Out of almost 700 scientists conducting research at HZDR, nearly one third are doctoral students. Among them is Trina De from the Center for Advanced Systems Understanding (CASUS) in Görlitz. As a member of the research group “Machine Learning for Infection and Disease” she is looking for ways to improve the analysis of microscopy-based clinical images using artificial intelligence (AI). In the this portrait from the HZDR series about female researchers at the research center, Trina De explains why high school was the place where she discovered the joy of science – and why she particularly likes the interdisciplinary aspect of her work.

What did you study and where do you work today?

I am a doctoral student in the “Machine Learning for Infection and Disease” group at CASUS. Previously, I completed my Bachelors in Statistics at St. Xavier’s College in Kolkata, India, and later on my Masters in Data Science in the field of Applied Mathematics at Chennai Mathematical Institute in Chennai, India.

What is your research topic or area of specialization?

My current research area is the automated morphological analysis of objects seen in microscopy-based clinical images, which means recognizing the shape, structure, or form of these objects. Primarily, my goal is to improve the data efficiency of machine learning and deep learning architectures. One important task is optimizing the so-called representation space. Modern microscopy images consist of information from multiple channels and are also dependent on the involved mechanics of the system. Incorporating all information in one artificial intelligence model is therefore challenging. I aim to train such models with as much meaningful information for the intended task as possible – at the expense of information that is likely not very important for this task. In mathematical terms, I aim to increase the expressibility and separability of the representation space.

What makes your job special for you? What do you find exciting about it?

This job is exciting for me because of the varied yet interconnected nature of the research at my institute and even within my group. It gives me exposure to the excellent work of veteran and fellow researchers and allows me to be up-to-date with unsolved problems in this field. It also allows me to further my knowledge and delve deeper into topics that I had learnt, back then only theoretically, in my bachelors and masters.

Tell us about the moment you decided to go into science.

I think it was an unconscious decision when I started enjoying discovering the rigorous mathematical concepts behind naturally occurring processes and doing related school work towards the end of my high school. Almost nothing else compared to the joy I felt there. I also found that developing a scientific thought process made thinking and finding solutions to problems in all aspects of life methodical and easier.

What goals or wishes do you have for the future?

I wish to make significant contributions to unsolved problems with the research that I am currently pursuing. I have other topics of interest such as stochastic processes and financial mathematics that I would like to make part of my research in the future. I wish to work in research & development for the later part of my career.

Who (or what) has particularly encouraged you in your career?

The constant support and encouragement from my family is a big reason why I could pursue this career. And I was lucky to never have been burdened by external factors with the feeling that I couldn’t do something until I tried it myself and either succeeded or failed.