Kexin Xu

Based in the United Kingdom

‍

With the vision of using AI to transform biomedical research, I dedicate myself to developing research tools to improve healthcare.

I am Kexin, a final-year B.Sc. Medical Biosciences student at Imperial College London. When Equitech Futures taught me to write "Hello World" in Python, the world of healthcare AI unfolded in front of my eyes. With the explosion of "big data" in biomedicine, we are in dire need of new research tools to exploit these data to unveil what they mean for human health. Here at Equitech Futures, I collaborate with computer scientists, who share a passion for healthcare AI, to develop AI tools that allow us to interrogate these newly-emerging health data.

In a recent project, I worked with computer scientists in the Health Lab to predict breast cancer stages from biopsy microscopy images. Under the supervision of Dr Abhilash Mishra, I worked closely with Aryan Gupta, Eugene Herasimov, Yash Gorana, and Kartik Chincholikar to develop a deep-learning pipeline for breast cancer prediction. I was able to combine my biomedical domain knowledge with their technical expertise to develop models that capture histopathological features in the images.

Ever since attending the Equitech Scholars program in 2022, I have been trying to deepen my understanding of AI through further training, including Imperial Graduate School, four summer schools, and three Coursera courses. I explored my interest in biomedical AI via three additional computational projects, including ensemble learning for NMR data, in-silico protein mutagenesis, and multi-omics prediction of biological age. I was able to publish a poster at the European Computational Biology Conference as the 2nd author and present at the International Directed Evolution Festival and the Cambridge AI in Medicine Program.