I am a Principal Research Scientist at Barts Health NHS Trust in London and a Research Fellow (Honorary) at the William Harvey Research Institute (Barts and The London School of Medicine and Dentistry), Queen Mary University of London.
My first encounter with machine learning (ML), or more generally, the field of artificial intelligence (AI) was during my Master’s degree in 2007 at The University of Sheffield. At that time, I was studying Control Systems Engineering, a subdiscipline of engineering that involves the modelling, simulation, analysis, optimisation and design of controllers for processes and systems (e.g. chemical, electrical, mechanical, financial, biological, aerospace, etc.). After that, I completed a PhD in control engineering working mainly in the areas of intelligent systems, structural optimization, dynamics and control at the same university. Control engineering principles and techniques are applicable in a wide range of fields, providing the needed flexibility to explore numerous career options.
After my PhD, I worked as a Research Fellow at Sheffield where I conducted research on data modelling and optimisation of material engineering processes and systems. Thereafter, I took up a teaching position at the University of Leicester. After few years, I decided to move back to scientific research – particularly, in an area that will involve addressing real-world problems where I could also exploit my strong mathematical background. To that end, I moved into a Scientific Software Developer role at University Hospitals Birmingham NHS Foundation, where I worked as part of a cross-functional team of medical physicists and clinical scientists to develop software servers and data collection systems for health records to assist in accelerating scientific research in radiotherapy analysis and treatment of cancer. Upon the successful completion of the project, I joined Barts Health NHS Trust, London, to work on research projects involving cardiovascular epidemiological data analysis and developing advanced AI algorithms and models to address challenging problems in cardiovascular medicine.
Reflecting on my journey so far in healthcare research, I have witnessed the initial stages of the development and deployment of AI-based technologies in addressing complex healthcare problems, including identifying relationships in patient phenotypes, optimising healthcare pathways, standardizing clinical diagnosis, developing predictive models, improving accuracy of medical related decision making and many more. Although AI-based technologies in healthcare are nascent tools, the impacts of their transformative capabilities are already being felt. However, it will require inter-disciplinary efforts to further realise the full potential of AI in medicine and health sciences. In other words, solutions to the really important problems can only be addressed by researchers from several disciplines – medicine, mathematics, computer vision, computer science, AI, radiology, engineering, and others – collaborating and working together. Biobanks, which are infrastructure for collecting biomedical data and data related to health record and lifestyle of voluntary and consented participants on a large scale with the purpose of advancing scientific research, will play a crucial role as a source of data for developing AI-based applications. As an AI research scientist within the healthcare sector, it is really an exciting time to be working on world-class research in medicine and healthcare, offering me the opportunity and platform to make a positive difference, directly or indirectly, in improving the quality of lives of millions of people, now and in the future.
How I found my ‘Why’: Benevolent AI’s Data Diversity Initiative
27 October 2020
Guest Blog - Adepeju Oshisanya AI Ethics & Data Diversity Advocate, Clinical Drug Development Leader at BenevolentAI