Dear Digital X-Ray Insider,
Nobody likes uncertainty in radiology reports, but could uncertainty have a more insidious effect beyond just annoying referring physicians?
A group of researchers from New York thinks it might. They looked at the impact of uncertain findings on the performance of chest radiography for pneumonia. Diagnosing pneumonia has historically been tricky for x-ray, with CT considered the reference standard.
The researchers found that previous research on the value of radiography for pneumonia has discarded uncertain findings, leaving a less-than-accurate perspective on x-ray's performance. When uncertain findings are added into the mix, x-ray records lower sensitivity and specificity. Read more in this edition's Insider Exclusive.
In other news in the Digital X-Ray Community, the annual RSNA meeting concluded earlier this month and a wealth of new research was presented in Chicago involving x-ray. RSNA speakers discussed the results of a study that used radiography to document injuries sustained by electronic scooters, as well as research that used an artificial intelligence (AI) algorithm to predict a patient's future healthcare costs by analyzing their chest radiographs.
Prior to RSNA, researchers from South Korea discussed a deep-learning algorithm they developed to detect lung cancer on chest radiographs, while another Korean group successfully used AI to detect significant abnormalities on chest x-rays in patients presenting to the emergency department.
We hope you've enjoyed reading these articles over the course of 2019 as much as we've enjoyed bringing them to you. Best wishes for a safe and happy holiday from everyone at AuntMinnie.com!