Enabling Next Generation Machine Learning for Large Scale Image Analysis

实现大规模图像分析的下一代机器学习

• 批准号: 10384903
• 负责人: Gerald Sabin
• 金额: $ 25.66万
• 依托单位: RNET TECHNOLOGIES, INC.
• 依托单位国家: 美国
• 财政年份: 2021
• 资助国家: 美国
• 起止时间: 2021-09-30 至 2022-03-29
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/10384903
• 关键词: 3-Dimensional; Address; Adoption; Architecture; Area; Cancer Detection; Caring; Classification; Clinical; Community Hospitals; Complex; Computer software; Data; Development; Diagnosis; Diagnostic; Disease; Evolution; Generations; Glass; Goals; Government Agencies; Graph; Growth; Health; Healthcare Systems; High Performance Computing; Human; Image; Image Analysis; Imaging problem; Incidence; Industry; Interobserver Variability; Investments; Judgment; Label; Lead; Learning; Lung; Machine Learning; Magnetic Resonance Imaging; Malignant Neoplasms; Maps; Medical; Medical Imaging; Medical Students; Memory; Methods; Modeling; Movement; Pathologist; Pathology; Patients; Performance; Phase; Play; Reproducibility; Research; Research Personnel; Resolution; Role; Rural Hospitals; Schedule; Screening for Prostate Cancer; Services; Slide; Stream; Techniques; Technology; TensorFlow; Time; Training; Work; aging population; algorithm development; base; clinical application; clinical diagnostics; convolutional neural network; deep learning; design; diagnostic accuracy; digital imaging; digital pathology; implementation efforts; improved; interest; learning network; learning strategy; medical specialties; network architecture; neural network; next generation; novel; overtreatment; pathology imaging; prototype; screening; tool; tumor; user-friendly

Project Summary/Abstract Deep learning has transformed medical image analysis by delivering clinically meaningful results on challenging problems like prostate cancer detection and lung screening. In pathology, industry is making significant invest- ments to develop deep learning tools for diagnostic use in clinical labs. FDA approval of whole-slide digital pathology images (WSIs) for use in primary diagnosis is further increasing interest, adoption, and investment in this technology. Judgments made by pathologists are the basis for the treatment of many diseases, yet in- terobserver variability among pathologists is significant, and errors can lead to overtreatment or even treatment of healthy patients. Pathology is also facing workforce issues as demand for pathologist services is outpacing growth of trained pathologists. Computational pathology tools based on deep learning can help address these problems by providing reproducible diagnoses, performing ”second reads” for human pathologists, automating tasks to improve pathologist efficiency, and helping general pathologists evaluate challenging cases. GPU accel- erators have played a significant role in advancing deep learning methods to build computational pathology tools, with machine learning frameworks (MLFs) like Pytorch and Tensorflow providing researchers with abstractions to quickly develop models that utilize GPUs. Evolution of GPUs and MLFs has been driven by analysis of small images, and so these tools cannot be easily applied directly WSIs or other large medical images like three dimen- sional MRI or CT. Adapting medical imaging problems to small image paradigms supported by GPUs and MLFs leads to suboptimal performance and increased implementation effort and complexity. More recent approaches that use streaming or ”unified memory” allow direct analysis of entire WSIs and have demonstrated performance advantages. These approaches can be slow, complex to implement, and are highly specific to a choice of network architecture which limits exploration and development of new architectures. More general-purpose, efficient, and user-friendly frameworks are required to allow the development of WSI scale deep learning. This project will develop techniques to automatically map deep learning networks implemented in common MLF architectures to one or more GPUs for arbitrarily large input images and activation layers. The proposed software will include a performance modeler to estimate the runtime of a given network on available GPU acceler- ators. These strategies will enable a new paradigm in deep learning for medical images, allowing the development of novel networks that are purpose-built for medical applications. Developers will be able to rapidly create and evaluate these networks using familiar MLF packages. This project will provide approaches to overcome GPU memory bottlenecks, a scheduler to map the network to available GPUs, integration with common MLFs, and demonstration using computational pathology use cases.

项目概要/摘要 深度学习通过在具有挑战性的情况下提供具有临床意义的结果来改变医学图像分析 前列腺癌检测和肺部筛查等问题。在病理学领域，工业界正在进行大量投资 开发用于临床实验室诊断用途的深度学习工具。 FDA 批准全幻灯片数字化 用于初步诊断的病理图像 (WSI) 进一步增加了人们的兴趣、采用和投资 在这项技术中。病理学家的判断是许多疾病治疗的基础，但在 病理学家之间的观察者差异很大，错误可能导致过度治疗甚至治疗 的健康患者。由于对病理学家服务的需求超过了病理学，病理学也面临着劳动力问题 训练有素的病理学家的成长。基于深度学习的计算病理学工具可以帮助解决这些问题 通过提供可重复的诊断、为人类病理学家执行“二次读取”、自动化来解决问题 提高病理学家效率的任务，并帮助普通病理学家评估具有挑战性的病例。 GPU加速 推动者在推进深度学习方法以构建计算病理学工具方面发挥了重要作用， 使用 Pytorch 和 Tensorflow 等机器学习框架 (MLF) 为研究人员提供抽象 快速开发利用 GPU 的模型。 GPU 和 MLF 的发展是由小型分析驱动的 图像，因此这些工具不能轻松地直接应用 WSI 或其他大型医学图像（例如三维） 局部 MRI 或 CT。将医学成像问题适应 GPU 和 MLF 支持的小图像范例 导致性能不佳并增加实施工作量和复杂性。最近的方法 使用流或“统一内存”允许直接分析整个 WSI 并展示了性能 优点。这些方法可能很慢、实施起来很复杂，并且对网络的选择有很强的针对性。 限制新架构探索和开发的架构。更通用、更高效、 需要用户友好的框架来开发 WSI 规模的深度学习。 该项目将开发自动映射常见深度学习网络的技术 MLF 架构到一个或多个 GPU，用于任意大的输入图像和激活层。拟议的 软件将包括一个性能建模器，用于估计给定网络在可用 GPU 加速器上的运行时间 演员。这些策略将为医学图像深度学习带来新的范式，从而促进医学图像的发展 专为医疗应用而构建的新型网络。开发人员将能够快速创建和 使用熟悉的 MLF 包评估这些网络。该项目将提供克服 GPU 的方法 内存瓶颈、将网络映射到可用 GPU 的调度程序、与常见 MLF 的集成，以及 使用计算病理学用例进行演示。