NeuroDataShare

Attentional and spatial deficits are common after brain damage (vascular, neurodegenerative or traumatic), and have dramatic consequences on patients’ everyday life and functional prognosis. Neurocognitive assessment and training is crucial for diagnosis, rehabilitation and prevention of these disorders. Data collection and processing in different countries, by different experts in hospitals and research centres are severely limited by the use of traditional paper-and-pencil tests. Despite their large diffusion, these tests lead to results difficult to evaluate and compare in populations of patients. A huge amount of potentially crucial information is thus lost. The introduction of Artificial Intelligence in the field of clinical diagnosis and rehabilitation is now changing the landscape. In recent years, novel techniques of technology-enhanced assessment entered the clinical practice. Their use has particularly benefitted the diagnosis and follow-up of visuo-spatial cognitive deficits, because these deficits typically manifest themselves during physical object manipulation or navigation. These techniques have the potential to collect a great amount of data, which however are not currently exploited in the best possible ways. Data from many patients with similar neurological conditions across several European countries could feed a public, anonymized database, open to researchers and clinicians. Such very large database would allow researchers to perform analyses with an unprecedented level of statistical power, depth and completeness. This approach needs diverse competences, in clinical and experimental neuropsychology, computer science, big data management and artificial intelligence. The objectives of the NeuroDataShare project are: 1. Develop an European Shared Database, using big data paradigms, openly sharing the data collected from neurocognitive assessment and trainings with a twofold scheme: (1) providing researchers with the opportunity to extract and apply data by using pattern recognition and Artificial Intelligence models; (2) providing clinicians with automatic algorithms for a rapid and easy data extraction from different patients. Design and apply the highest level of security for the data, including all the relevant features of security and anonymity, according to the most strict European laws. Create a common methodology in order to bridge the tools and procedures of the assessment and training of visuo-spatial and attentional abilities, allowing a quick digitalization of the previous and future data, feeding the database with data from neurocognitive assessment and training designing user-friendly platforms and applying co-design criteria involving clinicians, physicians, and practitioners. 2. Benefit from the potential of the data aggregation of the shared database to answer neuroscientific and clinical questions, by identifying neural and behavioural predictors of patients’ recovery and response to treatments. 3. Promote and include new methods for the tracking of the assessment and training sessions, using the new technologies, in particular applying tangible interfaces, augmented reality and gamification paradigms. 4. Apply Artificial Intelligence and machine learning methods to transfer, share and store collected data for researchers and professionals, to allow a common ground, and creating algorithms for automated interpretation of the clinical tests. 5. Increase the awareness on this theme in the scientific community, disseminating the results of the project and measuring the numerical impact of this process; 6. Create a multidisciplinary community of practices composed by neuroscientists, computer scientists, engineers, psychologists, physicians and practitioners, organizing an international conference about the digitalization of the data in the neurocognitive assessment and training; 7. Define plans to apply the set of methods developed within NeuroDataShare to other clinical fields.