Using Mahalanobis Distance to Evaluate Recovery in Acute Stroke

In the weeks immediately following a stroke, impairments across multiple cognitive domains are pervasive yet there is little literature that explores cognitive recovery during this period. This paper evaluates the use of Mahalanobis distance as a means of statistically evaluating cognitive change at the individual level.

A small battery of standardized neuropsychological tests was administered on five or six occasions across a 2-week period to the participants recovering from a stroke and a non-stroke control group. Mahalanobis distance was used to evaluate the change profile of those who were recovering from a stroke relative to the non-stroke control.

The outcomes of three patients show that Mahalanobis distance could statistically differentiate recovery, no change, and deterioration from normal repetition effects.

In the acute phase of stroke using Mahalanobis distance it is possible to distinguish between recovery, normal learning, and gerneralized learning deficits thereby identifying likely candidates for further cognitive assessment and rehabilitation.