In a bid to provide reliable healthcare delivery in remote areas at low cost, faculty from IIT Hyderabad has recently proposed a two-tier, telecardiology framework, where ECG records can be transmitted even when available resources such as power and bandwidth are limited. Prof Soumya Jana, Department of Electrical Engineering, IIT Hyderabad, shares the idea and ideology behind this cardiology framework, with Raelene Kambli
Tell us about the two-tier telecardiology framework that you have developed?
Typically, a patient is transferred to the diagnostic centre where experts look at the signal, evaluate it (if the signal is normal or abnormal) and make a recommendation. So that’s kind of a one-step process and doesn’t take into account any resource constraint (like power or bandwidth shortages). However, in remote areas, people may not be on the power grid and may not have access to a large bandwidth- phone or internet. In such cases, this framework may not be suitable for remote locations.
We have addressed this using the two-tier framework. As a first step, we picked 20 per cent of the samples. These samples helped us reconstruct the original signals. With reasonable accuracy of 95 per cent, the reconstruction helped us save power in two ways: by collecting 20 per cent signal (requiring 20 per cent of total power) and second by transmitting these signals we cut the process by 1/5th. In this way, we essentially can operate with 20 per cent of capacity.
In the second step, we transfer the signal directly to the local sub centre (instead of the diagnostic centre) for classifying the signal into normal and sub normal categories. If the signal is normal, the signal is not transferred any further. Only sub normal signals are transferred, saving a lot of time and resources.
How is it developed? What is the mechanism used to develop it?
What we are doing now is we are giving mathematical approval that works. For example; something needs to be transmitted; we transmit only the relevant part of it. That partial part is good enough.
This is called comparative sensitive technique, it actually demonstrates that this is indeed possible. Now, basically we demonstrate the mathematical terms and delegate it using some of the standard database that is available. We have used one of the databases hosted in the MIT, MIT BE database, to demonstrate the proof of concept.
Has it been tested so far? By whom?
Testing actually needs to happen at several levels. First level of testing is algorithmic level. We have to demonstrate the algorithm correct, that part has already been done.
The second level of testing is against the VL database to show that your algorithm is correct and works with real signals. So, that part has been done only using MIT BE data set.
The third thing is to take it to the field to test. This is not yet been done by us as the field version is in the process of manufacturing. One version of the system has already been manufactured, but not the field version. As soon as the product is ready, we will test it on field.
What are the benefits of this technology?
This system has been developed for the population in the remote parts of India and the world that cannot take advantage of regular healthcare. For example; somebody who lives in remote Arunachal Pradesh may not have access to medical services. Typically, the most basic healthcare system they have may be a health centre or a sub-centre. In some other places they may not have a doctor but only a trained medical worker.
So, we analysed the published data from government on Arunachal Pradesh; we found that people travel 9 km to access a sub centre. They do not have a doctor or proper treatment; they only have access to medical workers and some basic medicines. In addition, they do not have access to power or adequate communication systems like landline or internet facility. We can overcome these issues as we minimise using both power and communication bandwidth.
Who all are involved in developing this technology? What motivated you to develop it?
We have a team at IIT Hyderabad. Dr Shastri, a colleague from the mathematical department and I are from the facility side. Research scholar, Dr Sandeep Chandra and Rupak Tandoli have been instrumental in working out the integrity of technology. This is the core team. The other faculty members like Swamy Dutt help in manufacturing the device and chips. This team is completely dedicated to translating this vision into actual product prototype.
How does it work?
As mentioned above, the device has the capability of verifying sample signals as normal or abnormal. Using minimal power and bandwidth, the signals can be transmitted to the local sub-centre. The sub-centre will evaluate this data according to the algorithm. If the signal is abnormal, it will be sent to the diagnosis centre. The technicians at the diagnosis centre can look at the reconstructed signal acquired by other devices.
After this, it’s easy to diagnose the issue and send back the diagnosis through SMS. This saves a lot of time so that patients with serious issues can immediately meet a doctor to get treatment.
How can it be utilised to increase access to cardiac diagnoses, even in rural India?
It will be a boon to people in rural India who don’t have access to basic medical care. This solution will give them access to quick diagnosis and medical treatment if needed.
Are you planning to tie-up with the government to take this technology to the hinterlands of India?
This project is funded by the Ministry of Communication and Information Technology (MCIT) under the Department of Information Technology. We are also tying up with healthcare providers and companies. We are still working out the way we would like to take it to ground on a large scale.
What are your future plans for this product?
We are exploring more government support and generating industrial interest so that this actually can be made operational and sustainable.
Are there any more innovations in the pipeline?
Yes, there are many innovations, but in terms of cardiology we need to ensure that the targeted functionality is fully realised. For this solution specifically, we are making further innovations on the algorithm, the technology behind it. We are constantly improving the enabling technology. Our entire focus is on realising the functionalities, reducing cost, and reducing utilisation of resources like power and bandwidth.