Researchers at IIT Jodhpur have developed an automated Artificial Intelligence (AI) approach for predicting COVID-19 from chest X-rays. Over 2,500 chest X-ray pictures were used in the study, which yielded a sensitivity of over 96.75 percent.
The researchers at IIT Jodhpur have suggested a deep learning-based system called COMiT-Net, which learns the anomalies present in chest X-ray pictures to identify between a COVID-19 impacted lung and a non-COVID affected lung, according to a statement released by the institute.
Several research investigations in COVID-19 detection utilizing X-ray or CT scans have been conducted in recent years, however many of them have failed to give an explainable solution. The proposed study, which may clearly highlight the contaminated zone, is what makes this research special. The technology exclusively interprets data from the lungs.
"Not only can the developed Artificial Intelligence program predict whether the CXR contains COVID-19 pneumonia or not, but it can also find the infected locations in the lungs and explain them," IIT Jodhpur explained.