University of Antwerp and UZA Improve and Accelerate Cancer Research with Google Cloud

At the Center for Oncological Research (CORE), part of the University of Antwerp (UA), and the Multidisciplinary Oncology Center Antwerp (MOCA), part of the University Hospital Antwerp (UZA), scientists conduct daily research into new cancer treatments and diagnostic methods. For this diverse research, scientists gather a lot of data, which they can easily analyze using various technologies. Compared to humans, computers are, after all, better and faster at making diagnoses and predictions. This was previously done in an on-premise environment, but this presented many limitations in terms of capacity, speed, and scalability. In collaboration with the experts at Qodea, this local IT infrastructure was quickly migrated to Google Cloud. The result? There is now enough capacity for all scientists to carry out their research simultaneously and faster. In addition, the new cloud environment is fully prepared for the future.

Dr. Joe Ibrahim, Senior Researcher at the University of Antwerp, talks about the migration: "Due to the limited capacity in our previous on-premise environment, scientists often had to queue to run their processes, and that's something you don't want in cancer research. Despite having a high-performance cluster, the infrastructure at the time was often overloaded and even crashed if an application needed too much memory. Also, for certain studies there was a need for GPU computing, and that was not present in the former infrastructure." Enough reasons for the University of Antwerp to think about a better solution together with Qodea.

Cloud Solution

They found that solution in Google Cloud. In close consultation with the researchers at the University of Antwerp, the specialists at Qodea delivered a completely new research environment in Google Cloud in a relatively short time. The new cloud solution consists of three components: data pipelines, dashboarding & reporting, and an infrastructure that enables data sharing with third parties.

Prof. Dr. Timon Vandamme, Senior Staff Member in Digestive and Neuroendocrine Oncology at UZA, explains: "Thanks to the data pipelines in Google Cloud, automated research can be carried out via a number of predefined steps. We can store the output of such research in the cloud. In our previous on-premise environment, we also worked with data pipelines, but there too we ran into limitations."

The second part of the solution consists of monitoring used resources, generating reports and doing project management. For this, the University of Antwerp uses Google Looker Studio, which is designed to make data insightful. The realized infrastructure, the third part of the solution, makes it possible to share data with other (research) institutions. This infrastructure is already being used to share data pipelines. The actual sharing of data will take place at a later stage.

Flexibility

For the scientists, the new environment is easier to use, and because of its modularity, they can now choose the necessary resources per project in Google Cloud. “As a research institution, we do not constantly perform the same processes," says Joe. "Every research project is different and requires different resources." Google Cloud offers a flexible and scalable environment that scientists can work with. The cloud platform also offers extensive possibilities for Machine Learning (ML) and AI. The University of Antwerp has now carried out the first tests with ML in Google Cloud and wants to further explore the possibilities around AI and ML in the cloud environment in the future.

Results

In addition to the advantages already outlined, such as flexibility and ease of use, using Google Cloud also means that the University of Antwerp no longer has to invest in its own hardware, which always ages quickly and often proves insufficiently powerful to carry out scientific research. In addition, the extra capacity is incredibly important in studies for cancer patients. Timon: "When processing data for our research institute, time does not play such a big role. Then it doesn't matter if we have to wait two weeks for research results, for example. But when it comes to a patient's research data, we want to know the results as soon as possible, because then you can immediately advise the patient on the further course of treatment."

Approach

The University of Antwerp looks back on the collaboration with Qodea with satisfaction. A joint approach to setting up the new Google Cloud environment was very important to them. "We are, after all, scientists and we like to know how everything works," Timon clarifies. "Qodea was willing to take us through the process and share all their knowledge with us."

In a period of a few months, Qodea delivered a complete scalable and future-proof solution in Google Cloud for the University of Antwerp. The scientists now have access to an almost infinite amount of processing power, more memory capacity, and more flexibility in conducting research. By using the Google Cloud Platform, they are now able to scale cancer research much more effectively and securely. Timon: "We have managed to reduce the analysis time of certain studies from five days to about fifteen hours. That is a significant improvement. All in all, they are very satisfied with the professional cloud solution that has been implemented by the specialists at Qodea."