‘Cost of a blood test can come down to Rs 10’

Suman Chakraborty is a professor at the Indian Institute of Technology, Kharagpur. A recipient of the Infosys Prize 2022 in Engineering and Computer Science, he talks about his latest creation that could reduce the cost of a blood test to Rs 10 and allow it to be conducted without a medical laboratory. He also explains other technologies that he and his team are working on that could make rural healthcare much more efficient at spotting diseases early on. Q: You have been trying to revolutionise rural healthcare. Tell us more…A: Healthcare is currently a one-dimensional practice which implies that there will be a doctor, a structured medical system and people who will be able to access and afford these.The use of technology has enabled us to create a healthcare model which will basically be able to serve or cater to the needs of all the people exactly in the same way as it might to those who can afford [it].I am looking to put science and technology together and develop technologies which are simple to use and are low cost, without sacrificing accuracy. medium98156521Q: What are the new technologies that you are using?A: My domain is primarily diagnostic technology. If you go to any place where people do not have good diagnostic facilities, infectious diseases remain undetected.My interventions have primarily been targeted at detecting these diseases at an early stage.One type of test that we commonly do is a routine blood test. There are certain basic kidney function tests, liver function tests, complete blood count tests among others. All of these tests have certain common features. They are traditional, where a phlebotomist [professional trained to draw blood] comes in and collects the blood. The sample is then transferred to a laboratory which has a high-end machine and the test is conducted. We then get our results from a pathology lab.My tests are simpler without expensive requirements.Q: How practical are the tests?They are simple and feasible. Let me tell you how it works. For instance, in the haemoglobin test, you put a few drops of blood on a paper that already has the reagent.The paper absorbs the blood. The design of the paper is such that the blood will move to the portion where the reagent is present.Depending on the time needed for the reaction, within a few minutes a change in colour will be noticed on the paper.So, for someone who has a high level of haemoglobin, the colour will be a strong bluish green and if the patient is anaemic, it will be a lighter version of the colour. Now in case of a blood sugar test, depending on the content of sugar in the blood, a reddish brown colour will be noticed on the paper. The shade of the colour will depend on the amount of haemoglobin in the blood.Now, we need to convert this test into an objective test and we can do so by using a simple smartphone.We already have the data of a good number of test cases which we have stored. When the colour of the paper is uploaded on the smartphone, we use image analysis algorithms. These can be implemented through a mobile app.Suman Chakraborty and his work on the field mediumhttps://www.youtube.com/watch?v=uqMTTHxUHY8Q: What happens next?A: So, we have a database on the relationship between the colour and say, haemoglobin. That is standardised and the information is then fed into the mobile app. When the data of a particular colour is fed into the mobile app, it uses machine-learning technology to make calculations.The calculations are based on how the app has been taught to correlate colour and details of haemoglobin.Once the data is fed, the app will instantly calculate and tell us the value of haemoglobin, blood sugar etc. This is standardised to such an extent that we can say whatever the report we get from the laboratory, there is no more than 5% difference of that value from the value of a test that is done outside. medium98157051Q: What does this cost? And is it being implemented?A: I will tell you the cost as per actuals. For every test, you need a strip and a small volume of reagent. That is it.So, overall, for haemoglobin test or glucose test, it is likely to be about Rs 1.50 per test per patient.Now if we put in a business supply chain and other auxiliary costs, it will be amplified to a maximum of Rs 10.It is being used in a field trial mode, because there are companies which have taken it off for commercialisation. Also, if you consider the demand from places like Africa, where countries do not have a structured health system available, I think the requirement is huge.Q: Where is it being used now?A: It is mostly being used in West Bengal. We have our collaborator, Dr Satadal Saha. He is a fellow of the Royal College of Surgeons, a surgeon and a rural healthcare entrepreneur. He runs a foundation for innovations in health.This foundation is mandated by various governments to run rural healthcare centres with the help of certified frontline workers. These frontline workers are similar to Asha workers, but they are being trained and certified under the National Skill Development Mission programme.These people are primarily rural girls who work in the rural centres, before the technologies were introduced.So, what they used to do is when a patient goes there, they would feed information into a tablet, enter the various issues of the patient and then transfer the data to a remote doctor who would then prescribe the medicines.Where the system failed is that the remote doctor would recommend some diagnostic tests and medicines during the first consultation. Though medicines are readily available nowadays, that is not the case for diagnostic tests.Even a basic sugar test will require at least Rs 50-100, which most people can't afford.Q: I have read that you're developing a kind of screening test for oral cancer?A: Oral cancer is one of the top five killers in India. There are two things that may lead to oral cancer – tobacco chewing or previous HPV [human papillomavirus] infection or other issues unconnected with tobacco.In India, oral cancer is a public health challenge and it is primarily because of tobacco consumption. It starts as pre-cancer, which can be a prelude to cancer.What we see initially are lesions appearing in the mouth. They may be very very minute, small patches [which could go] undetected or unaddressed over a period of time [until] the patient’s condition may get very difficult — like the patient not being able to open their mouth.The big challenge is that in rural areas there are hardly any dentists. So, in most cases it remains undetected.We have developed a technology which is like a small torch. You focus that torch inside your mouth to the place where you have the suspected lesion for 30 seconds and it takes an image.This is not a standard image, but a thermal image. There are additional sensors in the device that take additional information from inside the mouth like humidity.There is a computer algorithm by which we convert that information to another set of information or data that can tell us how the blood is flowing through the blood vessels in the suspected location.The physiology is that when you have an abrupt increase in blood flow in an area, it may imply that new blood vessels have been created to support the advancement of cancer. This is known as angiogenesis. So, cancer has started.By digitising that information and analysing the app, again in about 10 minutes, we are able to give an interpretation of three very precise classifications we are currently making pre-cancer, cancer or no cancer. This is done by artificial intelligence, machine learning and image analysis.It does not require individual human analysis or interpretation. It is all done by the software. This is something that enables us to do the screening at a clinic or a grassroots level, where no other facility is required.You can listen to the podcast here