Kigali: Rwandan startup using artificial intelligence to scale up Covid-19 testing

As the world awaits medical breakthroughs to understand and curb the Covid-19 pandemic, bets are that solutions will come from developed markets.

Kigali: Rwandan startup using artificial intelligence to scale up Covid-19 testing

However, Insightiv, a Rwandan startup in artificial intelligence is staying up in an attempt to make a breakthrough that could change the diagnosis process trimming the process to minutes and increasing accuracy.

These could be a game-changer, not only in Rwanda but across the world.

Currently, Covid-19 tests are largely conducted using “swab” test, relying on Reverse Transcription Polymerase Chain Reaction (RT-PCR).  The process is somewhat time consuming and has been said to have about 80 per cent accuracy sensitivity according to the World Health Organisation.

On the other hand, CT scan exams have been found to have 98 per cent sensitivity but remain expensive and inaccessible for many. In Rwanda, a CT scan exam is estimated to cost around $150 and there are only about 10 CT machines in Rwanda.

Audace Nakeshimana, the Founder of Insightiv, told The New Times that they are building on the accuracy of CT scans and previous findings from other diseases which have shown that if radiologists study the pattern of the disease from CT, they can then eventually be able to diagnose the disease by only depending on chest x-rays.

X-rays would be more suitable as they are cheaper in comparison to CT scans and are accessible in Rwanda since they are available in all District and referral hospitals in Rwanda.

“We are therefore studying the capabilities and limitations in using chest x-ray for Covid-19 detection, given that x-ray is more accessible and less expensive. Insightiv is also developing and training artificial intelligence (AI) algorithms that can be used for assisting in automated detection of Covid-19 from medical images, which could speed up the detection and isolation of COVID-19 suspects, hence helping to reduce community exposure and to prevent infections,” Nakeshimana told The New Times.

To improve Rwanda and Africa’s resilience on the Covid-19 pandemic, a team of AI Researchers, Radiologists and Engineers at Insightiv is exploring medical imaging and Artificial Intelligence (AI) for faster, safer and cost-effective ways of detecting Covid-19 early and reducing community exposure to the disease.

From their small office at Makuza Plaza, the team led by Nakeshimana is working to leverage Teleradiology and AI to ensure any number of medical images could be collected for different patients who are suspected of Covid-19 in Rwanda.

The system under development will see artificial intelligence pick up cases it suspects of Covid-19 so that they can be brought to Radiologists’ attention faster, and radiologists would instantly diagnose all the images and decide who has or doesn’t have Covid-19.

The innovation has already gained acceptance in some European markets with the start-up signing agreements with healthcare facilities in Europe who are keen on taking up the technology.

“The agreements came about when we were talking with different healthcare facilities to understand their challenges and how our system can create value for them. Some hospitals, especially in developed countries in Europe or Americas, have solutions for Teleradiology, but they are still trying to find ways that artificial intelligence can help make their work efficient. Our European clients are interested in whether we can train AI algorithms to detect disease patterns, including Covid-19, on medical images that they have,” Nakeshimana said.

Is the solution applicable locally?

On whether the solution would be feasible and applicable locally, the computer science engineer says that their solution is tailored to work in an array of healthcare providers ranging from clinics to the most modern facilities.

“The system is designed to be compatible with international medical image formats. One challenge that we had when we first tested the system on medical images from Rwanda is that some of the machines in Rwanda generate older image formats that we didn’t support back then, and we ended up developing an additional software package that allows decoding images from machines used in Rwanda. Today, local compatibility with digital machines is not a problem anymore,” he said.

Insightiv was launched pre-covid19 improve access to medical diagnostics services in Africa. Nakeshimana is a recent Massachusetts Institute of Technology graduate with a bachelor’s degree in Computer Science and Engineering (CSE), a minor in Economics and an Advanced Research Certificate in Machine Learning and Artificial Intelligence.

Prior to founding Insightiv, he worked in Silicon Valley and at the Massachusetts Institute of Technology Labs.

His firm among other things seeks to address shortages of Radiologists in Rwanda, there are about 12 in the country.

Medical imaging using AI is an alternative solution to make radiology services accessible to anyone in Rwanda no matter where they are, and using advanced technology such as AI to reduce patient waiting time.