Back

Vascular Territory, White Matter Connections and the Neuroanatomical Basis of Memory Impairment after Stroke

Wright, P.; O'Sullivan, M. J.

2019-11-19 neurology
10.1101/19012385 medRxiv
Show abstract

ObjectiveTo investigate neuroanatomical correlates of memory impairment 30-90 days after stroke, including the role of white matter connections in the core circuit for episodic memory. MethodsA cohort of 179 patients with first symptomatic ischaemic stroke were enrolled into a longitudinal cognitive study, STRATEGIC. Verbal and visual memory were assessed at 50{+/-}19 (range 22-109) days. Lesion topography was defined by imaging (n=152). In a representative subgroup (n=53), 3T MRI and tractography was used to define patterns of tract injury and microstructure of uninjured tracts. ResultsLesion location, defined by arterial territory, was associated with verbal memory impairment (F(12,164)=2.62, p=0.003), independent of other factors such as age, risk factor status and lesion volume. Independent lesion symptom mapping identified regions of the left posterior temporal white matter, within the left posterior cerebral artery territory, associated specifically with verbal recall. Visual recognition memory was associated with microstructure of the uninjured fornix but not with lesion location. ConclusionsInfarct location strongly influenced verbal recall performance 50 days after stroke. Damage in two locations underpinned this relationship: the thalamus; and within the left PCA territory, where disconnection of parahippocampal white matter projections contributed to verbal memory impairment in some cases. The correlates of early cognitive prognosis were domain-specific with a different pattern of associations with executive function. The association between microstructure of the uninjured fornix and visual memory might reflect the effects of comorbid pathology on hippocampal circuits.

Matching journals

The top 4 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.