[论文解读] OncoPetNet: A Deep Learning based AI system for mitotic figure counting on H&E stained whole slide digital images in a large veterinary diagnostic lab setting.

Michael Fitzke, Derick Whitley+8 more|arXiv (Cornell University)|Aug 17, 2021

AI in cancer detection参考文献 39被引用 5

一句话总结

OncoPetNet 是一种深度学习系统，可在高通量的兽医诊断病理学环境中自动对苏木精和伊红（H&E）染色的全切片图像进行有丝分裂象计数。该系统实现每张切片 0.27 分钟的实时、专家级性能，并在 21.9% 的病例中改变了肿瘤分级，与人类专家相比存在差异。

ABSTRACT

Background: Histopathology is an important modality for the diagnosis and management of many diseases in modern healthcare, and plays a critical role in cancer care. Pathology samples can be large and require multi-site sampling, leading to upwards of 20 slides for a single tumor, and the human-expert tasks of site selection and and quantitative assessment of mitotic figures are time consuming and subjective. Automating these tasks in the setting of a digital pathology service presents significant opportunities to improve workflow efficiency and augment human experts in practice. Approach: Multiple state-of-the-art deep learning techniques for histopathology image classification and mitotic figure detection were used in the development of OncoPetNet. Additionally, model-free approaches were used to increase speed and accuracy. The robust and scalable inference engine leverages Pytorch's performance optimizations as well as specifically developed speed up techniques in inference. Results: The proposed system, demonstrated significantly improved mitotic counting performance for 41 cancer cases across 14 cancer types compared to human expert baselines. In 21.9% of cases use of OncoPetNet led to change in tumor grading compared to human expert evaluation. In deployment, an effective 0.27 min/slide inference was achieved in a high throughput veterinary diagnostic pathology service across 2 centers processing 3,323 digital whole slide images daily. Conclusion: This work represents the first successful automated deployment of deep learning systems for real-time expert-level performance on important histopathology tasks at scale in a high volume clinical practice. The resulting impact outlines important considerations for model development, deployment, clinical decision making, and informs best practices for implementation of deep learning systems in digital histopathology practices.

研究动机与目标

• 开发一种可扩展的深度学习系统，用于 H&E 染色全切片图像中的自动化有丝分裂象计数。
• 解决临床病理学工作流程中手动有丝分裂象计数耗时且主观的问题。
• 在高通量、多机构的兽医诊断实验室环境中部署该系统。
• 评估人工智能对肿瘤分级准确性和工作流程效率的影响。
• 建立在真实世界数字病理学环境中部署深度学习的最佳实践。

提出的方法

• 采用最先进的深度学习模型，用于全切片图像分类和有丝分裂象检测。
• 结合无模型优化技术与 PyTorch 性能增强，以加速推理过程。
• 实现了一个稳健、可扩展的推理引擎，专为在多个临床机构中实时部署而设计。
• 在来自 14 种癌症类型的 3,323 张全切片图像数据集上进行训练，数据来自 2 家兽医诊断中心。
• 由获得认证的病理学家对有丝分裂象进行标注，并验证模型预测结果。
• 将切片分类与有丝分裂象检测作为两级任务进行整合，以提高准确性和效率。

实验结果

研究问题

• RQ1 深度学习系统是否能在真实世界、高通量的兽医病理学环境中实现专家级的有丝分裂象计数性能？
• RQ2 基于人工智能的有丝分裂象计数与人类专家在肿瘤分级中的一致性与变异性相比如何？
• RQ3 哪些技术和操作策略能够实现人工智能在临床数字病理学工作流程中的可扩展、实时部署？
• RQ4 与人类评估相比，人工智能集成在多大程度上改变了临床决策，如肿瘤分级？
• RQ5 在准确性之外，将人工智能系统部署于临床病理学中还需考虑哪些关键因素，包括可重复性和泛化能力？

主要发现

• OncoPetNet 在每日报量处理 3,323 张全切片图像的高通量环境中，实现了每张切片 0.27 分钟的推理时间。
• 与人类专家基线相比，该系统在涵盖 14 种肿瘤类型的 41 例癌症病例中，显著提升了有丝分裂象计数性能。
• 在 21.9% 的病例中，AI 系统导致肿瘤分级与人类专家评估结果不同。
• 该系统是首个在临床实践中实现大规模、实时部署深度学习技术以达到专家级有丝分裂象计数性能的成功案例。
• 模型部署提升了工作流程效率，减少了有丝分裂象计数评估的主观性，对临床决策具有重要影响。
• 本研究为数字病理学中模型开发、部署及工作流程整合提供了关键洞见。