USA: In a diagnostic study published in JAMA Network Open, an artificial intelligence (AI) algorithm may improve the diagnostic performance of radiologists for the detection of pulmonary nodules on chest radiographs versus unaided interpretation.
The study showed that with help from AI-aided interpretation of chest radiographs, four of 9 radiologists had a lower number of missed and false-positive pulmonary nodules.
Chest radiographs are used frequently for early detection of pulmonary nodules despite their inferiority to low-dose computed tomography (CT). This is because of its accessibility, low cost, low radiation dose exposure, and reasonable diagnostic accuracy. Most early lung cancers present as pulmonary nodules on imaging, but on chest radiographs, these can be easily missed when they are subtle, small, or in difficult areas.
Against the above background, Fatemeh Homayounieh, Department of Radiology, Massachusetts General Hospital and Harvard Medical School, Boston, Massachusetts, and colleagues aimed to assess if a novel AI algorithm can help detect pulmonary nodules on radiographs at different levels of detection difficulty.
The study included 100 posteroanterior chest radiograph images that were between 2000 and 2010 of adult patients from an ambulatory health care center in Germany and a lung image database in the US. The images included were selected to represent nodules with different levels of detection difficulties (from easy to difficult), and comprised both normal and nonnormal control.
The presence of 5 findings was recorded by each test radiologist (atelectasis, pulmonary nodules, pneumothorax, consolidation, and pleural effusion). Their level of confidence for detecting the individual findings on a scale of 1 to 10.
The study included images from 100 patients (mean age of 55 years and including 64 men and 36 women).
Following were the study's key findings:
Mean detection accuracy across the 9 radiologists improved by 6.4% with AI-aided interpretation compared with unaided interpretation.
Partial AUCs within the effective interval range of 0 to 0.2 false-positive rate improved by 5.6% with AI-aided interpretation.
Junior radiologists saw greater improvement in sensitivity for nodule detection with AI-aided interpretation as compared with their senior counterparts (12%; 4% to 19% vs 9%; 1% to 17%) while senior radiologists experienced similar improvement in specificity (4%; −2% to 9%) as compared with junior radiologists (4%; −3% to 5%).
"Our study found that a novel AI algorithm was associated with improved accuracy and AUCs for junior and senior radiologists for detecting pulmonary nodules," concluded the authors.
Reference:
Homayounieh F, Digumarthy S, Ebrahimian S, et al. An Artificial Intelligence–Based Chest X-ray Model on Human Nodule Detection Accuracy From a Multicenter Study. JAMA Netw Open. 2021;4(12):e2141096. doi:10.1001/jamanetworkopen.2021.41096