A real-time machine-learning neurofeedback system for facilitating sustained attention and mindfulness (Neurofeed)

Different states of mind are presumably reflected as differences in the ongoing brain activation. However, depicting the critical features in the electromagnetic brain recordings, that would separate these states, is extremely challenging. For sensory stimuli (pictures, words etc.), that are presented in periodic and/or repeated manner, the associated brain activation can be extracted fairly easily. For understanding human behavior in real-life contexts, more ‘ecologically valid’ experimental conditions need to be utilized. This will offer also potential for application development. The aim of this project is 1) to extract features in ongoing brain activation that differentiate between undefined resting state, non emotional ‘mind wandering’, emotionally engaged (anxious) mind wandering and sustaind attention and 2) to develop a neurofeedback system whose purpose is to facilitate sustained attention by detecting the state of wandering thoughts. Sustained attention is a key component in mindfulness –based interventions, that are used to decrease stress and increase well-being. With neurofeedback, selected features of brain activity can be visualized to the subject real-time, with the goal of self-regulation. Big data samples are collected and machine learning methods are developed that can be utilized also in real-time closed-loop systems. The consortium is led by Professor Aapo Hyvärinen, the other subproject leaders are Lauri Parkkonen from Aalto University and Tiina Parviainen from University of Jyväskylä.


Project team



  • Academy of Finland ICT2023 program