Cloud computing is now a fundamental part of many services we use daily, often without realising it. From streaming our favourite shows, storing photos, and accessing health information, many of the services we rely on are powered by cloud technology. For research colleague Dr. Taylor Hanayik, it has been pivotal in scaling up neuroimaging analysis software quickly and effectively. Dr. Hanayik worked closely with the University’s Cloud Computing for Research Competency Centre, a part of Oxford’s Digital Transformation programme, which offers researchers consultancy and support to make use of the cloud services available to them.
Firstly, what is cloud computing?
Cloud computing provides the option for researchers to create and manage digital infrastructure in a flexible way, allowing users to scale up and down projects accordingly, which has become increasingly important with the use of AI, simulation and data analysis. Cloud service providers, such as Amazon Web Services (AWS), Google Cloud Platform (GCP) or Microsoft Azure, offer pay-as-you-go on-demand delivery of IT resources over the Internet. This reduces the need for researchers to have their own physical infrastructure units which can be highly costly and inefficient for large scale projects. Programs held on the cloud are also more broadly available, allowing access globally rather than needing a centralised team.
This is why cloud computing made sense for Dr. Hanayik, a Senior Research Software Engineer working with the Analysis Group, who was looking to enhance and distribute existing brain imaging analysis software. The software package, FSL, is a library of individual programs for analysing MRI-based images, used in 150 countries across thousands of labs. The FSL application represents 20 years of software development but navigating it and making usable disease specific programs requires a lot of technical know-how – something which clinicians working with medical imaging may not have.
The Analysis Group aimed to solve this problem by building a new platform, named FSL Clinical, extending the existing FSL research software to offer analysis solutions which are accessible to clinicians. The aim is that the software can analyse magnetic resonance imaging (MRI) scans and provide teams with accurate reports and data to inform clinical decisions, saving time and increasing confidence for radiologists who would usually need to analyse images by eye without helpful insights from software.
When creating this platform, the FSL Clinical team wanted the platform to be accessible from anywhere in the world without having to set up costly infrastructure which would require a great deal of technical expertise to maintain – this is where cloud computing came in handy. Having contacted the Cloud Computing for Research Competency Centre, the FSL Clinical team began working with Dr. Ansar Rafique, a Senior Research Software Engineer with expertise in cloud platforms, who was able to quickly start creating the cloud-based infrastructure needed for the FSL Clinical applications. Having the support of the Centre meant that Dr. Hanayik was able to focus on other aspects of the platform, including setting up the data analysis pipeline and user interface aspects for web applications. Dr. Hanayik said, “Having the support of the Competency Centre made this project timeline happen faster”.
In addition to setting up the cloud-based software, Dr. Rafique and the Competency Centre also helped to develop a system to automate the creation and destruction of cloud resources, such as storage and databases, using the infrastructure-as-a-code software tool Terraform. This is important for the FSL Clinical application as removing files and programs from the cloud once they’ve been used ensures the security of the data contained within them. Automation greatly reduces the time and effort required in doing this manually, as well as minimising the need for technical expertise and continuous monitoring of the cloud.
In all, the partnership between FSL Clinical and the Cloud Computing for Research Competency Centre has allowed the FSL Clinical team to upscale the existing software and create an accessible application prototype ready for feedback from their first clinical users. The first prototype of this went online at the end of June and is being used as a clinical research tool in the Oxford Brain Health Clinic to start analysing patient data from MRI scans in the next few weeks.
This is just one example of how the Cloud Computing for Research Competency Centre can help researchers reach their goals. The Competency Centre can offer personnel with technical expertise to fill gaps in research teams, leaving those teams free to focus on their core work. At the same time, the Centre is open to conversations with all colleagues to explore research needs and whether a cloud-based approach is appropriate for a specific project.
If you’re interested in seeing where the cloud could take your work, get in touch with the Cloud Computing for Research Competency Centre team to learn more.
For more information about Oxford’s digital transformation journey, please visit the Digital Transformation webpage.
