New app utilises artificial intelligence to assess memory fitness
The prototyping team at the Ģֱ’s Faculty of Information Technology has created an Android-based app for assessing the user’s memory fitness. The app detects symptoms of memory disorders and refers the user to professional assessment if needed. The project was conducted in collaboration with the Alzheimer Society of Finland.
The test is largely based on speech recognition. It includes questions relating to the user’s location, arithmetic tasks and tasks regarding object recognition as well as functional segments in which the user uses the device’s touch screen. Performance is assessed using natural language processing, image recognition and other methods.
“The original tasks were adapted in collaboration with the Alzheimer Society of Finland,” says Tommi Riipinen, the researcher leading the prototype team. “Substitute tasks similar to the original ones, and which measure the same memory disorder related traits, were then constructed. Data collected from speech is easily contaminated by interference from the surroundings, which influences the accuracy of the interpretation of the data, and therefore the reliability of the test results.”
The scoring of the app-based assessment is not, and nor is it intended to be, a replacement for an examination performed by a professional. The aim of the app is to lower the threshold for having the state of a family member’s memory tested and to refer them for a professional memory assessment if needed. When a memory disorder is detected at an early stage and the treatment can be initiated promptly, the progression of the memory disorder can be slowed and its negative effects, such as the patient getting lost, forgetting to take their medication, chronic stress and depression, can be avoided.
The memory testing app prototype emerged from a larger ongoing project at the Ģֱ’s IT faculty, Watson Health Cloud Finland. “As part of the project, we have arranged many workshop events,” explains the leader of the research project, Professor Pekka Neittaanmäki. “At these events the utilisation of artificial intelligence in different walks of life has been systematically explored. Over 250 experts participated in the events and over 30 feasible use cases were selected.”
Neittaanmäki continues: “A prototyping team consisting of students in the IT field has already been successful in creating many practical applications, of which the newly released memory testing application is an excellent example.”
For more on Watson Health Cloud Finland, see the report (in Finnish): /it/fi/tutkimus/julkaisut/tekes-raportteja/tekoalyn_soveltaminen_terveydenhuollossa_ja_hyvinvoinnissa.pdf
Additional information:
Professor Pekka Neittaanmäki, pekka.neittaanmaki@jyu.fi, Tel: +358 40 550 7005, Faculty of Information Technology, Ģֱ,
Researcher Tommi Riipinen, tommi.k.riipinen@jyu.fi, Tel: +358 40 805 3507, Faculty of Information Technology, Ģֱ