AI in Health Researcher – Dr. Rose Nakasi

Meet Rose, a distinguished Artificial Intelligence (AI) researcher whose work is at the forefront of technological innovation in healthcare. Her research interests lie at the intersection of Computer Vision and Machine Learning, specifically focusing on applying these cutting-edge technologies for automated microscopic diagnosis and spatial modeling of diseases in low-resourced countries;

Hello Rose. Thank you for sparing your time to have this interview with us. We appreciate you adding your voice to this call
I am delighted to be here. Thank you for having me.

Let us start with you introducing yourself to our audience
My name is Rose Nakasi, and I am from Uganda. I hold a PhD in computer science from Makerere University. Specifically, I am an Artificial Intelligence (AI) researcher. I am also a Lecturer at Makerere University and the head of the new Makerere Artificial Intelligence Health Lab, specifically focusing on utilizing and harnessing AI-based technologies to support Health Care. We are geared towards providing improved patient care. I also chair the International Telecommunication Union (ITU)/WIPO/WHO Topic Group on artificial intelligence for malaria detection under the global initiative AI for Health (GI-AI4H) – formerly the Focus Group AI for Health (FG-AI4H).

My research interest is artificial intelligence, particularly in developing low-cost tools for improved automated solutions for diagnostic health challenges. For example, we have developed tools to support microscopy diagnostics for diseases like malaria, cervical cancer, tuberculosis, and intestinal parasites. We are focusing on diseases that present a bigger public health burden to the country.

Apart from that, I have also done some projects supported by Google. I am a 2023 AI for Global Goals Fellow, a position that supports us in using automated tools to support disease diagnostics specifically for malaria, TB, and cervical cancer. We also have a DS-I malaria project under the Data Science Initiative Africa sponsored by the National Institute of Health (NIH). This is also to support malaria diagnostics and disease surveillance across Uganda. Our work at the lab has been supported by different initiatives, for example, the Lacuna Fund, Villgro Africa, and the Science for Africa Foundation. I think specifically that’s about me. I will end it there.

This is very impressive, Rose! Your extensive work stems from a deep passion and interest in the field. Thank you for your outstanding contributions!

AI and machine learning are fairly new concepts in Africa. So, growing up, did you envision yourself being at this point in your career or working on anything tech-related?
Well, I was a very good student in subjects like physics, mathematics, and chemistry. So part of the science we do around AI sparks from that background, especially mathematics. AI is mainly mathematical models behind the scenes.

My first chance at AI was during my Masters. We had a Lecturer, who now doubles as my mentor and collaborator – Dr. John Quinn, who introduced us to the concept of AI. I had experienced some elements of computing during my undergraduate degree, but it was more of the development side. Things like website development and application software that could support us perform specific tasks. That was great but examining how and why a type of technology works was at masters. That is where I got most of the motivation to get involved in AI.

Away from the mathematics, and the graph theory, we were also introduced to advanced programming, especially with Python. I eventually learned that computing through Artificial Intelligence could be used to solve real-life challenges in health, agriculture, and many other domains! Technology then became not just a science that I appreciated from first principles, but an impact support tool to our context.

This appreciation pushed me to work on different projects – that was way back, I think in 2012! I started working with my team to build AI tools that we could use in health disease detection and games like Sudoku. This became very interesting because we could not just rely on the science behind it, but the application as well. And I remember when the microscopy project came up to support disease diagnostics, it became more appealing! We were adding a brick to solving some of the most pressing challenges, especially in the medical field. This is what has mostly laid the foundation for my interest in the use of artificial intelligence and then specifically relying on it in the health sector.

As a country, not just as a person, we have all had very pressing health challenges at individual, family, and country levels. So, to have this opportunity to use technology to support in ways it can is not a small feat. We are not only using AI for disease diagnostics but also for prediction of disease prevalence, surveillance, and preventive medication which spans the entire health sector.

The health sector is one of the most sensitive areas, and also one that critically needs support. We need to find ways to empower the few experts we have in our context. What really motivated the work that we’re doing with microscopy diagnosis is the few technicians that we have. We realized early enough that it is going to take Uganda so many years to be able to have enough trained medical experts. So we decided to leverage technologies such as artificial intelligence to support our health sector. That and so many other pressing health challenges is how we built the AI in Health lab.

Someone reading might be wondering what artificial intelligence or AI is in layman’s terms. How would you explain to them?
This is a very good question! Many people don’t know what AI is! Some even confuse it with “artificial insemination”, haha. I’m glad that you have brought this question up.

With AI, we want to be able to train the computer or build computer models that work as if they are mimicking the brain. To do this, we train the computer to recognize things the brain can recognize. For example, you can recognize patterns out of objects. If I bring it back to the context of the disease detection that we are doing, for example, malaria, can the computer be trained to learn the characteristic features of the parasites? Can it identify the parasites microscopists see? That is artificial intelligence.

We train a computer to learn patterns and once it has learned patterns, it can also make future predictions. And, computers work faster and smarter than the human brain, because they take in a lot more information than a human brain can at a given time. So, AI helps in understanding object features, learning the patterns around them, and then making future predictions. It’s more of acting like an artificial brain.

You have broken it down very well, especially using such simple examples that anyone can understand

With all the hats you wear, there must be a list of challenges or obstacles you have had to overcome. Can you share some of those with us?
When developing technologies, particularly those using Machine Learning and AI, we rely heavily on data, mostly large datasets. Early on, one significant challenge we faced was obtaining medical data. However, over time, there has been an increasing appreciation of our work. We have been able to engage in meaningful discussions with medical experts and authorities who provide us with the necessary data. Today, accessing data is no longer a challenge, allowing us to develop reliable tools that support the medical field.

One of the motivating factors for our work has always been its potential to make a real impact, especially in healthcare facilities. This goal comes with numerous demands, requirements, and the need for substantial support. Developing tools that can transition from the lab environment to practical use in various facilities is challenging. A significant obstacle we faced was the lack of funding needed to move from the proof-of-concept stage to pilot projects and ultimately to scaling up our tools. We are grateful to our development partners who have supported us in reaching our current level. However, we still require more support to achieve nationwide adoption and acceptance of our tools, ensuring they can make a meaningful impact.

Another significant challenge we faced was gaining acceptance for the multidisciplinary nature of our team and the “black box” models we initially developed. When we first started, getting these models appreciated by the domain experts, particularly in the medical field, was difficult. However, over time, through collaborative efforts between the medical team and computer scientists, we have learned to appreciate each other’s expertise. This collaboration has allowed us to develop solutions informed by all our expert domains.

I am grateful for the different engagements and opportunities, such as Words That Count, initiatives that have disseminated and promoted our work. These platforms help us raise awareness, build confidence in the masses in the tools we are developing, and disseminate our work to all necessary stakeholders. Media coverage and presentations at various conferences have been instrumental in achieving acceptance and appreciation of our work. I am confident that continued collaboration and communication will further enhance the impact of our research and the solutions we are developing.

Thank you! But, I just wanted to narrow it down a little bit. If it is okay with you, can we talk about one or two challenges that you have faced as an African woman in this male-dominated field of technology and its advancements?

One of the main challenges we face, especially in Africa, is cultural hindrances. Traditionally, many fields, including certain academic subjects and professions, have been male-dominated and often perceived as such. This cultural bias has made it difficult for girls and young women to even consider these fields as viable options. However, I am encouraged to see that this is gradually changing. With different prominent female role models and girl-child empowerment initiatives available now, these are now motivating and empowering women and young girls to pursue careers in these traditionally male-dominated areas. This progress is uplifting, particularly for women like us who are working in these domains. Continuous support and encouragement are helping us break down these cultural barriers and inspire more women to join us in these fields.

In addition to cultural hindrances, another significant challenge for women in these fields is balancing professional and personal responsibilities. As a woman, I am not only a researcher but also a wife and mother. This dual role can sometimes slow down my work, as I need to allocate time for family and childcare. Without an enabling environment, this balancing act can be particularly challenging. In many African contexts, once a woman gets married, some limitations and boundaries hinder her progress. This often slows down women’s ability to pursue their dreams. Therefore, having a supportive family can make a significant difference in overcoming these barriers and achieving one’s goals.

I am extremely grateful for the support of my family, especially my husband, who has been at the forefront of encouraging and pushing me to achieve greater heights in my research. This support is invaluable and highlights the importance of families and spouses encouraging their partners to pursue their passions, particularly in science-dominated fields. Having such support from the home front is inspiring and empowering, and it has been a crucial factor in my ability to move forward with my work.

What have been some of the most prestigious moments on this journey of the innovations and work that you do in AI?
I have had many memorable moments, but one that really stands out is when our team won the Google AI for Global Goals challenge and the NIH DS-I Africa grant. Knowing that so many competitive teams participated and that we emerged as one of the winners was incredibly inspiring. It validated our hard work and showed us that our efforts are being appreciated on a global scale.

Another standout moment was when CNN reached out to cover our work. Initially, we only focused on the impact of our work within Uganda, but being approached by a global network like CNN was a significant validation. It showed us that our work was useful and appreciated on a larger global scale. At that time, we didn’t have significant support, but the coverage gave us confidence, motivation, and inspiration. It was a cornerstone for our team, reinforcing the idea that our efforts were making a difference and were valued globally. These moments have been crucial in motivating us and ensuring we never stop pursuing our goals.

Congratulations! I know a lot is coming your way because you are doing amazing work.

You have talked about who you are inside the lab with AI and technology. But then outside computers, who are you as a person?
I enjoy spending my time in relaxing environments, especially by the beach. If you call me over the weekend to relax by the beachside, I am always up for it. Although I am not a good swimmer, I find being near water incredibly refreshing.

I also love taking trips and vacations with my husband. We often plan these getaways to take a break from our work and recharge. Additionally, I enjoy listening to music, particularly slow music. I do not have many hobbies, but these activities bring me a lot of joy and relaxation.

What can you say is your favorite quote?
I do not think this is a quote but something I can say, maybe.

I have often seen people give excuses and let their hardships define them, instead of taking action to change their circumstances. One thing I always keep in mind is to not let difficult situations become excuses for giving up or stopping efforts to achieve my goals. Even in the hardest times, I believe in pushing through and not allowing excuses to dictate my actions. I strive to do something about the challenges I face.

That is very evident in your output. With all the work you do, excuses have no room in your schedule!

If there is a young girl out there admiring what you are talking about, what advice would you give them to pave their academic path into studying AI or contributing to projects in AI?
I would urge fellow women or young girls out there that as women, we also have a role to play in the development of our country. AI is one of those fields that allow us to impact every field around us. We have seen its potential in almost every sector that is data-related. It can be used in health, agriculture, banking, and any field where data is collected and analyzed. As women, we now can work in different fields and showcase our potential. I believe that no other technology offers as much potential as AI to help us achieve this.

What I want people to take home is that we face many pressing challenges, not only as a country but globally. These challenges are waiting for a young girl out there to solve. I encourage everyone to embrace their talents and seize any opportunity to be part of the solution to the bigger problems around us. You never know – your next innovation could solve a significant challenge in Uganda or even globally. It’s important to realize that we all have a part to play in the well-being of our communities, our country, and the world. Reflect on this responsibility and take up the mantle. This is just an encouragement to each of you out there.

So, I want to encourage girls to take up the opportunity to explore the capabilities that AI offers. As a collaborative effort for both men and women in this country, we can impact the societies that we live in for a better tomorrow.

Rose, this marks the end of our session. Thank you so much for squeezing in this interview. I know you run a tight ship. So, thank you very much
One last thing, I want to thank you, Winnie, for this engaging discourse. Thank you and your team at Words That Count for sourcing us out and sharing our work. Such engagements help us move our work from the labs to the target masses. We don’t take that for granted as a lab.

We are open to collaboration. Our social media platforms are open on Twitter and LinkedIn. We also have a website. Personally, my email is open; rose.nakasi@mak.ac.ug. Thank you very much. It has been great!

The pleasure is ours.

To our dear audience, thank you for engaging.

See you next time, BYE!