Indian medical scientists based in Kyoto, with the support of a team of a professor and students from IIT Roorkee, have developed an Artificial Intelligence (AI) based software which uses X-ray to detect novel coronavirus in symptomatic and asymptomatic patients. The Osaka-Kobe consulate general of India supported the friendly cooperation for research on the issue. 

The testing software can ascertain results based on X-Rays alone, scanning 100 images in under four seconds, which makes rapid screening possible. Since it is based only on the X-Ray scans, invasive swab tests are rendered unnecessary. The software design is a 2-step algorithm which processes input images sequentially in two streams that can answer Yes/No for COVID and distinguish non-COVID cases such as pneumonia. 

“The approach of this system is very simple. From the real-time perspective, we take the chest x-ray image of the patient and send it to the Deep and Machine Learning Model, which in a fraction of a second, can determine whether the patient is affected with COVID-19 or Pneumonia or is healthy,” says Kyoto-based medical scientist Namasivayam Ganesh Pandiyan, who led the research along with this student Vinod Joseph and IIT Roorkee professor Balasubramanian Raman’s team. 

Although the software has been tested on small sample size, so far the accuracy has remained at 99.96 per cent. While the possibility of a large sample size cutting into accuracy initially is not ruled out, it is ultimately expected to remain in the range of 99.69 to 100 per cent. Currently, the University of Cyberjava in Malaysia is contributing to the team by providing data for the analysis. Further, Pandian said that the software, when ready, will be made available free of cost to those who request the software for on-field validation. “We need large-scale analysis with randomised patient data for validation and future application,” clarifies Pandian. 

As it will take some time for the machine to get intelligent to handle screening patients by itself, in the interim, the software can help authorities to predict and decide who should be tested for the polymerise chain reaction (PCR) for confirmation. The AI software, as it continues to improve, will eventually ease the burden related to the dearth of testing kits and the workforce shortage, perform tests in resource-constrained settings such hinterlands, quantify disease progress and recovery in real-time. As to when the software will be available for use in India. At least three state governments are already in touch with the team in Kyoto.