while local LLMs preserve privacy but often fail at complex clinical interpretation tasks. We propose MedOrchestra, a hybrid framework where a cloud LLM decomposes complex clinical tasks into manageable subtasks and prompt generation, while a local LLM executes these subtasks in a privacy-preserving manner. Without accessing clinical data, the cloud LLM generates and validates subtask prompts using clinical guidelines and synthetic test cases. The local LLM executes subtasks locally and synthesizes outputs generated by the cloud LLM. We evaluate MedOrchestra on pancreatic cancer staging using 100 radiology reports under NCCN guidelines. On free-text reports, MedOrchestra achieves 70.21% accuracy, outperforming local model baselines (without guideline: 48.94%, with guideline: 56.59%) and board-certified clinicians (gastroenterologists: 59.57%, surgeons: 65.96%, radiologists: 55.32%). On structured reports, MedOrchestra reaches 85.42% accuracy, showing clear superiority across all settings. |
| title | MedOrchestra: A Hybrid Cloud-Local LLM Approach for Clinical Data Interpretation |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2505.23806 |