Cara Murez HealthDay Reporter
(Health Day)
TUESDAY, April 18, 2023 (HealthDay News) — Artificial intelligence (AI) could help clinicians diagnose tuberculosis in parts of the world where radiologists are in short supply, a new study says. Research suggests.
AI software has successfully identified tuberculosis (TB) from cell phone pictures of chest x-rays, researchers reported at a European conference this week.
Dr Frauke Rudolf, Research Leader, Department of Infectious Diseases, Aarhus University Hospital, said: ,Denmark.
Chest X-rays play an important role in detecting tuberculosis in patients who are unable to produce good quality sputum (sputum) samples for microbiological analysis. Using software to help diagnose conditions based on x-rays can help areas with limited resources and few radiologists.
Rudolf’s team set out to determine the accuracy of this method. They compared the performance of his AI software in evaluating chest radiographs to that of his two Ethiopian radiologists with different levels of experience.
AI was given a mobile phone picture of a non-digital chest radiograph.
Of the 498 patients, 11% were diagnosed with TB, 41 were diagnosed clinically, and 16 were diagnosed by PCR testing, the study authors note.
The software performed as well as or better than trained radiologists in identifying PCR-confirmed cases. It correctly identified 75% of all PCR-confirmed cases and approximately 86% of non-TB cases.
By comparison, evaluation by inexperienced radiologists correctly detected about 63% of PCR-confirmed cases and correctly identified almost 92% of people without TB.
Experienced radiologists correctly selected 75% of PCR-confirmed cases and 82% of non-TB cases.
“With an estimated 3 million undiagnosed cases in 2021, there is an urgent need to develop new strategies and technologies aimed at improving tuberculosis detection in low-resource, high-incidence settings. Yes,” Rudolph said in a news release from the European Clinical Council. Microbiology and Infectious Diseases (ECCMID).
Tuberculosis is a leading cause of death and disease worldwide, killing 1.6 million people each year. It is her 13th leading cause of death worldwide and her second leading cause of infectious death after COVID-19.
“In resource-poor areas with high rates of tuberculosis and a shortage of radiologists, a mobile phone can take a chest x-ray and send the image for remote analysis by AI.” says Rudolf. “This will allow us to read more chest x-rays properly and, more importantly, diagnose more tuberculosis cases.”
The study was to be presented at ECCMID in Denmark on Monday. Results presented at medical conferences should be considered preliminary until publication in a peer-reviewed journal.
The US Centers for Disease Control and Prevention has more information on tuberculosis.
Source: European Conference on Clinical Microbiology and Infectious Diseases, News Release, 17 April 2023
Copyright © 2023 HealthDay. All rights reserved.
