AI software wins award for improving chest X-ray readings, tackling healthcare disparities

SumiBit AI's AIRead-CXR software has earned recognition at the 11th Human Tech Awards for its innovative approach to chest X-ray analysis. The software aims to alleviate the burden on medical professionals and improve access to quality diagnostics, particularly in regions facing a shortage of radiology specialists.

“

We are including all the diverse postures patients exhibit when undergoing chest X-rays in our analysis scope, and we can handle most of the findings that a radiologist needs to identify.

South Korea experiences a high volume of X-ray and CT scans per capita, with demand expected to rise due to an aging population and advancements in medical technology. However, a scarcity of radiologists presents a significant challenge. SumiBit AI's AIRead-CXR leverages generative AI to address this issue by automatically producing draft diagnostic reports. Radiologists can then review and edit these AI-generated drafts, streamlining the diagnostic process.

Unlike chatbot-style AI, AIRead-CXR analyzes only chest X-ray images, ensuring consistency and reducing variability in results. Bae Woong, CEO of SumiBit AI, highlighted the software's ability to analyze diverse patient postures and cover most findings that radiologists typically identify. He expressed a vision for exporting their proprietary AI models in the medical field.

“

We have a vision of exporting AI models developed in-house in the medical field.

The software has achieved significant regulatory milestones, including South Korea's first clinical trial approval for generative AI medical software in November 2024 and subsequent approval from the Ministry of Food and Drug Safety as a Class 3 medical device in April 2024. Following FDA clinical trial approval in April 2025, this marks a pioneering instance of a standalone generative AI product receiving national medical device approval for its ability to analyze medical images and generate draft reports in Europe.

“

The fact that it can reduce the reading burden of radiologists and provide rapid reading support in areas with a shortage of medical personnel or in emergency situations is outstanding.