Reconstructing mental experience from brain activity is becoming increasingly feasible through advances in neuroimaging and deep learning. Neural signals have been translated into images, text, and speech and have been applied clinically to restore communication and movement in patients with motor paralysis. Extending reconstruction to patients with disorders of consciousness (DoCs) represents the next critical step. DoCs encompass conditions such as the vegetative and minimally conscious states, in which wakefulness is preserved but behavioral signs of awareness are absent, inconsistent, or difficult to interpret. These behavioral signs may not reflect patients' underlying cognitive capacities, as neuroimaging studies have shown that a subset retains cognitive function. Reconstruction could offer insight into these otherwise inaccessible experiences and potentially restore communication. However, if applied incorrectly, reconstruction risks mischaracterizing a patient's inner life and compromising their autonomy. To clarify the current landscape, this article reviews the development of reconstruction methods, their emerging clinical applications, and the distinct interpretive challenges associated with applying these approaches to DoCs. It then offers recommendations for evaluating reconstruction results, centered on identifying awareness, validating reconstruction models, and protecting patient autonomy. The aim is to support the responsible advancement of reconstruction and its potential to transform understanding of DoCs.
Kolisnyk et al. (Wed,) studied this question.