Back

Prediction of dynamic balance state and recovery following stroke using fMRI graph analysis

Symonitz, O.; Joubran, K.; Puzis, R.; shmuelof, l.

2024-10-29 neurology
10.1101/2024.10.28.24316247 medRxiv
Show abstract

Stroke is associated with damage to neural tissue and is the leading cause of long-term sensory-motor disability in adults. Dynamic balance impairments are one of the most debilitating outcomes of stroke, leading to increased falls and loss of mobility. While the recovery of motor functions following stroke was shown to be affected by the initial brain damage, the ability to predict recovery based on neural markers is limited due to the involvement of multiple brain areas in dynamic balance, and the limited size of available datasets. We apply graph-theory-based neural markers to predict the extent of recovery in the presence of rehabilitative treatment and the passage of time on a dataset of 21 subjects after stroke. We report that global features are more informative than local features, describing individual regions. We also report that recovery level is predicted more accurately (85%) than dynamic balance state (76%). Our results demonstrate the feasibility of graph-based analyses on limited datasets and may contribute to clinical goal setting and to mapping the neural substrates of dynamic balance.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.