2023 cohort

Proteins, the macromolecular machines of life, make us see the world, fight diseases, carry oxygen to our cells, and allow cells to be alive. They are on the basis of all major biological processes that occur on us, animals, plants, bacteria and viruses. Their complexity and versatility is the reason for their ubiquity making their study a rich field of study full of practical outcomes. With the help of AI, I aim to harvest some of this potential to help understand ourselves better and take advances in the fields of medicine and health. 

In my research, I want to address problems from the intersection of biology, ecology and health to inspire creative machine learning-based solutions. With my academic background in data science and economic research, coupled with professional experience in designing AI-based products, I now want to further community-driven machine learning research around computer vision and multimodality. More specifically, I want to dive deeper into unsupervised and self-supervised paradigms as well as causal-graph learning.

I am broadly insterested in developing deep learning algorithms like natural language processing and computer vision for some biomedical problems. Specifically, I was interested in applying VAEs, RNNs, GANs, Transformers & Flow-based models in bioinformatics to design novel molecules for accelerating traditional drug discosvery in my previous master's study. I was also interested in developing graph neural networks for some sing-cell related problems. I hope I could find out my real interests in the Biomedical AI and develop some state-of-the-art models based on deep learning algorthrithms during my first year at the CDT in Biomedical AI.

My current research interests are centered around Natural Language Processing (NLP) within the biomedical domain, where challenges including hallucination, comprehension of domain-specific terminologies, and limited data availability hinder the effective deployment of current state-of-the-art Large Language Models (LLMs) and NLP methodologies originally developed for general-domain text.

My educational background is in Psychology and Neuroscience, whereas my academic work focused on public health and epidemiology in subgroups of the population, such as older adults and autistic people. My research interests relate to the complexity, diversity and underlying neural mechanisms of psychiatric pathologies, with a particular interest in autism and heterogeneity. I am excited to apply AI methodology to gain a deeper understanding of these topics.

I am a bioinformatician with a research interest in the application of machine learning and mathematical modelling in omics data. Particularly, I am interested in Granger causality in data with a time series component. I believe it will help researchers who study processes such as haematopoiesis.

Personal website

My research interest lies at the intersection of Artificial Intelligence and biomedicine. Specifically, I am fascinated by the prospect of using structured data from genomics, electronic health records, and clinical data to uncover patterns and make personalized decisions. I firmly believe that combining AI and biomedicine brings us closer to the dream of precision medicine.

My background is in molecular biology, studying the underlying mechanisms and cellular interactions that drive biological systems. Specifically, I looked at how we can use machine learning and immune cell data to diagnose diseases such as cancer. My current research interests lie in reinforcement learning and finding ways to apply these intelligent algorithms to further assist clinicians and patients. I am also passionate about ML engineering, connecting the gap between novel research techniques and end user application.

My current research interest lies in developing methods that combine machine learning (ML) and prior knowledge to better analyse muti-omics (multimodal) single-cell RNA-seq (scRNA-seq) datasets. I am also eager to convert the methods to tools with a graphical user interface (GUI) and bridge the gap between biologists/clinicians with relatively less programming experience and informaticians.

Personal website

My research interests lie in the intersection of medical image analysis and deep learning/artificial intelligence (DL/AI). Specifically, I am interested in using DL/AI for image segmentation, registration, and classification. In addition to exploring efficient model structures and training strategies, I believe the interpretability of deep learning models is an important research topic for biomedical AI. As such, I am also interested in model bias & fairness.

In general, my passion lies not only in developing efficient algorithms to solve technical problems but also in bridging the gap between research and practical applications. I am interested in tackling various challenges in biomedical AI, including data privacy, data imbalance, and domain shift problems. In the coming years, I will focus on the intersection of medical image analysis and deep learning, with a vision to improve clinical workflow via technical innovation.