基于肺功能数据训练的深度学习算法对胸片中阻塞性肺病的预测。

Prediction of Obstructive Lung Disease from Chest Radiographs via Deep Learning Trained on Pulmonary Function Data.

机构信息

Department of Radiology and Imaging Sciences, School of Medicine, University of Utah, Salt Lake City, UT, USA.

Department of Electrical and Computer Engineering, Scientific Computing and Imaging Institute (SCI), University of Utah, Salt Lake City, UT, USA.

出版信息

Int J Chron Obstruct Pulmon Dis. 2021 Jan 5;15:3455-3466. doi: 10.2147/COPD.S279850. eCollection 2020.

DOI:10.2147/COPD.S279850

PMID:33447023

原文链接:https://pmc.ncbi.nlm.nih.gov/articles/PMC7801924/

Abstract

BACKGROUND

Chronic obstructive pulmonary disease (COPD), the third leading cause of death worldwide, is often underdiagnosed.

PURPOSE

To develop machine learning methods to predict COPD using chest radiographs and a convolutional neural network (CNN) trained with near-concurrent pulmonary function test (PFT) data. Comparison is made to natural language processing (NLP) of the associated radiologist text reports.

MATERIALS AND METHODS

This IRB-approved single-institution retrospective study uses 6749 two-view chest radiograph exams (2012-2017, 4436 unique subjects, 54% female, 46% male), same-day associated radiologist text reports, and PFT exams acquired within 180 days. The Image Model (Resnet18 pre-trained with ImageNet CNN) is trained using frontal and lateral radiographs and PFTs with 10% of the subjects for validation and 19% for testing. The NLP Model is trained using radiologist text reports and PFTs. The primary metric of model comparison is the area under the receiver operating characteristic curve (AUC).

RESULTS

The Image Model achieves an AUC of 0.814 for prediction of obstructive lung disease (FEV1/FVC <0.7) from chest radiographs and performs better than the NLP Model (AUC 0.704, p<0.001) from radiologist text reports where FEV1 = forced expiratory volume in 1 second and FVC = forced vital capacity. The Image Model performs better for prediction of severe or very severe COPD (FEV1 <0.5) with an AUC of 0.837 versus the NLP model AUC of 0.770 (p<0.001).

CONCLUSION

A CNN Image Model trained on physiologic lung function data (PFTs) can be applied to chest radiographs for quantitative prediction of obstructive lung disease with good accuracy.

摘要

背景

慢性阻塞性肺疾病（COPD）是全球第三大致死原因，但其常常被漏诊。

目的

开发机器学习方法，使用胸部 X 光片和使用近乎同时进行的肺功能测试（PFT）数据进行训练的卷积神经网络（CNN）来预测 COPD。并与相关放射科医生文本报告的自然语言处理（NLP）进行比较。

材料和方法

这项经机构审查委员会批准的单中心回顾性研究使用了 6749 例双视图胸部 X 光检查（2012-2017 年，4436 例独特的受试者，54%为女性，46%为男性）、同一天的相关放射科医生文本报告以及在 180 天内获得的 PFT 检查。使用 10%的受试者进行验证，19%的受试者进行测试，对预训练有 ImageNet CNN 的 Resnet18 进行训练，以训练图像模型。使用放射科医生的文本报告和 PFT 训练 NLP 模型。模型比较的主要指标是接受者操作特征曲线下的面积（AUC）。

结果

图像模型通过胸部 X 光片预测阻塞性肺病（FEV1/FVC<0.7）的 AUC 为 0.814，优于放射科医生文本报告中的 NLP 模型（AUC 为 0.704，p<0.001），其中 FEV1 为 1 秒用力呼气量，FVC 为用力肺活量。图像模型在预测严重或非常严重的 COPD（FEV1<0.5）时表现更好，AUC 为 0.837，而 NLP 模型的 AUC 为 0.770（p<0.001）。