Superiority of Supervised Machine Learning on Reading Chest X-Rays in Intensive Care Units.

Kumiko Tanaka,Takuro Horikoshi,Nozomi Takahashi,Toshiya Nakaguchi,Taka-Aki Nakada,Yuichiro Yoshimura,Koichiro Shinozaki,Hajime Yokota,Takahiro Dozono

doi:10.3389/fmed.2021.676277

Abstract

Purpose: Portable chest radiographs are diagnostically indispensable in intensive care units (ICU). This study aimed to determine if the proposed machine learning technique increased in accuracy as the number of radiograph readings increased and if it was accurate in a clinical setting.Methods: Two independent data sets of portable chest radiographs (n = 380, a single Japanese hospital; n = 1,720, The National Institution of Health [NIH] ChestX-ray8 dataset) were analyzed. Each data set was divided training data and study data. Images were classified as atelectasis, pleural effusion, pneumonia, or no emergency. DenseNet-121, as a pre-trained deep convolutional neural network was used and ensemble learning was performed on the best-performing algorithms. Diagnostic accuracy and processing time were compared to those of ICU physicians.Results: In the single Japanese hospital data, the area under the curve (AUC) of diagnostic accuracy was 0.768. The area under the curve (AUC) of diagnostic accuracy significantly improved as the number of radiograph readings increased from 25 to 100% in the NIH data set. The AUC was higher than 0.9 for all categories toward the end of training with a large sample size. The time to complete 53 radiographs by machine learning was 70 times faster than the time taken by ICU physicians (9.66 s vs. 12 min). The diagnostic accuracy was higher by machine learning than by ICU physicians in most categories (atelectasis, AUC 0.744 vs. 0.555, P < 0.05; pleural effusion, 0.856 vs. 0.706, P < 0.01; pneumonia, 0.720 vs. 0.744, P = 0.88; no emergency, 0.751 vs. 0.698, P = 0.47).Conclusions: We developed an automatic detection system for portable chest radiographs in ICU setting; its performance was superior and quite faster than ICU physicians.

Highlights

Ill patients often have organ dysfunction and require frequent and intense monitoring
The time to complete 53 radiographs by machine learning was 70 times faster than the time taken by intensive care unit (ICU) physicians (9.66 s vs. 12 min)
The diagnostic accuracy was higher by machine learning than by ICU physicians in most categories

Summary

Introduction

Ill patients often have organ dysfunction and require frequent and intense monitoring. Portable chest radiography is key to assessing cardiopulmonary function in the intensive care unit (ICU), allowing clinicians to identify pathological findings such as pneumonia, pneumothorax, pleural effusion, and atelectasis (1–7). A review of a large number of portable chest radiographs with high accuracy is important for the improvement of ICU patient outcomes; this can be challenging, primarily due to a lack of manpower (8, 9). Machine learning technology is effective in analyzing a large amount of data, including image data (10–14). This promising technique potentially supports interpretations of radiographs, which may improve quality of care and patient safety by reducing physician’s workload in ICU. Chest radiographs have been the most investigated; insufficient accuracy limits its clinical use. Investigations on chest radiographs in the ICU have not been well elucidated

Objectives

Methods

Results

Discussion

Conclusion