---
title: "A Survey of Large Language Models for Healthcare: from Data,"
description: "This survey asks how large language models are changing healthcare AI and what still prevents their… 23 references, cited by 59 papers here."
authors: ["Kai He", "Rui Mao", "Qika Lin", "Yucheng Ruan", "Xiang Lan", "Mengling Feng", "Erik Cambria"]
year: 2023
venue: "arXiv:2310.05694"
source: "https://arxiv.org/abs/2310.05694"
doi: "10.48550/arxiv.2310.05694"
arxiv: "2310.05694"
licence: "CC BY-SA 4.0 (arXiv)"
licence_url: "http://creativecommons.org/licenses/by-sa/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/he-2023-survey-large-language-models-healthcare/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/he-2023-survey-large-language-models-healthcare/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2023](https://md.research.elamur.ai/free/en/papers/2023/page/6/) / A Survey of Large Language Models for Healthcare: from Data, Technology, and Applications to Accountability and Ethics

# A Survey of Large Language Models for Healthcare: from Data, Technology, and Applications to Accountability and Ethics

**Authors:** Kai He, [Rui Mao](https://md.research.elamur.ai/free/en/authors/rui-mao/), [Qika Lin](https://md.research.elamur.ai/free/en/authors/qika-lin/), Yucheng Ruan, [Xiang Lan](https://md.research.elamur.ai/free/en/authors/xiang-lan/), [Mengling Feng](https://md.research.elamur.ai/free/en/authors/mengling-feng/) and [Erik Cambria](https://md.research.elamur.ai/free/en/authors/erik-cambria/) · **Year:** 2023 · **Venue:** arXiv:2310.05694 · **Licence:** [CC BY-SA 4.0 (arXiv)](http://creativecommons.org/licenses/by-sa/4.0/)

**Links:** [Publisher's page](https://arxiv.org/abs/2310.05694) · [PDF](https://arxiv.org/pdf/2310.05694) · [DOI 10.48550/arxiv.2310.05694](https://doi.org/10.48550/arxiv.2310.05694) · [arXiv 2310.05694](https://arxiv.org/abs/2310.05694) · [OpenAlex](https://openalex.org/W4387560284)

**Citations:** 27 citations (OpenAlex, 2026-10-14) · [59](https://md.research.elamur.ai/free/en/papers/he-2023-survey-large-language-models-healthcare/cited-by/) in our library, of the 141,840 papers whose bibliographies we read

## Summary
This survey asks how large language models are changing healthcare AI and what still prevents their real-world use. The authors trace the field's evolution from earlier pretrained models such as BERT and RoBERTa, which were usually built for one task at a time, to generative systems like GPT-3, GPT-4, Llama and Med-PaLM 2. They cover two levels of capability: basic tasks (entity and relation extraction, text classification, question answering, dialogue) and more advanced uses (clinical decision support, virtual assistants, early disease prediction, literature analysis). They also review training data, learning techniques, training costs and benchmark results for medical LLMs, and they maintain an open-source list of resources. The authors identify two main shifts in the field. The first is a move from discriminative to generative modelling. The second is a move from model-focused to data-focused development. Their central conclusion is that raw performance is becoming less of a barrier. The main obstacles are now fairness, accountability, transparency and ethics, along with hospital IT integration and differing regulations across countries. This makes the survey a practical starting point for anyone choosing or building a domain-specific LLM.

## Key points
- GPT-3, with more than 100 billion parameters, is presented as the turning point from PLMs to LLMs. Med-PaLM 2 is cited for reaching expert-level accuracy on the USMLE.
- The survey highlights specialised models: HuatuoGPT, which actively asks patients questions; OphGLM for ophthalmology; and SoulChat for mental health. It also covers prompting methods such as LLM-NERRE, CARP and AMuLaP.
- LLMs are argued to outperform PLMs most clearly on medical text classification, because clinical text depends heavily on context. For example, "MI" can mean either myocardial infarction or mitral insufficiency.
- Three barriers to deployment are named: difficulty integrating with hospital IT systems, fragmented systems after hospital mergers, and unclear regulation. Federated learning and shared governance are suggested to support cross-border collaboration.
- The curated resource list is hosted at github.com/KaiHe-CatOwner/LLM-for-Healthcare. The final version appeared in Information Fusion in 2025.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [National University of Singapore](https://md.research.elamur.ai/free/en/universities/national-university-of-singapore/) — University, Singapore, Singapore · 84 works · 32,204 citations
- [Nanyang Technological University](https://md.research.elamur.ai/free/en/universities/nanyang-technological-university/) — University, Singapore, Singapore · 73 works · 24,390 citations

📚 Shelves (6): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,086 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,889 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works · [🩺 Medicine](https://md.research.elamur.ai/free/en/shelves/medicine/) · 80 works · [🩺 Health Informatics](https://md.research.elamur.ai/free/en/shelves/medicine/health-informatics/) · 53 works

🧭 Topics (1): [🛡️ Bias, fairness & safety](https://md.research.elamur.ai/free/en/topics/bias-fairness-safety/) · 209 works

## Cite

Cite as: Kai He, Rui Mao, Qika Lin, Yucheng Ruan, Xiang Lan, Mengling Feng and Erik Cambria. 2023. A Survey of Large Language Models for Healthcare: from Data, Technology, and Applications to Accountability and Ethics. arXiv:2310.05694.

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@misc{he2023survey,
  title = {{A Survey of Large Language Models for Healthcare: from Data, Technology, and Applications to Accountability and Ethics}},
  author = {Kai He, Rui Mao, Qika Lin, Yucheng Ruan, Xiang Lan, Mengling Feng and Erik Cambria},
  year = {2023},
  doi = {10.48550/arxiv.2310.05694},
  eprint = {2310.05694},
  archivePrefix = {arXiv},
  url = {https://doi.org/10.48550/arxiv.2310.05694}
}
```

- [← previous RL4F: Generating Natural Language Feedback with Reinforcement Learning for Repairing Model Outputs](https://md.research.elamur.ai/free/en/papers/akyurek-2023-rl4f-generating-natural-language-feedback/)
- [next → PromptNER: Prompting For Named Entity Recognition](https://md.research.elamur.ai/free/en/papers/ashok-2023-promptner-prompting-named-entity-recognition/)
- [↑ back to papers of 2023](https://md.research.elamur.ai/free/en/papers/2023/page/6/)

## References

23 references, 9 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Karan Singhal, et al., [Towards expert-level medical question answering with large language models](https://md.research.elamur.ai/free/en/papers/singhal-2023-towards-expert-level-medical-question/), 2023, arXiv preprint arXiv:2305.09617.
2. Jacob Devlin Ming-Wei Chang Kenton, Lee Kristina Toutanova, [BERT: Pre-training of deep bidirectional transformers for language understanding](https://md.research.elamur.ai/free/en/papers/devlin-2019-bert-pre-training-deep-bidirectional/), in: Proceedings of NAACL-HLT, 2019, pp. 4171–4186.
3. Yinhan Liu, et al., RoBERTa: A robustly optimized BERT pretraining approach, 2019. [find](https://scholar.google.com/scholar?q=Yinhan+Liu%2C+et+al%2E%2C+RoBERTa%3A+A+robustly+optimized+BERT+pretraining+approach%2C+2019%2E)
4. Kai He, et al., Understanding the patient perspective of epilepsy treatment through text mining of online patient support groups, Epilepsy Behav. 94 (2019) 65–71. [find](https://scholar.google.com/scholar?q=Kai+He%2C+et+al%2E%2C+Understanding+the+patient+perspective+of+epilepsy+treatment+through+text+mining+of+online+patient+support+groups%2C+Epilepsy+Behav%2E+94+%282019%29+65%E2%80%9371%2E)
5. Tom Brown, et al., [Language models are few-shot learners](https://md.research.elamur.ai/free/en/papers/brown-2020-language-models-few-shot-learners/), Adv. Neural Inf. Process. Syst. 33 (2020) 1877–1901.
6. Jason Wei, et al., [Chain-of-thought prompting elicits reasoning in large language models](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/), Adv. Neural Inf. Process. Syst. 35 (2022) 24824–24837.
7. Hongjian Zhou, et al., [A survey of large language models in medicine: Progress, application, and challenge](https://md.research.elamur.ai/free/en/papers/zhou-2023-survey-large-language-models-medicine/), 2023, arXiv preprint arXiv:2311.05112.
8. Michael Moor, et al., Med-flamingo: a multimodal medical few-shot learner, 2023, arXiv preprint arXiv:2307.15189. [find](https://scholar.google.com/scholar?q=Michael+Moor%2C+et+al%2E%2C+Med%2Dflamingo%3A+a+multimodal+medical+few%2Dshot+learner%2C+2023%2C+arXiv+preprint+arXiv%3A2307%2E15189%2E)
9. Chunyuan Li, et al., [Llava-med: Training a large language-and-vision assistant for biomedicine in one day](https://md.research.elamur.ai/free/en/papers/li-2023-llava-med-training-large-language/), 2023, arXiv preprint arXiv:2306.00890.
10. Juexiao Zhou, et al., Path to medical AGI: Unify domain-specific medical LLMs with the lowest cost, 2023, arXiv preprint arXiv:2306.10765. — [arXiv:2306.10765](https://arxiv.org/abs/2306.10765) [find](https://scholar.google.com/scholar?q=Juexiao+Zhou%2C+et+al%2E%2C+Path+to+medical+AGI%3A+Unify+domain%2Dspecific+medical+LLMs+with+the+lowest+cost%2C+2023%2C+arXiv+preprint+arXiv%3A2306%2E10765%2E+%E2%80%94+%5BarXiv%3A2306%2E10765%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F2306%2E10765%29)
11. Hongbo Zhang, et al., [HuatuoGPT, towards taming language model to be a doctor](https://md.research.elamur.ai/free/en/papers/zhang-2023-huatuogpt-taming-language-model-be/), 2023, arXiv preprint arXiv:2305.15075.
12. Weihao Gao, et al., OphGLM: Training an ophthalmology large language-and-vision assistant based on instructions and dialogue, 2023, arXiv preprint arXiv:2306.12174. — [arXiv:2306.12174](https://arxiv.org/abs/2306.12174) [find](https://scholar.google.com/scholar?q=Weihao+Gao%2C+et+al%2E%2C+OphGLM%3A+Training+an+ophthalmology+large+language%2Dand%2Dvision+assistant+based+on+instructions+and+dialogue%2C+2023%2C+arXiv+preprint+arXiv%3A2306%2E12174%2E+%E2%80%94+%5BarXiv%3A2306%2E12174%5D%28https%3A%2F%2Farxiv%2E)
13. Chen Yirong, et al., SoulChat: The "empathy" ability of the large model is improved by mixing and fine-tuning the data set of long text consultation instructions and multiple rounds of empathy dialogue, 2023. [find](https://scholar.google.com/scholar?q=Chen+Yirong%2C+et+al%2E%2C+SoulChat%3A+The+%22empathy%22+ability+of+the+large+model+is+improved+by+mixing+and+fine%2Dtuning+the+data+set+of+long+text+consultation+instructions+and+multiple+rounds+of+empathy+dialogu)
14. Jesutofunmi A. Omiye, et al., Large language models in medicine: the potentials and pitfalls, 2023. [find](https://scholar.google.com/scholar?q=Jesutofunmi+A%2E+Omiye%2C+et+al%2E%2C+Large+language+models+in+medicine%3A+the+potentials+and+pitfalls%2C+2023%2E)
15. Linmei Hu, et al., A survey of knowledge enhanced pre-trained language models, IEEE Trans. Knowl. Data Eng. (2023). [find](https://scholar.google.com/scholar?q=Linmei+Hu%2C+et+al%2E%2C+A+survey+of+knowledge+enhanced+pre%2Dtrained+language+models%2C+IEEE+Trans%2E+Knowl%2E+Data+Eng%2E+%282023%29%2E)
16. Shubo Tian, Qiao Jin, Lana Yeganova, Po-Ting Lai, Qingqing Zhu, Xiuying Chen, Yifan Yang, Qingyu Chen, Won Kim, Donald C. Comeau, et al., Opportunities and challenges for ChatGPT and large language models in biomedicine and health, Brief. Bioinform. 25 (1) (2024) bbad493. [find](https://scholar.google.com/scholar?q=Shubo+Tian%2C+Qiao+Jin%2C+Lana+Yeganova%2C+Po%2DTing+Lai%2C+Qingqing+Zhu%2C+Xiuying+Chen%2C+Yifan+Yang%2C+Qingyu+Chen%2C+Won+Kim%2C+Donald+C%2E+Comeau%2C+et+al%2E%2C+Opportunities+and+challenges+for+ChatGPT+and+large+language+mo)
17. Wayne Xin Zhao, et al., [A survey of large language models](https://md.research.elamur.ai/free/en/papers/zhao-2023-survey-large-language-models/), 2023, arXiv preprint arXiv:2303.18223.
18. Bonan Min, et al., [Recent advances in natural language processing via large pre-trained language models: A survey](https://md.research.elamur.ai/free/en/papers/min-2021-recent-advances-natural-language-processing/), ACM Comput. Surv. (2021).
19. Suhana Bedi, Yutong Liu, Lucy Orr-Ewing, Dev Dash, Sanmi Koyejo, Alison Callahan, Jason A. Fries, Michael Wornow, Akshay Swaminathan, Lisa Soleymani Lehmann, et al., Testing and evaluation of health care applications of large language models: a systematic review, JAMA (2024). [find](https://scholar.google.com/scholar?q=Suhana+Bedi%2C+Yutong+Liu%2C+Lucy+Orr%2DEwing%2C+Dev+Dash%2C+Sanmi+Koyejo%2C+Alison+Callahan%2C+Jason+A%2E+Fries%2C+Michael+Wornow%2C+Akshay+Swaminathan%2C+Lisa+Soleymani+Lehmann%2C+et+al%2E%2C+Testing+and+evaluation+of+health+c)
20. Elizabeth C. Stade, Shannon Wiltsey Stirman, Lyle H. Ungar, Cody L. Boland, H. Andrew Schwartz, David B. Yaden, João Sedoc, Robert J. DeRubeis, Robb Willer, Johannes C. Eichstaedt, Large language models could change the future of behavioral healthcare: a proposal for responsible development and eval… [find](https://scholar.google.com/scholar?q=Elizabeth+C%2E+Stade%2C+Shannon+Wiltsey+Stirman%2C+Lyle+H%2E+Ungar%2C+Cody+L%2E+Boland%2C+H%2E+Andrew+Schwartz%2C+David+B%2E+Yaden%2C+Jo%C3%A3o+Sedoc%2C+Robert+J%2E+DeRubeis%2C+Robb+Willer%2C+Johannes+C%2E+Eichstaedt%2C+Large+language+mode)
21. P.M. Lavanya, E. Sasikala, Deep learning techniques on text classification using natural language processing (NLP) in social healthcare network: A comprehensive survey, in: 2021 3rd International Conference on Signal Processing and Communication, ICPSC, IEEE, 2021, pp. 603–609. [find](https://scholar.google.com/scholar?q=P%2EM%2E+Lavanya%2C+E%2E+Sasikala%2C+Deep+learning+techniques+on+text+classification+using+natural+language+processing+%28NLP%29+in+social+healthcare+network%3A+A+comprehensive+survey%2C+in%3A+2021+3rd+International+Conf)
22. Kai He, Lixia Yao, Knowledge enhanced coreference resolution via gated attention, in: 2022 IEEE International Conference on Bioinformatics and Biomedicine, BIBM, IEEE, 2022, pp. 2287–2293. [find](https://scholar.google.com/scholar?q=Kai+He%2C+Lixia+Yao%2C+Knowledge+enhanced+coreference+resolution+via+gated+attention%2C+in%3A+2022+IEEE+International+Conference+on+Bioinformatics+and+Biomedicine%2C+BIBM%2C+IEEE%2C+2022%2C+pp%2E+2287%E2%80%932293%2E)
23. Kai He, et al., Construction of genealogical knowledge graphs from obituaries: Multitask neural network extraction system, J. Med. Internet Res. 23 (8) (2021) e25670. \#\# 7. Conclusion [find](https://scholar.google.com/scholar?q=Kai+He%2C+et+al%2E%2C+Construction+of+genealogical+knowledge+graphs+from+obituaries%3A+Multitask+neural+network+extraction+system%2C+J%2E+Med%2E+Internet+Res%2E+23+%288%29+%282021%29+e25670%2E+%5C%23%5C%23+7%2E+Conclusion)

*Updated 24 October 2026*