Washington: According to a study published in Radiology, a publication of the Radiological Society of North America, an artificial intelligence (AI) tool can successfully discern between normal and abnormal. chest x-rays in a clinical context (RSNA).
Chest x-rays are used to identify a wide range of heart and lung diseases. Atypical chest X-rays can indicate a variety of problems, including cancer and chronic lung disorders.
An artificial intelligence tool that can accurately differentiate between normal and abnormal chest X-rays would greatly alleviate the heavy workload experienced by clinicians. radiologists globally.
“There is an exponentially increasing demand for medical imaging, especially cross-sectional such as CT and magnetic resonance“said study co-author Louis Lind Plesner, MD, of the Department of Radiology at Herlev and Gentofte Hospital in Copenhagen, Denmark, adding: “Meanwhile, there is a global shortage of trained radiologists. Artificial intelligence has shown great promise, but it should always be thoroughly tested before any implementation.”
For this retrospective, multicenter study, Dr. Plesner and colleagues wanted to determine the reliability of using an artificial intelligence tool that can identify normal and abnormal chest X-rays.
The researchers used a commercially available artificial intelligence tool to analyze the chest X-rays of 1,529 patients from four hospitals in the capital region of Denmark. Chest radiographs of emergency, inpatient, and outpatient patients were included. The AI tool classified the X-rays as “high-confidence normal” or “non-high-confidence normal” as normal and abnormal, respectively.
Two board-certified thoracic (chest) radiologists were used as the reference standard. A third radiologist was used in cases of disagreement, and the three physicians were blinded to the results of the AI.
Of the 429 chest X-rays that were classified as normal, 120, or 28 percent, were also classified by the AI tool as normal. These X-rays, or 7.8 percent of all X-rays, could potentially be safely automated using an AI tool. The AI tool identified abnormal chest X-rays with 99.1 percent sensitivity.
“The most surprising finding was how sensitive this AI tool was for all types of chest disease,” said Dr. Plesner, adding, “In fact, we couldn’t find a single chest X-ray in our database where the algorithm made a big mistake. In addition, the AI tool had better overall sensitivity than clinical board-certified radiologists.”
According to the researchers, further studies could be directed toward a broader prospective implementation of the AI tool where radiologists still review reported chest X-rays autonomously.
The AI tool performed especially well in identifying normal x-rays from the outpatient pool at a rate of 11.6 percent. This suggests that the AI model would work especially well in outpatient settings with a high prevalence of normal chest radiographs.
“Chest X-rays are one of the most common imaging exams performed worldwide,” said Dr. Plesner, adding, “Even a small percentage of automation can save radiologists time, who can prioritize on higher issues.” complex”.
