[论文解读] Foundation Model for Advancing Healthcare: Challenges, Opportunities, and Future Directions

Yuting He, Fuxiang Huang+5 more|arXiv (Cornell University)|Apr 4, 2024

Primary Care and Health Outcomes被引用 5

一句话总结

对 Healthcare Foundation Models (HFMs) 在语言、视觉、生物信息学和多模态子领域的全面综述，详细介绍方法、数据资源、应用、挑战及未来方向。

ABSTRACT

Foundation model, which is pre-trained on broad data and is able to adapt to a wide range of tasks, is advancing healthcare. It promotes the development of healthcare artificial intelligence (AI) models, breaking the contradiction between limited AI models and diverse healthcare practices. Much more widespread healthcare scenarios will benefit from the development of a healthcare foundation model (HFM), improving their advanced intelligent healthcare services. Despite the impending widespread deployment of HFMs, there is currently a lack of clear understanding about how they work in the healthcare field, their current challenges, and where they are headed in the future. To answer these questions, a comprehensive and deep survey of the challenges, opportunities, and future directions of HFMs is presented in this survey. It first conducted a comprehensive overview of the HFM including the methods, data, and applications for a quick grasp of the current progress. Then, it made an in-depth exploration of the challenges present in data, algorithms, and computing infrastructures for constructing and widespread application of foundation models in healthcare. This survey also identifies emerging and promising directions in this field for future development. We believe that this survey will enhance the community's comprehension of the current progress of HFM and serve as a valuable source of guidance for future development in this field. The latest HFM papers and related resources are maintained on our website: https://github.com/YutingHe-list/Awesome-Foundation-Models-for-Advancing-Healthcare.

研究动机与目标

• 总结 Healthcare Foundation Models (HFMs) 在四个子领域——语言、视觉、生物信息学和多模态——的当前进展。
• 分析阻碍 HFM 部署的数据、算法和计算基础设施方面的挑战。
• 提供 HFMs 中使用的预训练与适应方法的分类法。
• 确定数据集、应用和新兴方向，为未来研究提供指导。

提出的方法

• 按子领域（LFM、VFM、BFM、MFM）和预训练范式（生成式、对比式、混合、监督式）提供系统的 HFMs 分类。
• 评估适应策略（微调、适配器微调、提示工程）及其在将 HFMs 转移到医疗任务中的作用。
• 总结四个子领域的数据集和资源，以评估数据可用性与局限性。
• 调查应用并将最新的 HFMs 映射到临床情境，以说明实际影响。
• 讨论在数据质量/多样性、算法可靠性和计算基础设施方面的关键挑战。）

实验结果

研究问题

• RQ1 在语言、视觉、生物信息学和多模态领域，HFMs 在医疗保健方面的当前进展如何？
• RQ2 HFMs 在医疗保健中面临的主要数据、算法和基础设施挑战是什么？
• RQ3 哪些未来方向和机会最有望推动 HFMs 在临床实践中的应用？

主要发现

• 四个子领域的 HFMs 通过使通用能力适用于多样化任务，推动了医疗保健 AI 的发展。
• 在数据伦理、多样性、异质性和成本方面存在重大挑战，阻碍大规模 HFM 的训练与部署。
• 高维医疗数据的计算基础设施需求相当庞大，且对环境具有影响。
• 对预训练与适应方法的广泛分类显示，在预训练阶段广泛使用生成式和对比学习，在下游任务中使用微调或基于提示的适应。
• 该综述提供了大量数据集、应用，以及面向未来的方向，指导 HFMs 的后续发展。