Picture this: a day with clear sky, children playing around a lake, and a picnic blanket spread out on the grass with people ready to enjoy their meal. This is a simple summer scene that researchers at University of Montreal propose to use for a rapid assessment for language impairment, known as aphasia, a common outcome after stroke. People with aphasia may encounter difficulties ranging from speech production to comprehension. And, it is difficult to predict how long it will take a person to recover.

In-hospital speech pathologists encounter another challenge to correctly evaluate people with aphasia: Detailed assessment batteries exist, but these take time and tend to be exhausting. The researchers at University of Montral, led by Amélie Brisebois, have developed a new evaluation, by analyzing the description of a picnic scene from the Western Aphasia Battery, a classic language assessment test. This test was conceived to explore the discourse abilities of patients, or in other words, how they describe, without further instructions, an image or stimulus. A closer look at their descriptions may tell us more than we think.

The researchers investigated a variable called thematic units (TUs), which are specific references to the people or ideas present in a scene. For example, the picnic scene includes mentions of “child”, “drinking”, “dog”, “fishing”, and so on. They analyzed how patients referred to the elements in the picture, both during an initial evaluation and then in a follow-up six months later.

Brisebois and her colleagues found that the quantity of TUs a person with aphasia expresses after a stroke is correlated with the recovery of their general language abilities. This simple assessment can give clinicians cues to make faster and accurate predictions of long-term recovery in people with aphasia and invites both researchers and clinicians to explore more naturalistic ways to assess language impairments. 

Disclosure: The author is a student at the same research lab as Amélie Brisebois and other authors of the published study, but did not participate in the study described here.