Findings from a study of a large group of tuberculosis patients from Dhaka, Bangladesh, comparing five commercially available artificial intelligence (AI) algorithms, showed that “all five AI algorithms outperformed experienced certified radiologists and could avoid follow-on Xpert testing and reduce the [number of people needed to test] while maintaining high sensitivity,” according to the study authors.

The researchers identified the five AI companies and products used in the study as “CAD4TB (version 7) by Delft Imaging Systems (Netherlands), InferRead DR (version 2) by Infervision (China), Lunit INSIGHT CXR for Chest Radiography (version 4.9.0) by Lunit (South Korea), JF CXR-1 (version 2) by JF Healthcare (China), and qXR (version 3) by Qure.ai (India).”

The clinical implications are clear: “AI algorithms can be highly accurate and useful triage tools for tuberculosis detection in high-burden regions, and outperform human readers,” according to the study.

Sources:

Zhi Zhen Qin, Shahriar Ahmed, Mohammad Shahnewaz Sarker, Kishor Paul, Ahammad Shafiq Sikder Adel, Tasneem Naheyan, Rachael Barrett, Sayera Banu, Jacob Creswell. “Tuberculosis detection from chest x-rays for triaging in a high tuberculosis-burden setting: an evaluation of five artificial intelligence algorithms.” The Lancet Digital Health September 2021. Accessed September 14, 2021, at https://www.thelancet.com/journals/landig/article/PIIS2589-7500(21)00116-3/fulltext

Morton, Will. “AI could help diagnose TB in developing countries.” AuntMinnie.com. August 27, 2021. Accessed September 14, 2021, at https://www.auntminnie.com/index.aspx?sec=sup&sub=xra&pag=dis&ItemID=133338