Can machines be trusted? Robustification of deep learning for medical imaging

机器可以信任吗？

• 批准号: 10208969
• 负责人: John William Garrett
• 金额: $ 31.89万
• 依托单位: UNIVERSITY OF WISCONSIN-MADISON
• 依托单位国家: 美国
• 财政年份: 2020
• 资助国家: 美国
• 起止时间: 2020-07-02 至 2024-03-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10208969
• 关键词: Adopted; Algorithms; Attention; Brain; Cardiac; Classification; Clinical; Clinical Research; Critiques; Dangerous Behavior; Data; Data Set; Diagnostic radiologic examination; Disease; Dose; Effectiveness; Ensure; Exhibits; Exposure to; Failure; Goals; Human; Image; Image Analysis; Label; Machine Learning; Magnetic Resonance Imaging; Mathematics; Medical Imaging; Methods; Modeling; Morphologic artifacts; Motion; Noise; Output; Pattern; Performance; Physics; Positron-Emission Tomography; Predisposition; Recommendation; Research; Research Design; Research Personnel; Scheme; Source; Structure; System; Thoracic Radiography; Training; Trust; Tumor Tissue; Variant; Work; X-Ray Computed Tomography; base; classification algorithm; clinical implementation; clinically relevant; deep learning; deep learning algorithm; design; disease diagnosis; human error; imaging Segmentation; improved; instrument; learning community; loss of function; machine learning algorithm; neural network; novel; operation; performance tests; physical process; radiologist; reconstruction; resilience; statistics; success; tumor

Machine learning algorithms have become increasing popular in medical imaging, where highly functional algorithms have been trained to recognize patterns or features within image data sets and perform clinically relevant tasks such as tumor segmentation and disease diagnosis. In recent years, an approach known as deep learning has revolutionized the field of machine learning, by leveraging massive datasets and immense computing power to extract features from data. Deep learning is ideally suited for problems in medical imaging, and has enjoyed success in diverse tasks such as segmenting cardiac structures, tumors, and tissues. However, research in machine learning has also shown that deep learning is fragile in the sense that carefully designed perturbations to an image can cause the algorithm to fail. These perturbations can be designed to be imperceptible by humans, so that a trained radiologist would not make the same mistakes. As deep learning approaches gain acceptance and move toward clinical implementation, it is therefore crucial to develop a better understanding of the performance of neural networks. Specifically, it is critical to understand the limits of deep learning when presented with noisy or imperfect data. The goal of this project is to explore these questions in the context of medical imaging—to better identify strengths, weaknesses, and failure points of deep learning algorithms. We posit that malicious perturbations, of the type studied in theoretical machine learning, may not be representative of the sort of noise encountered in medical images. Although noise is inevitable in a physical system, the noise arising from sources such as subject motion, operator error, or instrument malfunction may have less deleterious effects on a deep learning algorithm. We propose to characterize the effect of these perturbations on the performance of deep learning algorithms. Furthermore, we will study the effect of random labeling error introduced into the data set, as might arise due to honest human error. We will also develop new methods for making deep learning algorithms more robust to the types of clinically relevant perturbations described above. In summary, although the susceptibility of neural networks to small errors in the inputs is widely recognized in the deep learning community, our work will investigate these general phenomena in the specific case of medical imaging tasks, and also conduct the first study of average-case errors that could realistically arise in clinical studies. Furthermore, we will produce novel recommendations for how to quantify and improve the resiliency of deep learning approaches in medical imaging.

机器学习算法在医学成像中越来越流行，在医学成像中高度功能 已经对算法进行了训练以识别图像数据集中的模式或功能并在临床上执行 相关任务，例如肿瘤分割和疾病诊断。近年来，一种被称为 深度学习通过利用大量数据集彻底改变了机器学习领域 从数据中提取功能的计算能力。深度学习非常适合医学成像中的问题， 并在细分心脏结构，肿瘤和组织等潜水任务中取得了成功。 但是，机器学习的研究还表明，从仔细的意义上讲，深度学习是脆弱的 设计对图像的扰动可能会导致该算法失败。这些扰动可以设计为 人类无法察觉，因此受过训练的辐射主义者不会犯同样的错误。作为深度学习 方法获得接受并朝着临床实施迈进，因此，开发一个至关重要的 更好地了解神经网络的性能。具体而言，了解的限制至关重要 深度学习时，请介绍噪声或不完美的数据。该项目的目的是探索这些 在医学成像背景下的问题 - 更好地识别优势，劣势和失败点 深度学习算法。 我们指出，理论机器学习中研究类型的恶意扰动可能不是 代表医学图像中遇到的噪音。虽然在物理上不可避免地噪音 系统，诸如主题运动，操作员错误或仪器故障等来源引起的噪音可能 对深度学习算法的有害影响较小。我们建议表征这些效果 对深度学习算法的性能的扰动。此外，我们将研究随机的效果 由于诚实的人为错误而可能引起的标记错误引入了数据集。我们还将开发新的 使深度学习算法对临床相关扰动的类型更强大的方法 上面描述。 总而言之，尽管神经网络对输入中较小错误的敏感性得到了广泛认可 深度学习社区，我们的工作将在特定情况下调查这些一般现象 医学成像任务，还对平均案例错误进行了首次研究，这些错误实际上可能在 临床研究。此外，我们将为如何量化和改进提出新的建议 在医学成像中深入学习方法的弹性。