Recently, Josef Murad, the Chief Executive Officer (CEO) of Byon8, a global med-tech solutions provider from Sweden, was in Kampala to sign a Memorandum of Understanding with Victoria University and One World Health, a US based organization, that will see the trio run a pilot project of an artificial intelligence software that helps health workers determine a patient’s diagnosis by imputing the patient’s symptoms in Masindi district, western Uganda.
In this elaborative exclusive interview, Murad explains who they are, their software AITOPYA, working in rural Masindi and the future of technology in health care services provision. Below are excerpts from the interview.
Tell us about Byon8 and the partnership you have signed with Victoria University and One World Health?
Byon8 is a Swedish med-tech company developing intelligent technology for the health care sector – in Sweden and globally. The company works to develop the next generation of health care using technologies such as Artificial Intelligence (AI), machine learning and blockchain, with the dual goal of improving both quality of care and efficiency in the delivery of health care services.
The flagship product AITOPYA, an AI software, supports health care providers in diagnosing patients by using AI technology with the dual goal of both improving quality of care and save time and resources in primary heath care settings.
With the goal of improving operations at OWH, Byon8 will implement AITOPYA, a diagnostic tool, at OWH three clinics in and around Masindi. This will be done through following the plan laid out.
BYon8 and AITOPYA will install the needed hard and software at the clinics then the staff at the health care will be introduced to AITOPYA through a two day training course.
Victoria University, with a strong health sciences department, will lead the work of evaluating the use of an AI digital tool for increasing efficiency and improving quality of care.
How does this AITOPYA diagnostic software work?
Basically the software works in such away that either the health care professionals or patients can input their data. This means that if for example you are at home and you feel that you need health care, you can bring your smartphone or PC and start typing in symptoms and medical history.
This information will then be sent to a health professional, for example at a clinic. They will get an evaluation of your information say what sort of diagnosis that might be and what sort of problems you might have and so on.
Then, they can recommend where to go for treatment. You don’t have to necessarily go to a clinic to get a recommendation, you can get it remotely. Of course you can do it at the clinic or anywhere.
The most important thing is that the data is calculated and is presented to the actual clinician. What the clinicians chooses to do with the information is up to them. They can call you to their clinic or have a chat with you on video for consultation.
The other choice is the AI itself actually gives you a recommendation if you don’t have access to a clinician. Instead of Googling the symptoms, you can actually use this software.
How reliable and accurate is this AITOPYA software since it is not a human being?
Of course it needs a lot of validation and that is why we are having this collaboration with Victoria University. We want to validate the data we are using. We are continuously updating our database. We are continuously making improvements.
The software has a diagnostic accuracy of around 92 percent which is very high with this kind of softwares. But as we said, it is also a support system. It is not there to confirm the diagnosis.
It is there to help the physician or clinician to actually diagnose you. It can never decide that this is the diagnosis, it just gives you suggestions.
If I input my data, in this case my symptoms if I am a patient, how long will the clinician get this information?
In a second. It is synchronized (between the patient and health care provider) and cloud based. As long as you have internet connection, you will get an assessment very fast.
You are working in rural Masindi district, it is not an urban area like you see Kampala and in developing countries like Uganda we have internet infrastructure challenges, connectivity is not reliable, cost of smartphones and computers is high, how effective will the software be in a place were ICTs are not common and internet is not reliable?
We are operating in a remote area like Masindi because it is part of the strategy we use in other areas. When you are doing a pilot project, a study in a remote area like Masindi, you can get a much small controlled group of respondents.
And since it is a small area you get to know what sort of patients they are; which makes it easier to assess and see what the actual results of diagnosis are.
In Kampala, there would be no such control because there would be a much higher demand or higher number of patients than in Masindi. We will do a pilot project here in Kampala but that will come later. Our strategy is to come into Kampala and release the product here.
On the smartphone and internet issue – we are not letting patients fill in this data by themselves in Masindi. In Masindi, we are letting the health professionals to fill in this information.
This is because the digital literacy and access to smartphones is not as good as Kampala. So you have to be strategic. In Kampala, we will let more patients fill in their data.
How different is this software from other softwares in the market?
The software is called AITOPYA because it is based on Artificial Intelligence (AI). The difference is in how we have constructed the actual AI. We have something we call the rule based model and the machine learning model.
The machine learning is usually what you connect to the AI and learn from the patient that is consulting. That is something that all competitors have but we have combined that with the rule learning model and within medicine there are a lot of rules.
Combining those rules with machine learning makes it unique and similar to how physicians learn these diagnosis. That is the main difference from our competitors.
Sweden, where you come from, and Uganda have different illnesses, how is this working out in Sweden, for example we have diseases here that don’t exist in Sweden?
When working in Uganda, we try to implement local diseases like malaria which we might not have in Sweden. The good thing with the software is that you can modify it depending on each country needs. You can tailor make this software to each country.
What is the future like for med-tech and products like Byon8’s AITOPYA?
It is a very expensive market, growing like never before especially within AI. If you look at AI technology within the health care sector, it is going from a million dollar market to a billion dollar market in just a few years so it is growing rapidly. That is why we are very aware that we have to make an aggressive global expansion with this sort of product to make sure we can use it in as many countries as possible.
Does that come with affordability especially on the side of the user or patients?
If you look at the market right now, there is a high demand for this sort of product but all products need some sort of validation and this is why it costs so much much money, to validate everything.
We have been developing this for the last six years – the algorithm, the AI – and we are at a place were we feel comfortable that we have validated enough, that it can go to the market.
This interview first appeared on HEALTH NEWS UGANDA here