[mattkrause]

There's a huge chunk of neuroscience devoted to questions like this.

Several groups have shown that they can "decode" a remembered image from brain activity. This is comparatively easy when the images are simple and there are only a few possibilities, but can generalize to larger sets of images and even (sort of) never-before-seen ones. Sensory and motor information is relatively accessible; I don't know that anyone's making great progress decoding thoughts like "I should be home by 8pm".

If this is up your alley, you may want to check out work done by Frank Tong's lab at Princeton (e.g. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1808230/) or Jack Gallant's group at UC Berkley (e.g, https://www.sciencedirect.com/science/article/pii/S105381191...), among others.