arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.08749cs.CVcs.LG
收藏

使用深度学习和可解释人工智能自动检测脑MR图像中的运动伪影

Automated detection of motion artifacts in brain MR images using deep learning and explainable artificial intelligence

Columbia University in the City of New York(哥伦比亚大学);GE Healthcare(通用电气医疗);Columbia University Medical Center(哥伦比亚大学医学中心);John Hopkins University(约翰斯·霍普金斯大学)

Marina Manso Jimeno, Keerthi Sravan Ravi, Maggie Fung, John Thomas Vaughan,, Sairam Geethanath

AI总结:

本研究提出一种基于2D CNN和Grad-CAM的深度学习模型,用于T1加权脑图像中刚性运动伪影的自动检测,并作为ArtifactID工具的一部分,有效自动化了MRI质量评估过程。

摘要:

质量评估(包括检查图像中的伪影)是MRI数据采集过程中的关键步骤,旨在确保数据质量及下游分析或解读的成功。本研究展示了一个深度学习模型,用于检测T1加权脑图像中的刚性运动。我们利用2D CNN进行三分类,并在公开的回顾性和前瞻性数据集上进行了测试。Grad-CAM热力图能够识别失败模式,并为模型结果提供解释。该模型在六个运动模拟回顾性数据集上实现了85%的平均精确率和80%的召回率。此外,模型在前瞻性数据集上的分类结果与平均边缘强度(一种指示运动的图像质量指标)显示出强烈的负相关(-0.84)。该模型是ArtifactID工具的一部分,旨在内联自动检测Gibbs振铃、卷绕和运动伪影。该工具自动化了部分耗时的QA过程,并增强了现场专业知识,这在本地MR知识稀缺的低资源环境中尤为相关。

英文摘要:

Quality assessment, including inspecting the images for artifacts, is a critical step during MRI data acquisition to ensure data quality and downstream analysis or interpretation success. This study demonstrates a deep learning model to detect rigid motion in T1-weighted brain images. We leveraged a 2D CNN for three-class classification and tested it on publicly available retrospective and prospective datasets. Grad-CAM heatmaps enabled the identification of failure modes and provided an interpretation of the model's results. The model achieved average precision and recall metrics of 85% and 80% on six motion-simulated retrospective datasets. Additionally, the model's classifications on the prospective dataset showed a strong inverse correlation (-0.84) compared to average edge strength, an image quality metric indicative of motion. This model is part of the ArtifactID tool, aimed at inline automatic detection of Gibbs ringing, wrap-around, and motion artifacts. This tool automates part of the time-consuming QA process and augments expertise on-site, particularly relevant in low-resource settings where local MR knowledge is scarce.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑