Researchers tested an artificial intelligence tool to identify and diagnose abnormal chest x-rays in a clinical setting accurately. This will speed up the process to assist health professionals and reduce their workload.
(Photo : ALBERTO PIZZOLI/AFP via Getty Images)
Italian Professor Roberto Mezzanotte and a nurse nurse view the lung x-rays of a Covid patient on October 13, 2021 at the Casalpalocco hospital, south of Rome.
AI-Diagnosing Tool
Doctors need to identify an abnormal chest as soon as possible to know if a patient is suffering from lung cancer, emphysema, pneumonia, tuberculosis, heart failure ad more chest-related problems. Additionally, there is a growing demand for medical imaging and a global shortage of trained radiologists.
Fortunately, researchers studied a new artificial intelligence tool that is claimed to speed up the process of diagnosing chest and lung problems. Interesting Engineering reported that the tool's accuracy is at par with clinical board-certified radiologists.
Department of Radiology at the Herlev and Gentofte Hospital Doctor and Co-Author Louis Line Plesner, M.D led the whole team. The claims are backed by the findings from an experiment with 1,529 patients involved, consisting of outpatients, in-hospital patients, and emergency patients.
Testing the Tool
All of these patients' chest x-rays were studied by the researchers and the AI program assorted some x-rays as "high-confidence normal" and "not high-confidence normal." For comparison, the performance comes close to three medical board-certified chest radiologists. The two were used as the reference standard while the other one was used in cases of disagreements.
Eureka Alert reported that the AI tool was very careful with abnormal x-rays with a sensitivity of 99.1 percent. Plesner stated, "The most surprising finding was just how sensitive this AI tool was for all kinds of chest disease. In fact, we could not find a single chest X-ray in our database where the algorithm made a major mistake."
Despite these, the AI tool identified a negative finding after diagnosing the chest x-ray of a 44 -year-old patient with pneumonia as normal. But most of the time, the tool did excellently to identify problems. Out of 429 429 chest x-rays, human radiologists stated that the AI program classified 120 as normal.
Also Read: Babies Beat AI in 'Commonsense Psychology', According to New Study
Plesner added that the AI tool automatically checked nearly 8 percent of all the x-rays they examined through the study, which is very important since every small percentage of automatization can save time for health experts and prioritize other matters during the testing process.
Science Daily reported that the team sees the AI tool's potential as it spots abnormal x-rays very quickly. This can speed up the traditional diagnosing process for chests at hospitals. But for this to release. Several studies were required to test the program to check it on a large scale and enhance the system.
"Chest X-rays are one of the most common imaging examinations performed worldwide. Even a small percentage of automatization can lead to saved time for radiologists, which they can prioritize on more complex matters," Plesner noted.
Relate Article: AI Predicts Alzheimer's Risk With Above 90% Accuracy
