Last weekend, at the 15th Informaticians Meeting on the Harghita (the HinTa), we had the opportunity to present our ongoing FeatureCloud project and share our experience with federated machine learning in bioinformatics.
Students, entrepreneurs and developers – almost 100 persons attended the 15th edition of the event organized by the IT Plus Cluster. In light of digitalization, many professionals presented inspiring ideas about cutting-edge technologies, innovations and challenges. We care about future generations, so Egnosis did his part as a gold sponsor. Accordingly, we helped the Cluster to open the road for youngsters to boost their knowledge in a student-budget-friendly way.
Within the FeatureCloud presentation, we had many ideas to share about federated machine learning approaches with the HInTa participants. Firstly, Béla Bihari (CEO) presented the vision and motivation of the company and briefly introduced ongoing Horizon+20 projects of Egnosis. Secondly, Sándor Fejér (senior software developer) presented the FeatureCloud projects’ architecture, solution design and main challenges.
Traditional machine learning approaches, so important in understanding diseases and inventing new cures, are infeasible when the biomedical research data is globally distributed: establishing central clouds for research presents cyber security risks, not to mention legal restrictions, like EU’s GDPR.
FeatureCloud solves these issues using federated machine learning without violating patient privacy and without loss of scientific accuracy. With this approach, all biomedical data remain within the hospitals’ secure IT infrastructure. The federated machine algorithms don’t share any personal data, only model parameters (called also features, hence the name FeatureCloud). FeatureCloud is an All-In-One solution for all actors participating in the process. It offers tools for developers implementing federated machine learning algorithms and an easy-to-use environment for researchers to conduct their biomedical research.
As the project nears completion in its final year, the consortium is preparing the last deliverables that will assemble this lifelike and game-changer AI platform.
“FeatureCloud innovates a novel strategy to overcome the legal barrier of exchanging raw patient data and thus enable true large-scale medical data mining in a cyber-risk-minimizing manner. Having such a privacy-preserving artificial intelligence, by design, will generate rapid patient benefits as it will be able to combine data sets on a massive scale. FeatureCloud will finally allow medical research to enter the “big data” era in practical reality.” Source of the excerpt: https://featurecloud.eu/
Overall, we wanted to convey to young entrants that, besides a regular software developer’s career, Egnosis offers a large scale for growth. In addition to the industrial developments, there is an opportunity to participate in global change projects for those brave enough who are open to research and are not afraid of the new challenges of bioinformatics. On the one hand, these projects are excellent opportunities for them to stay in the academic environment for further studies. On the other hand, they can gain practical knowledge and experience at a company where industry-validated solutions are incorporated into global research projects.
Thank you for reading!