Using Artificial Intelligence for Chest Radiograph Interpretation: A Retrospective Multi-reader-multi-case (MRMC) Study of the Automatic Detection of Multiple Abnormalities and Generation of Diagnostic Report System

Guo, Lin1; Cheng, Guanxun2; Wang, Lifei3; Zheng, Bin4; Jaeger, Stefan5; Giger, Maryellen L.6; Fuhrman, Jordan6; Li, Hui6; Divekar, Ajay7; Xiao, Qian1; Qian, Lingjun1; Xia, Li1; Li, Hongjun8; Lure, Fleming Y. M.9

10.1117/12.3005136

2024

Conference on Medical Imaging - Computer-Aided Diagnosis

1605-7422

COMPUTER-AIDED DIAGNOSIS, MEDICAL IMAGING 2024

12927

FEB 19-22, 2024

null,San Diego,CA

1000 20TH ST, PO BOX 10, BELLINGHAM, WA 98227-0010 USA

SPIE-INT SOC OPTICAL ENGINEERING

In the study, we first introduce a novel AI-based system (MOM-ClaSeg) for multiple abnormality/disease detection and diagnostic report generation on PA/AP CXR images, which was recently developed by applying augmented Mask R-CNN deep learning and Decision Fusion Networks. We then evaluate performance of MOM-ClaSeg system in assisting radiologists in image interpretation and diagnostic report generation through a multi-reader-multi-case (MRMC) study. A total of 33,439 PA/AP CXR images were retrospectively collected from 15 hospitals, which were divided into an experimental group of 25,840 images and a control group of 7,599 images with and without processed by MOM-ClaSeg system, respectively. In this MRMC study, 6 junior radiologists (5 similar to 10yr experience) first read these images and generated initial diagnostic reports with/without viewing MOM-ClaSeg-generated results. Next, the initial reports were reviewed by 2 senior radiologists (>15yr experience) to generate final reports. Additionally, 3 consensus expert radiologists (>25yr experience) reconciled the potential difference between initial and final reports. Comparison results showed that using MOM-ClaSeg, diagnostic sensitivity of junior radiologists increased significantly by 18.67% (from 70.76% to 89.43%, P<0.001), while specificity decreased by 3.36% (from 99.49% to 96.13%, P<0.001). Average reading/diagnostic time in experimental group with MOM-ClaSeg reduced by 27.07% (P<0.001), with a particularly significant reduction of 66.48% (P<0.001) on abnormal images, indicating that MOM-ClaSeg system has potential for fast lung abnormality/disease triaging. This study demonstrates feasibility of applying the first AI-based system to assist radiologists in image interpretation and diagnostic report generation, which is a promising step toward improved diagnostic performance and productivity in future clinical practice.

Artificial intelligence Radiograph interpretation Multi-abnormality detection Automated diagnostic report Observer assessment study

其他

英语

CPCI-S ; EI

Shenzhen Science and Technology Program["KQTD2017033110081833","JSGG20201102162802008","JCYJ20220531093817040"] ; Guangzhou Science and Technology Planning Project[2023A03J0536]

Computer Science ; Radiology, Nuclear Medicine & Medical Imaging

Computer Science, Artificial Intelligence ; Computer Science, Theory & Methods ; Radiology, Nuclear Medicine & Medical Imaging

WOS:001208134600031

Web of Science

1.Shenzhen Zhying Med Imaging Co Ltd, Shenzhen, Peoples R China
2.Peking Univ, Dept Radiol, Shenzhen Hosp, Shenzhen, Guangdong, Peoples R China
3.Southern Univ Sci & Technol, Shenzhen Peoples Hosp 3, Natl Clin Res Ctr Infect Dis, Dept Radiol,Hosp 2, Shenzhen, Guangdong, Peoples R China
4.Univ Oklahoma, Sch Elect & Comp Engn, Norman, OK USA
5.NIH, Natl Lib Med, Bethesda, MD USA
6.Univ Chicago, Dept Radiol, Chicago, IL USA
7.Caddie Technol Inc, Potomac, MD USA
8.Capital Med Univ, Beijing YouAn Hosp, Dept Radiol, Beijing, Peoples R China
9.MS Technol Corp, Rockville, MD USA

Guo, Lin,Cheng, Guanxun,Wang, Lifei,et al. Using Artificial Intelligence for Chest Radiograph Interpretation: A Retrospective Multi-reader-multi-case (MRMC) Study of the Automatic Detection of Multiple Abnormalities and Generation of Diagnostic Report System[C]. 1000 20TH ST, PO BOX 10, BELLINGHAM, WA 98227-0010 USA:SPIE-INT SOC OPTICAL ENGINEERING,2024.