Researchers in the U.K. are claiming they can use functional MRIs and computer algorithms to predict whether patients will suffer from continuous psychosis (which is normally considered to require institutionalization) or episodic psychosis (which is not).  They claim a prediction accuracy of around 70%, while the legal standard for involuntary commitment to a mental institution is merely preponderance of the evidence.  Does this mean that courts should allow the use of this technique (or perhaps even compel it) when making involuntary commitment rulings?