We are pleased to welcome Dr Vivek Singh to the Barts Cancer Institute (BCI) at Queen Mary University of London. As our newest Lecturer and group leader, Dr Singh is establishing his first independent lab within our Centre for Cancer Biomarkers & Biotherapeutics, researching innovative artificial intelligence approaches to medical data analysis.
Dr Singh’s expertise lies at the intersection of computing and biomedicine. He focuses on using artificial intelligence (AI) to interpret medical images and a variety of other patient data – an area holding huge promise for improving our ability to detect and diagnose cancer.
Thanks to ongoing advances in computational techniques and power, AI tools are enabling us to analyse complex medical data sets with unprecedented speed. In addition, deep learning algorithms can now discern subtle patterns in data that may elude human detection. This ability promises to aid clinicians in detecting cancer earlier than ever before – a crucial factor in treating patients successfully.
AI can also help us to take advantage of the exponential increases in available medical data – from genomic sequences to medical images and clinical records. Analysing this rich data promises to help forecast the trajectory of a patient’s cancer and enable personalized treatment plans that are tailored to each person's unique circumstances. Dr Singh’s work builds on these exciting possibilities.
We sat down with Dr Singh to hear more about his career path from electronics to cancer research and his ambitions for his new lab at the BCI.
My scientific journey began in India, where I completed my bachelor's and master's degrees focusing on electronic communication engineering and then signal processing. I entered the realm of medical research for the first time during my PhD in Tarragona, Spain. There, I researched the use of AI to detect possible breast cancer, and its molecular subtypes, in medical images such as mammograms and ultrasound images. This was a period when AI was really starting to gain traction. The development of new deep learning techniques allowed programs to learn to find subtle patterns in data without human supervision, which showed great promise for helping doctors to make the best decisions about their patients' care more quickly, effectively and affordably.
Interestingly, my wife, who is a GP, played an important role in motivating me to transition from pure electronics into the medical domain. As I immersed myself in the biomedical world for the first time, it was great to discuss my plans and ideas with her. Over time, our interests have slowly converged, and she has also developed an interest in computational biology and AI and learned how to code.
The pandemic started shortly after I graduated from my PhD. Thanks to the computational focus of my research, I was able to undertake a postdoc position remotely at Durham University in the UK. I worked with Professor Boguslaw Obara, using machine learning to detect structures in the retina in eye-scans from patients with macular holes – breaks that appear in the centre of the retina. After 6 months, when travel had resumed, I moved to the USA to undertake a postdoc at Center for Ultrasound Research and Translation led by Dr Anthony Samir at Harvard Medical School within the Massachusetts General Hospital in Boston. There, I researched the use of AI models to detect liver disease from ultrasound images. Additionally, I worked on an industrial collaboration project for quality assessment for ultrasound images.
In 2022, I undertook a final postdoc position at Queen’s University Belfast in the UK under the leadership of Professor Manuel Salto-Tellez, this time concentrating on developing an AI algorithm to analyse microscopic images of tissues (histopathology images) to predict whether patients had a mutation in the gene KRAS, which is associated with poor survival and increased tumour aggressiveness.
The BCI's clinical ties and the access it provides to biobanks made it especially attractive. Here, we can access not only a wealth of imaging data but also clinical records and genomic data, all of which I can apply my AI tools to. The BCI’s reputation as a supportive cancer research community where you can have lively discussions with knowledgeable colleagues also suggested it would be the ideal place to establish my first independent lab.
The seed funding that the institute provides as part of the Precision Medicine Part 2 program funded by Barts Charity will also be instrumental in helping me establish my lab swiftly. It is allowing me to bypass some of the initial intensive grant-writing phase typical when establishing a lab, helping me to get my first projects off the ground quickly. Thanks to this funding, I'm in the process of recruiting my very first lab member – a PhD student.
Our research will integrate diverse forms of data - such as medical imaging, genomic data and patient health records – to develop robust AI tools that aid early diagnosis across a range of cancer types for the benefit of patients and the public. The ultimate aim is to translate our research and build tools for cancer diagnosis that could help the NHS save time and money and reduce the burden on doctors. We're already looking to build partnerships with industry partners and apply for funding to support this goal. But the first step is to build the lab team by recruiting some PhDs and postdocs.
Outside the lab, I enjoy reading. Cricket is another passion of mine – I was part of a club back in Belfast. Family also plays a central role in my life. I have a young son who I love spending time with, and my wife and I are excitedly anticipating the arrival of a new family member soon, who will no doubt occupy much of my spare time!