The next AIMed Breakfast Briefing will be held in Bern, Switzerland. Please click here to follow the session via live-streaming.
Switzerland has been ranked the most innovative country in the World since year 2011, according to the Global Innovation Index co-developed by Cornell University, the World Intellectual Property Organization (WIPO) and INSEAD.
Based on a separate, independent finding conducted by ASGARD – human venture capital for artificial intelligence (AI), a consultancy set up by entrepreneur Fabian Westerheide, Switzerland also has the highest number of AI companies per capita. Of all domains, data analytics dominates most of the European AI industries.
Hence, it’s not at all surprising for University Hospital of Bern, or Inselspital, one of the top university hospitals in Switzerland which has a long history tracing back to 1354, to have a head start in establishing a clinical data warehouse.
Interoperability of data
All along, scientists and physicians, especially those engaged in AI research, understood the importance of clinical data. As mentioned in a recent Journal of Laboratory Medicine paper written by Fried-Michael Dahlweid, Chief Technology & Innovation Officer of Inselspital and his colleagues, “data is a treasure chest… However, it is also a box with (probably more than) seven locks”.
Indeed, because each institution has their own data storage system and these systems seldom communicate with one another. Coupled with the fact that clinical data itself comes in many forms ranging from images, test results, to written notes and history records. It is hard for researchers to get their hands on all of them, let alone the tedious process of turning unstructured data into the ones that can be used for analysis purpose.
The Insel Clinical Data Warehouse (IDWH) was set up to ascertain a communal platform for data management that is not only catered for research purpose, but also for hospital operations and patient treatments. Synergy optimization is the platform’s primary aim. The warehouse is also built in a way which facilitates horizontal scalability and is flexible towards future adaption.
The core of IDWH is a two-fold data lake, which captures structured data and the Apache Hadoop (Hadoop) project designed for unstructured and high-volume data. The data lake acts as a channel for seamless flow of data into the platform. There is also an intermediate layer to pre-process, reference, and partly validate the incoming data before they are eventually transfer out for different uses.
On top of data, Inselspital also active in improving patient care through non-invasive neuroimaging technologies. The Support Center for Advanced Neuroimaging (SCAN) gathers experts from the fields of physics, computer science, neuro-radiology, neurology, to develop and validate new technology-driven neuroimaging solutions.
SCAN is headed by Professor Roland Wiest, who specialized in the use of neuroimaging techniques in the diagnoses of epilepsy. Last year, he was awarded by the Swiss League Against Epilepsy for his work on the measurement of changes in the magnetic field associated with epileptic brain activity.
Furthermore, the Heart Foundation Project taking place at SCAN is developing machine learning (ML) took to predict stroke tissue damage and clinically reconstruct interpretable perfusion maps. Research Fellow, Richard McKinley in-charge of the project believes this will provide a faster and more accurate identification of at-risk tissues, rendering more treatment options for patients beyond the six-hour window after the stroke.
AIMed Breakfast Briefing in Switzerland
This coming Thursday (4 July), Dr. Dahlweid, will be chairing the latest AIMed Breakfast Briefing – Transforming Healthcare with AI at Switzerland’s Largest Hospital at the Insel Data Science Center. Together with him, there will be Professor Wiest, Dr. Ursula Widmer of Widmer & Partners, and Dr. Christian B. Westermann, Leaders Data & Analytics, PwC Switzerland.
Through an interactive panel discussion, the speakers will bring attendees through the current state of AI development in Switzerland. They will highlight present research interests, case-studies, challenges, and future trends. The event is open to the public and you may register here. You may also watch it live or revisit the session here. Do follow us on AIMed events page, Twitter, Instagram, Facebook, and Youtube, for more details and updates on the latest events.
A science writer with data background and an interest in the current affair, culture,