Hosted by Cardiff University's School of Medicine

Abstract: The typical life-cycle of a clinical algorithm remains: train on historical data, publish a good receiver-operator curve (ROC), and then collect dust in the ‘model graveyard’. This begs the question: if model performance is so promising, why is there such a chasm between development and deployment? In order to bridge this implementation gap, our focus must shift away from optimizing an Area Under the Curve (AUC) toward three more practical aspects of model design: actionability, safety and utility. These same principles apply in this new era of large language models (LLMs). This presentation will focus on three (often-overlooked) components of algorithm design and provide template use-cases for future machine learning research using clinical data.

Register for your free place on Eventbrite

Biography: Martin Seneviratne is a doctor-turned-data scientist, with interest in clinical informatics, aiming to bridge the divide between technology and healthcare. He is one of the co-founders of an AI startup company, Phare Health.

Having undertaken both his undergraduate degree in Physics and his Medical degree at the University of Sydney, Dr Seneviratne was working as a junior doctor at Sydney’s Royal Prince Alfred Hospital when he made the decision to leap into the world of digital health. While still at university, he developed an app to support task management across multidisciplinary clinical teams (WardConnect) and worked with the George Institute for Global Health on an app for community cardiovascular screening (HealthNavigator). Shortly after graduating, he was appointed to advisory roles with the Australian Digital Health Agency and the Australasian Institute of Digital Health. Then in 2016, he joined healthtech startup CancerAid – a ground-breaking app supporting cancer patients and their caregivers. In 2017, he began a two-year research masters in clinical informatics at Stanford University, focusing on machine learning over hospital data. During his Masters, he was made a Digital Health Fellow at Stanford Medicine X – a think tank at the intersection of technology, design and healthcare – and quickly became a leading voice on translating machine learning to the bedside. In 2019, Dr Seneviratne joined London-based AI research agency DeepMind as a Clinician Scientist, later merging with Google Health. In 2022, he was announced the Global Australian Emerging Leader Awardee. Since Aug 2023, he has been a co-CEO of Phare Health.