Eoin Brophy is a PhD candidate; his research focuses on developing deep learning methods as a reliable tool to generate and process medical time-series data that adhere to privacy-preserving regulations. Ultimately, this will allow for sharing and disseminating of medical data and, in turn, advancing clinical research in the relevant fields, such as in the NICU. Eoin is passionate about using AI for improved consumer health and patient outcomes and has extensive research expertise in developing machine learning applications and rich time series data solutions. In addition, he successfully developed novel deep learning techniques to generate synthetic physiological data.
In 2017, Eoin received his MEng in Electronic Engineering from Maynooth University and worked as a lecturer for IADT in Electronics and Microcontrollers. Following this, Eoin held a research assistant position in SAP and Insight SFI Research Centre for Data Analytics in 2018. He started his PhD journey in late 2018, working out of DCU School of Computing and INFANT Research Centre.
Eoin recently defended his PhD entitled “Deep Learning-based Signal Processing Approaches for Improved Tracking of Human Health and Behaviour with Wearable Sensors”
The outcomes of his PhD were as follows:
· The first contribution lies in forwarding the field of medical time series data generation using recurrent Generative Adversarial Networks. He developed a custom architecture and loss function towards improved physiological data generation and, more significantly, provided continuous multichannel time series generation using GANs for the first time. This advancement is crucial as it means this technology can be used in further clinical training and research applications because it allows for sharing and dissemination of sensitive medical data.
· The second contribution put forward by his work explores novel deep learning technologies applied to continuous physiological time series data and unlocks further insights from these datasets. In addition, he developed novel applications of existing architectures and pioneered loss functions to improve human-centric model performance.
These PhD contributions push the envelope for state-of-the-art deep learning methods for human physiological data processing.
Eoin actively provides specialist consultancy services in the Irish technology space and hopes to undertake commercialisation opportunities in the future. Eoin was supervised by Prof. Tomas Ward (DCU) along with Prof. Geraldine Boylan (INFANT), Prof. Maarten de Vos (KU Leuven), and Dr Lorcan Walsh (Novartis).