E!12748 AD Shield Lytics Health AB and RISE
| Reference number | |
| Coordinator | Sentiantechnologies AB |
| Funding from Vinnova | SEK 4 912 912 |
| Project duration | April 2019 - September 2022 |
| Status | Completed |
| Venture | Eurostars |
Purpose and goal
The aim was to use AI to enhance the functionality of the Alzheimer´s detection game Starry Night. Sentian has processed, validated, and studied data from the gaming platform and clinical studies. Correlations have been made between game data, and efforts have been made to improve algorithms for game scoring. RISE intended to assist with interaction design expertise and has done so. Another goal was to lead a clinical evaluation of the developed technical solution. This task was redefined, and the goal dropped. Both Sentiantechnologies and RISE have met their goals.
Expected results and effects
The work in analytics and AI has led to insights into causal relationships and shed light on a need for more granular data. A tangible result is an executable machine code with an algorithm that can be used for future data sets. The contribution from RISE is expected to amount to a more user-oriented solution as a result of the project.
Planned approach and implementation
The AI project had three parts. First, a preparatory part where game data and its architecture were analyzed, and a literature study was carried out. An analysis platform was developed in the second part to accelerate the analysis. Finally, game data were correlated with patient data from a clinical study. The project was planned and executed through periodic meetings between the project partners and precisely planned singular work sessions with only RISE and with other actors.