Periodic Reporting for period 3 - MES-CoBraD (Multidisciplinary Expert System for the Assessment & Management of Complex Brain Disorders)
Reporting period: 2024-04-01 to 2024-09-30
In the European Union alone, around 165 million people are affected by a brain disorder. Around 1 in 3 people are estimated to eventually suffer from a neurological or mental disorder. The Multidisciplinary Expert System for the Assessment & Management of Complex Brain Disorders (MES-CoBraD) project aims to improve the diagnostic accuracy and therapeutic outcomes of complex brain disorders, such as epilepsy, dementia, and sleep disorders. The work of this project was mainly aimed at addressing these issues by focusing on improving the tools available to researchers both to conduct traditional and AI research and to gather usable data from their patients, ready for use by themselves and other researchers in a fast and efficient process. In addition, the project also conducted research on aspects of complex brain disorders research such as detection using novel techniques and the impact on patients. This research was conducted by taking advantage of the gathered data and the tools developed during the project’s course.\nTo contribute to this, the MES-CoBraD project created a unified online accessible platform that anyone can use, which hosts most of the tools developed during its development. This platform allows researchers to gather data, upload data, pre-process them, set up and save workflows that describe common or novel research or therapeutic processes and analyse them using the over 100 analytics techniques present in the platform. This whole process is completed in the researchers’ browser without needing any specific technical knowledge regarding the installation and maintenance of the underlying tools, nor, more importantly, the need for highly demanding computational resources. The platform, using the semi-automated tools, allows researchers to easily adjust and repeat processes, enhancing their efficiency and their capabilities.
During the course of the project, the following results were achieved:
• An Expert System comprised of the following modules:
o A Data Analytics module was created, comprised of xx modules. Those modules focus on four categories: tools for the analysis of time series data, mainly EEG data; tools for the analysis of MRIs; tools for actigraphy analysis; and tools for general statistical analysis. The module includes a GUI and, with small alterations, can be run independently from the platform.
o An AI Analysis module that contains both traditional ML techniques for use with tabular data and tools developed for new specific use cases such as detection of ictal segments in EEG signals and focal cortical dysplasia.
o A workflow manager that allows researchers to create and save workflows of diagnostic or treatment processes and connect those steps with the analytics modules and alter the flow dynamically based on results of each.
o A query builder allowing efficient preprocessing of tabular data and creation of new datasets.
• A questionnaire builder integrated with the metadata manager, tracking added variables.
• A database of RWD variables of interest collected from the data of participating organisations and used by the expert system.
• An Edge application that can be deployed at the location of researchers allowing uploading of data to the platform and depersonalisation.
• A live version of this platform hosted online that can be accessed by the public.
• The ETHAI (Ethical AI) model outlining ethical requirements for the use of AI..
In total, the work of the project led to the creation of 22 publications in journals and conferences.
• An Expert System comprised of the following modules:
o A Data Analytics module was created, comprised of xx modules. Those modules focus on four categories: tools for the analysis of time series data, mainly EEG data; tools for the analysis of MRIs; tools for actigraphy analysis; and tools for general statistical analysis. The module includes a GUI and, with small alterations, can be run independently from the platform.
o An AI Analysis module that contains both traditional ML techniques for use with tabular data and tools developed for new specific use cases such as detection of ictal segments in EEG signals and focal cortical dysplasia.
o A workflow manager that allows researchers to create and save workflows of diagnostic or treatment processes and connect those steps with the analytics modules and alter the flow dynamically based on results of each.
o A query builder allowing efficient preprocessing of tabular data and creation of new datasets.
• A questionnaire builder integrated with the metadata manager, tracking added variables.
• A database of RWD variables of interest collected from the data of participating organisations and used by the expert system.
• An Edge application that can be deployed at the location of researchers allowing uploading of data to the platform and depersonalisation.
• A live version of this platform hosted online that can be accessed by the public.
• The ETHAI (Ethical AI) model outlining ethical requirements for the use of AI..
In total, the work of the project led to the creation of 22 publications in journals and conferences.
The MES-CoBraD project researched in depth the relationship between researchers, clinicians, and patients, and the data produced during the care and research. The developed platform can be used by researchers, giving them access to both a very efficient and convenient implementation of many of the most used tools in the research of complex brain disorders and new ones. Individual modules are hosted in online repositories as open-source code and can be appropriated and used by future researchers.
Real-world data have been gathered, anonymised, and have been used to create a database of metadata to aid with their analysis. Research papers indicating new detection methods and exploring effects of complex brain disorders have been developed. These results will improve the quality of care given to patients with complex brain disorders and help accelerate research of these disorders.
Real-world data have been gathered, anonymised, and have been used to create a database of metadata to aid with their analysis. Research papers indicating new detection methods and exploring effects of complex brain disorders have been developed. These results will improve the quality of care given to patients with complex brain disorders and help accelerate research of these disorders.