As Sweden's support to research in Uganda comes to a close, we interviewed Dr. Rose Nakasi on how her PhD research utilized Artificial Intelligence to detect Malaria in Uganda. This is what she had to share with us.
Can you tell us about who you are and what you do?
My name is Dr. Rose Nakasi and I am a lecturer in Busitema University where I teach computer science. I am also a Research Scientist working within the Makerere University Artificial Intelligence (AI) Laboratory at the College of Computing and Information Science.
Tell us about your PhD program.
My Doctorate research was on Computer Science, with the topic “Automated diagnosis of Malaria: Deep Neural Network”. My research focused on use of artificial intelligence and Mobile smart phones to augment the diagnosis of malaria by microscopy. This was multidisciplinary research including both computer science and health.
I realized that there is a lack of trained lab technicians in most of the health centres, especially in low developed countries like Uganda. My approach was to use Artificial Intelligence to detect malaria parasites from microscopic image data.
The innovation in this research is the coupling of a 3D printed Adapter onto the eyepiece of the microscope to enable attachment of the smartphone. The smart phone then captures images under the microscope. The artificial intelligence models that I developed then learn the characteristics of the malaria parasite, thus enabling it to predict a diagnosis on a new test image.
I worked closely with Mulago Referral Hospital under the Department of Internal Medicine and Lund University in Sweden. I was supported through a Sida funded project under the PI-ship of Prof. Gilbert Maiga, called Training for Sustainable Spatially Enabled e-Services Delivery (TSSEED). My research was supervised by Dr. Ernest Mwebaze and Dr. Aminah Zawedde with mentorship from Dr. John Quinn. I graduated from Makerere University in May 2022.
Do you feel that your PhD program can benefit Uganda as a country?
My invention is applicable in most rural areas that do not have access to reliable microscopy diagnoses. In Uganda, there is a lack of enough skilled and trained technicians to do microscopy diagnoses. Using smart phones and web based diagnostic platforms, we are able to work in highly endemic settings with malaria. The innovation also ensures that the malaria detector may not necessarily need access to the internet, which can make it possible to increase the number of patients diagnosed.
The benefits of using Artificial Intelligence (AI) to diagnose malaria include improved accuracy in the diagnosis of malaria as well as increased speed and consistency in malaria detection. I have worked closely with microscopists and lab technicians in Mulago Hospital to identify the characteristics of the malaria pathogens. Once my model is fed with new data, it can easily map out where the parasites are. It can also detect the severity of the disease by identifying the quantity of parasites in a given sample.
My hope is that this invention can augment and support the work of microscopists, especially in malaria endemic parts of Uganda and other Sub-Saharan Africa regions with hard to reach settings. I hope that the Ministry of Health will be interested in the outcomes of my research as well.
How do you expect your PhD research to help you in the advancement of your career? What is the next step?
My research has propelled me towards a career path in artificial intelligence in health. It has given me an opportunity to apply artificial intelligence in health while creating an impact. Malaria diagnosis will need big collaborations within the medical sector including both the government and development partners. I hope to look for further funding to scale up the project and further research in this area.
My model can also be utilized to diagnose cancer, TB and many other microscopically diagnosed diseases. There is limitless potential when it comes to the use of artificial intelligence for solving societal challenges.
Tell us about the Artificial Intelligence Lab.
I work under the Makerere Artificial Intelligence Lab as a Research Scientist. It is a laboratory for research in AI under the Department of Computer Science at Makerere University. Researchers in this lab focus on how to use artificial intelligence to solve challenges in health, agriculture, traffic monitoring and markets among other areas.
I joined the lab in 2015 when I was doing my Master’s in Computer Science with a focus on artificial intelligence. The lab is headed by Dr. Joyce Nakatumba-Nabende.
What challenges do you expect to face as you consider applying your research?
The model is data intensive, with many images needed to train the AI on how to detect the malaria parasites. Additionally, my research only focused on Mulago National Referral Hospital so I would like to study different environments and health facilities as well.
The next step is to scale this up and to include different settings. I would need support for this possibly from other development partners, as there will be a cost element if we want to role this out in endemic areas.
Aside from your professional achievements, have you been able to benefit from the support you received from Sweden in other ways?
I was able to receive career guidance from my supervisors in Uganda and Sweden that has helped me. I also developed management skills through my research and learned how to do good budgeting and reporting. I have now developed a bigger network both in Uganda and Sweden. I look forward to possibly collaborating more with Lund University with the hope of further developing my career path.
Do you have anything else you would like to share with us?
I would like to thank my supervisors, Dr. Amina Zawedde and Dr. Ernest Mwebaza, for all their assistance that helped me complete my research. I would also like to thank my PI, Prof. Gilbert Maiga for the continued mentorship and review meetings that pushed us to get results for our research. I also appreciate Prof. Buyinza Mukadasi and his team for ensuring our research study was efficiently coordinated. Lastly, I would like to thank the Government of Sweden for supporting me to complete my PhD in time. I am forever grateful.
Compiled by Andrew Byaruhanga, Communications Officer, Embassy of Sweden
