BadCLM: Backdoor Attack in Clinical Language Models for Electronic Health Records
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909245604102144 |
|---|---|
| author | Lyu, Weimin Bi, Zexin Wang, Fusheng Chen, Chao |
| author_facet | Lyu, Weimin Bi, Zexin Wang, Fusheng Chen, Chao |
| contents | The advent of clinical language models integrated into electronic health records (EHR) for clinical decision support has marked a significant advancement, leveraging the depth of clinical notes for improved decision-making. Despite their success, the potential vulnerabilities of these models remain largely unexplored. This paper delves into the realm of backdoor attacks on clinical language models, introducing an innovative attention-based backdoor attack method, BadCLM (Bad Clinical Language Models). This technique clandestinely embeds a backdoor within the models, causing them to produce incorrect predictions when a pre-defined trigger is present in inputs, while functioning accurately otherwise. We demonstrate the efficacy of BadCLM through an in-hospital mortality prediction task with MIMIC III dataset, showcasing its potential to compromise model integrity. Our findings illuminate a significant security risk in clinical decision support systems and pave the way for future endeavors in fortifying clinical language models against such vulnerabilities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_05213 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | BadCLM: Backdoor Attack in Clinical Language Models for Electronic Health Records Lyu, Weimin Bi, Zexin Wang, Fusheng Chen, Chao Computation and Language Artificial Intelligence The advent of clinical language models integrated into electronic health records (EHR) for clinical decision support has marked a significant advancement, leveraging the depth of clinical notes for improved decision-making. Despite their success, the potential vulnerabilities of these models remain largely unexplored. This paper delves into the realm of backdoor attacks on clinical language models, introducing an innovative attention-based backdoor attack method, BadCLM (Bad Clinical Language Models). This technique clandestinely embeds a backdoor within the models, causing them to produce incorrect predictions when a pre-defined trigger is present in inputs, while functioning accurately otherwise. We demonstrate the efficacy of BadCLM through an in-hospital mortality prediction task with MIMIC III dataset, showcasing its potential to compromise model integrity. Our findings illuminate a significant security risk in clinical decision support systems and pave the way for future endeavors in fortifying clinical language models against such vulnerabilities. |
| title | BadCLM: Backdoor Attack in Clinical Language Models for Electronic Health Records |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2407.05213 |