"Collaboration ensures that I can do effective research, which helps advance healthcare"

Long before people with dementia develop symptoms, changes can be observed in their brains. But how do we detect those changes in the brain, and when does a deviation increase risk of dementia? Questions like these intrigue Bo Li, PostDoc researcher with the scientific program Medical Delta Diagnostics 3.0: Dementia and stroke. She explores how to use artificial intelligence (AI) to better understand the brain patterns in normal aging and in disease.

This interview is the second in a series of interviews with Medical Delta-funded PhD students and PostDoc researchers. Bo's PostDoc research is funded by the scientific program Medical Delta Diagnostics 3.0: Dementia and stroke.

How did you get involved in this research?

I am socially engaged and have a passion to improve healthcare. During my bachelor I wanted to become a doctor. But along the way I discovered that I can also help people with algorithms - maybe even more than as a doctor. The development of new technology improves possibilities for early diagnosis and prognosis for a large number of people at the same time. I did my bachelor and master in China, and was co-supervised by Prof. Dr. Ir. Bart ter Haar Romeny. Through him I eventually joined the Biomedical Imaging Group Rotterdam (BIGR) of Erasmus MC as a PhD student.

I am fascinated by the human brain, the most complex organ of our body. Dementia is one of the biggest health problems of the 21^st century. Every three seconds, someone develops dementia in the world. What’s worse, is that we still cannot fully explain its pathology. Therefore, I hope that I can contribute my knowledge of AI and image analysis to improve our understanding of dementia through this research project.

What do you hope your algorithm will accomplish?

Given the rising cost of healthcare and the limited availability of healthcare personnel, it is obviously nice if you can save time with a quick decision that is also more accurate.

Looking further into the future, I hope that the algorithm will allow us to better understand changes in the brain, so that we can see earlier if someone is ill, or even will get ill. Hopefully, we are able in the future to start a treatment to slow down the cognitive decline of the brain, and people with dementia will then be able to live independently for longer.

Bárbara, the previous interviewee, is curious about your answer to the following question: 'Looking back, what would you have done differently, and why?'

I would have involved even more people in the project at the very beginning, and put even more energy into building the network that I now have. The value of Medical Delta for my research is that it brings together experts from all sorts of fields who have common ground with my project. By talking to them, I find out what the technological possibilities are and what clinical professionals need. The collaboration ensures that I can conduct effective research, which helps to advance healthcare. Also, people from different disciplines think differently, from different perspectives. The collaboration always generates new ideas for research into new solutions.

The fact that explanation is so important to them, was a new insight for me that I only could have gained by working with doctors.In the beginning I was very much focused on the technology for instance; I wanted to develop an algorithm that worked as quickly and accurately as possible. But in talking to doctors I found out that it is not so interesting to them whether an algorithm takes “half a second” or “a whole second” to analyse the data. And even accuracy might not be the most important thing in comparison with the explainability of the algorithm. Doctors mainly want to know how deep learning arrives at the prediction; they want an explanation of the algorithm. Then they know on what basis deep learning is making decisions. The fact that this explanation is so important to them, was a new insight for me that I only could have gained by working with doctors.