Fig. 1: Experimental six-back object-scene mnemonic discrimination paradigm and study design.

A–C The six-back object-scene mnemonic discrimination (MD) task was performed in the scanner (fMRI) and remotely on a computerized web-based training platform. Stimuli were presented in sequences of 12 items: six encoding (“first” trials) followed by six test phase images. Participants were asked to respond “new” or “old” to each image using their middle and index fingers. In the test phase, they had to recognize whether the stimulus was similar but slightly changed ("lure" correct response: “new”) or an identical repetition (“repeat” correct response: “old”) compared to the “first” images. Each sequence consisted of either objects (A) or scenes (B) only. Lure and repeat stimuli only differed in shape or geometry (C). D Participants completed the six-back object-scene task inside the MRI scanner. They were scanned pre- and post a 2-week remote web-based cognitive training intervention that consisted of three 45-minute sessions per week. One group (n = 26) trained using a web-based version of the above task, whereas an active control group (n = 27) was presented with the same images but did a psychomotor task by clicking on moving neuron-icons on top of the images. Note that different sets of images were shown for each fMRI session, and the cognitive training.