ADVERTISEMENT

Xray Setu: Can AI Help Doctors Detect COVID Using WhatsApp?

X-Ray Setu: Can this AI driven X-ray sent over WhatsApp help doctors detect COVID-19?

Updated
<div class="paragraphs"><p>Can Xray Setu help doctors detect COVID-19 infection in rural areas?</p></div>
i

The government has launched a new AI driven bot called Xray Setu to check for COVID-19 infections in early stages.

This service has been introduced especially for doctors working in rural areas where COVID-19 testing has been sparse owing to the lack of testing facilities and provisions.

Currently, the platform is available in its beta version. i.e. a version that is launched before the formal launch.

Xray Setu, using Artificial Intelligence (AI) technology, can help in early detection of COVID-19 in patients.

A picture of a person's chest X-ray is sent to the Xray Setu platform using WhatsApp. The bot then reveals in just a few minutes whether the person concerned is infected with COVID, and which part of the lungs is affected

But will it really help patients and doctors in rural areas? Before we explore this, let's understand how it will work.

ADVERTISEMENT

How Does Xray Setu Work?

Firstly, go to xraysetu.com.

You must then click 'Try The Free Xraysetu Beta'. It's a service free.

This will open a new page, which can be connected to a chatbot on this page.

A doctor will take a picture of the suspected patient's X-ray and send it to Xray Setu's WhatsApp number, +91 8046163838, via a chat bot.

A technician and artificial intelligence (AI) analysis will review the images to generate a report.

This will be followed by a 2-page diagnosis report sent back to the doctor in a few minutes.

What are the Key-Highlights?

The Xray Setu service analyses X-rays using artificial intelligence based special machine learning and deep learning algorithms.

The doctor can then read the report and give proper advice to the patient.

  • As it is meant to be used in rural areas, Xray Setu has been designed to read and screen digital as well as analog (Analog) X-rays.

  • The AI is capable of reading and processing low resolution images of X-rays.

  • It is said to have a sensitivity of 98.86% and specificity of 74.74%.

  • It can detect not only COVID-19, pneumonia but 15 types of lung abnormalities.

Xray Setu: Can AI Help Doctors Detect COVID Using WhatsApp?

(Photo: https://www.xraysetu.com/)

According to the website, the major motive behind developing is the drawbacks of RT-PCR tests. Apart from being expensive, RTPCR test reports can take a long time to arrive.

In the meantime, a patient's condition can worsen, and can even turn fatal. Getting CT scans done quickly right then isn't an option for many either.

In such situations, Xray Setu could help immensely.

What Do Doctors Think of It?

Though idea of it sounds like a Godsend, especially at a time when rural India lags dramatically behind when it comes to COVID-19 testing, treatment and vaccination, doctors are skeptical.

For one, they believe that the service is hardly available in rural areas. At the same time, whether the service will in reality be able to realise the purpose with which it was made has also been questioned.

Speaking to FIT, Dr Abhishek Ranjan consultant physician and former AIIMS Patna doctor in Gopalganj district of Bihar, says this,

“Doctors don’t know about it. There are daily training and discussions of practitioners in our locality, but many are not aware of it. Even in a city like Patna, government hospitals or private practitioners are not aware of it, so what can they say about remote rural areas.”
Dr Abhishek Ranjan, consultant physician, Gopalganj district, Bihar

He also adds making preparations to streamline it before launching it should be focused on.

However, they do not deny the usefulness of X-rays in the second wave of covid in India.

He explains, saying the 'delta' variant of COVID-19 found in India has been damaging the lungs instead of staying around the nose and throat.

ADVERTISEMENT

"In this case, the damage to the lungs in the X-ray is known just within 3-4 days after the symptom appears and has helped many patients get timely treatment in areas where ICU hospitals are not there."

But, having said that, X-rays themselves have certain drawbacks.

  • The first problem is that x-rays have to be digitised. Analog X-rays performed by an old machine are quite blurred. Markings are only vsible in very serious cases.This makes it difficult to catch primary infections.
  • How a technician carries out the Xray can influence the results and sometimes cause errors. Moreover, the patient has to hold his breath and give exposure, and a hurried X-ray can cause some odd shadow (image) that can be misinterpreted as Covid.

Dr Ranjan further adds,

“AI systems are formula based. They can catch things on a large scale, but which are shadow mixes, i.e., if someone has TB, bacterial pneumonia or viral pneumonia, AI can’t distinguish it. It can catch 70-80% lung damage cases, but not less lung damage than that.This is where the doctor’s experience comes in and a clinical examination of the patient is required.”
Dr Abhishek Ranjan, consultant physician, Gopalganj district, Bihar

According to Dr Ranjan, one of the serious consequences could be that a person may have to bear the burden of taking medicines unnecessarily if he receives a false positive report.

"There is a difference between claims and practical circumstances. These claims can work if we have logistics to support it," says Dr Ranjan.

“This may work in the future but before that such an infrastructure has to be created in primary health centres (PHC). There should be X-ray centres in villages. Rural hospitals do not have enough staff, which they do not have enough information. Without basic infrastructure, the benefit will be negligible.”
Dr Abhishek Ranjan

Dr Abhishek Shankar, Assistant Professor in Radiation Oncology Department, Lady Hardinge Medical College, Delhi, has a similar opinion.

"Accuracy is a big issue in such reporting of the disease," he says. "This technique cannot calculate the severity of the disease. In serious cases, hospitalization of the patient should be a priority instead of sending x-rays to the doctor," he says.

“If the doctor has an X-ray report, they will be able to read it themselves, why would they send it to read it on whatsapp. Will the doctor be able to ask for the patient’s history and advise the patient to be admitted or referred?”
Dr Abhishek Shankar, Assistant Professor in Radiation Oncology Department, Lady Hardinge Medical College, Delhi

He also adds that in a country where pulmonary infections like TB are quite common in rural areas, AI screenings like this can create confusion.

He also points out that x-ray is not a diagnostic tool for COVID-19.

Dr Shankar isn't optimistic of the impact that something like Xray Setu when it comes to changing the ground reality.

ADVERTISEMENT

"More patients have died due to lack of oxygen and non-availability of beds in the hospital. Strengthening primary health care will help more. Villages should have X-ray machines. We need convenience and doctors. the condition of the PHC should improve. A facility like x-ray bridge has little scope for changing something on the ground." he says.

He adds, "this will only help in data gathering and research."

How Accurate Is Xray Setu in Detecting COVID-19?

When it comes to precision, RTPCR tests are considered the 'gold standard' in COVID-19 testing. But in rural India, the big challenge isn't so much its accuracy as it is its availability on time.

According to Xray Setu's website, its model has been evaluated with a COVID-19 data set of NIH, UK.

The Xray Setu aims to give doctors another screening option for Covid-19 and help in early intervention to prevent lung damage and potentially save lives.

The platform has been developed by ARTPARK (AI & Robotics Technology Park), an NGO set up by the Indian Institute of Science (IISc), and the Department of Science and Technology (DST), Government of India, in collaboration with Niramay (Niramai), a health tech startup.

ARTPARK is a non-profit company supported by IISc and AI Foundry, funded by DST (Department of Science and Technology) and GoK (Government of Karnataka) for its work in the sector.

(Subscribe to FIT on Telegram)

Published: 
ADVERTISEMENT
Stay Up On Your Health

Subscribe To Our Daily Newsletter Now.

Join over 120,000 subscribers!
ADVERTISEMENT