Bucharest, 24 June 2022 – The groundbreaking European Research & Development project, DeepHealth (Deep-Learning and HPC to Boost Biomedical Applications for Health), has successfully concluded, marking a significant achievement in medical diagnostics. The project consortium, including the dedicated team from SIMAVI, has developed innovative Machine Learning models that enhance the speed and accuracy of diagnostics in urology.
These advanced models are capable of identifying areas of interest within CT scan images, a first-of-its-kind accomplishment, especially in the field of adrenal gland medicine. Prior to this project, such precision in diagnostics had not been achieved.
Project DeepHealth: Aiming for Excellence in Healthcare
The DeepHealth project aimed to develop high-performance tools based on Artificial Intelligence, Machine Learning, and Computer Vision technologies. These tools are intended for integration with existing biomedical applications in medical practices. The project’s ultimate goal, now realized, was to empower medical professionals with advanced technologies for the diagnosis, monitoring, and treatment of patients.
SIMAVI’s Key Contributions
Within the DeepHealth project, SIMAVI played a crucial role with two major contributions:
- Development of a User-Friendly Front-End Application: In collaboration with UNIMORE and using the EDDL and ECVL libraries, SIMAVI developed an intuitive front-end application as part of the DeepHealth toolkit. This interface simplifies the use of AI, Machine Learning, and Computer Vision processes for medical professionals, presenting results through easy-to-understand graphics.
- Creation and Implementation of Machine Learning Models: Utilizing state-of-the-art techniques in AI/ML and leveraging the project’s libraries, SIMAVI developed eight neural network models. These models assist medical personnel in making faster and more accurate diagnoses in urology, particularly in identifying critical areas in CT scan images. The clinical support for this development was provided by the team from the “Professor Doctor Theodor Burghele” Clinical Hospital, who facilitated data collection and anonymization.
These advanced diagnostic processes can be controlled via command-line interfaces, offering greater flexibility albeit with a steeper learning curve. To assist medical staff in navigating these technologies, a comprehensive user manual has been developed.
DeepHealth Consortium
The DeepHealth project brought together 22 partners from nine European countries, including Romania’s SIMAVI and “Prof. Dr. Theodor Burghele” Clinical Hospital. Other notable participants include NTT DATA Spain, Universitat Politècnica de València, Philips Medical Systems Netherlands BV, and several leading research institutions and healthcare providers.
This diverse consortium contributed a wide range of expertise, from medical knowledge to artificial intelligence and supercomputing, ensuring a robust and comprehensive approach to healthcare innovation.
For more information, please visit: DeepHealth Project Website
