female support worker going through paperwork with male service user in the therapy rooms of York House landscape

Dr Hannah Thompson from the University of Surrey was a speaker at the May edition of the continuing professional development (CPD) session in Kerwin Court. The session is open to the multidisciplinary team, who were keen to learn about the latest research developments in chronic stroke rehabilitation.

Understanding the world we live in is necessary in our interaction with it; we know that a mug can hold hot liquid, some musical instruments can be blown into to make noise and bananas taste nicest if they’re peeled before eating.

Dr Thompson introduced some of the most recent conceptualisations of how knowledge is organised and accessed, its underlying brain networks, and explained how information must be accessed to be relevant for the particular context. For example, the meaning of the word MATCH is different depending on whether you’re lighting a BBQ or at Wimbledon; and the use of object can be standard (e.g., using a NEWSPAPER to read) or different (e.g., using a NEWSPAPER to swat a fly).

The main focus of the talk was difficulties in retrieving these context-dependent meanings after stroke. Dr Thompson’s studies found that responses in people with left-sided (LH) stroke were biased by well-established (dominant) semantic links, such as that we make between SALT and VINEGAR, as opposed to less obvious links (non-dominant), such as SALT and WOUND, as in the expression RUBBING SALT IN A WOUND.

This reliance on pre-established links means it’s more difficult to learn new things. For example, after learning word-pairs, patients are much better at remembering previously related pairs (e.g., PARTY and BALLOON) compared to unrelated pairs (e.g., CAMERA and BISCUIT). Here, they are likely to say that they learnt something related (e.g., CAMERA and PHOTO). What’s more, their self-awareness of how well they are remembering is affected by how related the items are, rather than how accurate the patient is.

However, patients with stroke can show a positive impact of cueing; there are dramatic improvements in comprehension if given an appropriate cue. For example, this can be a sound cue in naming (e.g., /el/ for ELEPHANT) or a context cue (e.g., an image of a road for TRAFFIC). Dr Thompson showed evidence that even the room someone is in makes a difference to understanding objects; if the object is in its natural environment (such as a WHISK in the kitchen), it’s more likely to be comprehended.

These findings led to the development of a training task for patients with problems with comprehension. The training involves showing patients words to be associated, and included highly associated items like ONION and FRYING PAN, and a low associations like FLIP FLOPS and MARKET STALL. During training, they were given feedback to the correct association. So far, the researchers have found that six sessions over two-weeks of training have led to significant improvements in performance on this task. Importantly, there was some generalisation to other tasks, participants showed improved accuracy for associating words and pictures that had not been included in the training.

Alongside the behavioural results, brain imaging data suggested reduced activity in the brain when comparing the same person before and after training, suggesting more efficient processing and less effort. The session attendees were enthusiastic about these promising findings, particularly because the training was relatively brief, and some of the participants had suffered their stroke many years prior to the study, showing that change is possible even in chronic stroke patients.