Talk abstract: AI in Medicine: the good, the bad and the ugly. Advancements in AI have brought great promise of its applications in the fields of medicine and healthcare. AI applications have been proposed in almost every area of medicine: from tools that predict Covid-19 disease severity or treatment response in cancer to management systems that optimize appointment scheduling by accounting for missed appointments.
This talk is a first-hand data centric analysis of AI in medicine. We will analyse methodological limitations, root causes of failure and identify what predicts success of an AI model in this field. We will discuss and establish the healthcare sectors and methods with more potential for successful AI deployment and where AI can make an impact in clinical practice.
Bio: Miguel is a former MD turned statistician and data scientist who runs a consultancy business that provides data science and statistics support to companies ranging from start-ups to large, multinational organisations. He has experience working in the tech space, specially, bio- and medtech and has a portfolio of projects involving the use of standard statistical methodology, Bayesian statistics, machine learning and deep learning approaches and combination of bioinformatics and statistics to create novel analytical methods. He is passionate about educating and training the next generation of data science practitioners who can bring new ideas and approaches to the field.