This study aimed to investigate if a NeuroTracker intervention could improve cognitive abilities in older adults with subjective cognitive decline, and determine if AI models could be used to increase training efficacy.

What Was Studied

48 participants between 60 and 90 years of age with subjective cognitive complaints, but otherwise healthy, were assigned to NeuroTracker training group (26) or a control group (22). All participants provided detailed socio-demographic information via questionnaires and baseline neuropsychological assessments (California Verbal Learning Test, Digit Span, D-KEFS Trail Making Test, D-KEFS Verbal Fluency Test, and Stroop Test). The NeuroTracker group performed 7 weeks of training, the control group only performed NeuroTracker baseline assessments. Both groups performed follow-up neuropsychological assessments at 8 weeks and 11 weeks. Machine Learning models were used to analyze demographic and assessment data to test if cognitive performance and responsiveness to training could be predicted.

What Was Found

The NeuroTracker group experienced a large improvement in scores of around 70%, along with wide and robust performance transfer on the neuropsychological assessments at week 8, with further gains (without training) at week 11. AI models yielded highly accurate predictions of responsiveness to the training intervention.

Takeaways

The researchers propose that such models can be used to effectively tailor NeuroTracker programs to the needs of individuals.

‍

Reference: Brian Christie et al. ‘Artificial Intelligence models to enhance cognitive intervention in older adults with Subjective Cognitive Decline: pilot study’ 2020.

Looking for more information on NeuroTracker research? Feel free to contact us.