DARTH group mission
Ever-increasing healthcare provision demands strain national health systems globally, which struggle to meet patients' needs. There is a growing unmet clinical need to develop accurate, robust, practical decision support tools to facilitate diagnosis, assessment, and symptom monitoring to improve patient health and care.
We are a dynamic research group based at the Usher Institute, Edinburgh Medical School, University of Edinburgh working at the interface of engineering, mathematics, informatics, and medicine. We passionately believe we can design and provide effective solutions which will revolutionize contemporary healthcare delivery through capitalizing technology and harnessing data.
We develop and apply signal processing and statistical machine learning algorithms to explore data and decipher complicated concealed statistical relationships. Our algorithms are directly driven by and validated on complicated real-world problems, aiming to facilitate interpretation of the underlying key mechanisms of the modelled system. Our work is inherently multi-disciplinary and we collaborate with industrial partners and researchers worldwide. We tackle challenging problems in healthcare domains from neurodegenerative disorders and mental health, to asthma, cardiovascular disease, and neonatal monitoring.
We are organizing the EAI PervasiveHealth 2022; the best papers will be invited to submit work to journal special issues. Looking forward to an interesting meeting and catching up with colleagues working in this exciting area in person! https://pervasivehealth.eai-conferences.org/2022/
We are organizing a special issue in IEEE JBHI, where Thanasis serves as the Lead Guest Editor. This special issue invites extended versions of the best conference papers from the Pervasive Computing Technologies for Healthcare and is also an open call. See the call for papers and submit your best work!
We are organizing a special issue in Sensors, where Thanasis serves as the Lead Guest Editor. This special issue invites work broadly in smart sensing technologies. See the call for papers and submit your best work!
29 Sep 2022
28 Aug 2022
17 Aug 2022
15 Jul 2022
19 May 2022
27 Apr 2022
1 Apr 2022
31 Mar 2022
29 Mar 2022
15 Feb 2022
1 Feb 2022
Journal paper accepted in BMC Pulmonary Medicine! This study mined electronic prescription records from over 670,000 people and 41m prescriptions to estimate asthma symptom severity and trajectories towards informing asthma management.
Journal paper accepted in Computers in Biology and Medicine! This study develops an intelligent and lightweight system mining electromyography sensor data, thus automatically monitoring eating behaviour and providing real-time feedback.
Journal paper accepted in Sensors! The study builds a robust framework to analyze three dimensional acceleration data proposing a new robust acceleration summary measure. Check the accompanying MATLAB source code.
Alan Turing blogpost highlighting our recent work on feature selection that has been published in Patterns! The post makes an argument for developing robust and versatile algorithms. Check it out here.
Journal paper accepted in BMJ! The study includes a systematic review, pooling together data from 14 studies for meta analysis and the development of a decision support tool to assess the probability of acute heart failure.
Conference paper accepted in IEEE EMBC! This study focuses on estimating medication adherence from large electronic health records using asthma as a testbed.
Journal paper accepted in Springer Nature Computer Science! This study focuses on using speech to objectively identify Parkinson's subtypes. Results are generalizable across three countries, including external validation.
Two conference papers accepted at IWINAC! Both papers provide new mechanistic insights into Parkinson's through advanced processing of speech signals.
Journal paper accepted in Patterns - also check the accompanying source code! This study introduces a new robust feature selection algorithm and tests side-by-side performance of 20 feature selection algorithms across 12 datasets.