Driving innovation in clinical decision support through AI.
The mission of the DHICI Lab is to revolutionise healthcare through data science and artificial intelligence technology to create impactful, positive, sustainable change at scale.
The Rosemary Bryant AO Research Centre (RBRC) and the Industrial AI (IAI) Research Centre join forces in the DHICI Lab which is an open lab for project-based research across disciplines in digital health and clinical informatics.
Our focus
- Innovative digital health technology development
- Advanced analytic techniques for health system change
- Human-centred digital health design
- Standards-based interoperability in digital health
- Evaluation and application of AI in healthcare
The RBRC and IAI Research Centre have different but complementary capabilities which are combined in this lab to work on problems and challenges faced by industry and public institutions such as local health networks.
The lab consists of researchers and students from both centres who work together on projects that address specific problems. The lab provides a unique environment where the capabilities of both centres can be leveraged by external organisations, and knowledge and resources are shared to achieve better and quicker outcomes.
We invite you to explore the below projects to learn more about the impactful outcomes of our research. More information can be obtained from: A/Professor Georg Grossmann or Dr Jan Stanek from the Industrial AI Research Centre and Professor Marion Eckert or Mr Greg Sharplin from the Rosemary Bryant AO Research Centre.
-
Current Projects
Preharm
This project is in collaboration with the Digital Health Cooperative Research Centre (DH-CRC) and will develop a Predictive Harm Response Management (PReHRM) algorithmic tool to reduce adverse events across local health networks in South Australia.
Pattern analysis of events in patient pathways across ecosystems
Identifying patterns from observations and linking them to outcomes is an important analytical tasks for clinical decision making. This project addresses the challenge of analysing incomplete datasets where some observations of a patient pathway might not be observable and missing. This project will build a robust pattern mining framework that is able to identify patterns despite incomplete or missing data.
Adaptation of Clinical Decision Support Systems across ecosystems
Clinical Decision Support Systems (CDSS) are pivotal in modern healthcare, aiding healthcare practitioners in making accurate decisions. However, most of the existing systems are static and adaptation to a new environment or changes to the existing system are difficult to achieve. This project is investigating adaptation strategies that enable the system to adjust to changes by analysing its internal and external behaviour and data structures.
Data management and visual analytics of workforce climate monitoring of nurses and midwives
In health and social care, data currency and its visual representation is critical for informing researchers, practitioners, policy-makers, governments and industry in a meaningful way. This project will evaluate existing open-source products for increased automation of data management and data processing, and interactive data visualisation solutions that can be deployed online to customers.
-
Past Projects
Mawson project
Initiated in 2008, this project focused on creating a genetic variants registry, aiming to include all available variants, not just the prominent ones. Due to the lack of software with computer-readable outputs, a configurable tool was developed to automatically extract relevant gene variant data from human-oriented reports (using Mutation Surveyor) and securely submit them to the registry. The platform was built using Cache/Ensemble (Intersystems) and Java (on Windows, with monitored folder capabilities). The prototype was successfully tested across multiple laboratories in Australia.
xTardis
Funded by the Australian National Data Service (ANDS), this project involved collecting and managing data and metadata from experiments conducted at The Wark Institute at UniSA. The project adhered strictly to the data and metadata submission standards set by ANDS. A prototype was developed to gather data from two devices, testing different aspects of the project. This prototype was successfully tested, and the mass data extraction and handling components were later reused in data management for the Australian Synchrotron project.
Human Variome Project (NeCTAR)
As a continuation of the Mawson Project, this initiative focused on data collection for the Human Variome Project. The solution emphasized enhanced configurability and modularization, leading to a system that was successfully tested in laboratories across the nation.
