In advance directives, we tell our physicians and surrogate decision makers what types of care we want--and don't want--when we can no longer make decisions for ourselves. But research consistently shows a gap between patients' preferences for end-of-life care and what their surrogate decision makers think those patients want. Neuroscience and, more specifically, a specific form brain imaging -- fMRI -- may be able to help close that gap. By imaging a person's brain while he or she is making a decision, researchers can tell which part of the brain is being brought to the specific decision-making task. Such findings have shown that the part of the brain that becomes active when a research subject is making decisions based on purely personal preference is different from the part that becomes active when the subject is making socially or morally guided decisions. Instructing a surrogate to make one's end-of-life decisions should be a social, morally guided decision, not a purely personal one. So perhaps posing end-of-life care questions in a social, moral frame rather than a purely personal frame will elicit care decisions from patients that align more closely with decisions their surrogates would make.