About this seminar

Abstract: 

There is a relatively unexplored territory on how to systematically link clinical records with information acquired from sensing devices at home and provide user guidance and interoperability. This is an important research direction with direct socioeconomic impact in the healthcare clinical and technological sector. In this talk, Dr. Deligianni is going to discuss inherent challenges in processing heterogeneous healthcare data. For example, low prevalence of outcome events, inadequate labelled data, missing data along with shifts and mismatches between the training and target distribution. Subsequently, she will go through some of her current research work to address these challenges.

Short bio:

Dr. Fani Deligianni has a multi-disciplinary background, with qualifications and research experience from world-class leading institutions, which have led to high quality research contributions for challenging problems in computing science and healthcare. She has an engineering diploma in electrical and computer engineering from Aristotle University of Thessaloniki, a MSc in Advanced computing and a PhD in Medical Imaging Computing from Imperial College London and an MSc in Neuroscience from UCL. Her track record includes more than 70 peer-reviewed papers/abstracts/book chapters (25 journals, 4 book chapters) and it has been impactful in the machine learning for healthcare applications, along with medical imaging analysis community as it is reflected by her google scholar h-index 24 and 3683 citations. She has received substantial funding from MRC, EPSRC and Royal Society to develop machine learning algorithms to process medical imaging and neurophysiological data to improve risk prediction and diagnosis of diseases.

Zoom link: 

https://us02web.zoom.us/j/9662365566?pwd=ZmpaeUE1U0hyQlVzaXJBN3dVMW5aQT09