PGI’s DecXpert will help in rapid detection of tuberculosis

The Department of Pulmonology, in collaboration with Dectrocel, a startup operating under the Software Technology Park of India (STPI), has developed an AI-powered software tool that enables rapid detection of tuberculosis with an accuracy of around 95%, using only the patient’s chest X-rays.

This tool is called DecXpert.

Research on this artificial intelligence algorithm was published in the prestigious journal Nature Scientific and was appreciated by the international team of WHO in the UK.

The authors of this groundbreaking study are Dr Alok Nath, Dr Zia Hashim, Dr Saumya Shukla, Dr Zafar Neyaz, Dr Richa Mishra (SGPGIMS), Dr Ankit Shukla (University of Queensland Faculty of Medicine) and Nikhil Misra (IIT Kanpur).

This AI-assisted software application tool after clinical validation was introduced for patient use by Dr. Alok Nath, HoD, Department of Pulmonology, SGPGIMS. According to Dr. Alok Nath, this software application will have immense utility in remote and underserved areas of the country where about 30 to 40% of TB cases remain undetected, causing rapid spread of the disease and also increasing the burden of the disease.

In our country, about 250,000 cases of tuberculosis each year remain undetected and undiagnosed. This newly invented software tool DecXpert will revolutionize the concept of rapid diagnosis of tuberculosis cases at the point of care.

In this application software, a chest X-ray of a suspected tuberculosis patient is simply transferred and within seconds the diagnosis is displayed in the form of a graph and can be quickly used to make a decision to confirm the diagnosis. The ease of application is optimal because images can also be transferred via a mobile application.