Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Shifting Activation in Subacute to Chronic Stroke Recovery: A Functional Near Infrared Spectroscopy Study.
Poster Session D, Saturday, September 13, 5:00 - 6:30 pm, Field House
AnnaKate Spotts1, Hana Kim2, Argye E. Hillis1, Lisa D. Bunker5, Erin L. Meier6, Alexandra Z. Durfee3, Voss Neal1, Victoria Tilton-Bolowsky4, Melissa D. Stockbridge1; 1Johns Hopkins University, School of Medicine, 2University of South Florida, 3Towson University, 4Teachers College, Colombia University, 5Midwestern University, 6Northeastern University
Introduction: Discourse comprehension deficits following left hemisphere (LH) or right hemisphere (RH) stroke can severely affect participation in everyday conversations. The right prefrontal cortex (PFC) and left superior temporal gyrus (STG) are critical structures employed in a non-lesioned brain during discourse comprehension (Morales et al., 2022), but how discourse comprehension recovery unfolds over time following LH versus RH stroke remains incompletely understood (Barbey et al., 2014). Evidence suggests that recovery of auditory comprehension in LH stroke is marked by an average shift in peak activation from contralesional to ipsilesional structures including the STG (Saur et al, 2010). This study used functional near infrared spectroscopy (fNIRS) to examine how lesion lateralization influences longitudinal changes in bilateral PFC and STG activation during discourse comprehension from subacute to chronic stroke stages. Methods: Individuals with LH and RH ischemic stroke completed discourse comprehension tasks that included three alternating blocks of ~1min stories from the Discourse Comprehension Test (Brookshire & Nicholas, 1993) at 3 months (subacute; LH N=19; RH N=5), 6 months (LH N=22, RH N=5) and 12 months (LH N= 19, RH N = 9). FNIRS data were acquired using a NIRx NIRSport2 device with 16 sources and 16 detectors positioned over bilateral perisylvian regions, resulting in 46 long- and 8 short-distance channels. Raw fNIRS data were processed using standard processing steps before using a general linear model to estimate oxygenated hemoglobin (HbO) using Homer3 software. Averaged changes in HbO in bilateral ROIs (PFC, STG) were analyzed for each stroke group (LH/RH) at subacute (3m) and chronic (6m,12m) time points. Results: A priori analysis consisted of one-sample t-tests and Wilcoxon signed-rank tests, based on Shapiro-Wilk normality assessments, to test whether mean HbO changes differed from zero across ROIs at each timepoint. RH participants showed significant activation in the right lateral PFC at 3 months, t(4) = 2.44, p = .036, d = 1.09. LH participants showed modest activation in the left STG at 6 months, t(21) = 1.83, p = .041, d = 0.39. All other comparisons were non-significant (p > .05). Due to low group-level power and stroke heterogeneity, Fisher-Freeman-Halton exact tests with Monte Carlo sampling (10,000 samples, 99% CI) were used for participant-level analysis. All tests were non-significant (Monte Carlo p = 1.000), including for LH strokes (left/right STG, left/right PFC) and RH strokes (left/right STG, left/right PFC) at all timepoints. Pearson chi-square and likelihood ratio tests also yielded non-significant results (p > .14), showing no consistent activation patterns over time within individuals. Conclusions: Contrary to the literature-based hypothesis, no consistent shift from contralesional to ipsilesional activation was observed across regions or timepoints. A key limitation of this preliminary study is that lesion size and location were not accounted for—an ongoing challenge in fNIRS research. These findings suggest that compensatory mechanisms may be highly individualized and depend on individual characteristics, such as lesion location and health of the contralesional hemisphere. Future studies will evaluate the distinct patterns of change in activation across individuals, and how these changes relate to accuracy of comprehension.
Topic Areas: Disorders: Acquired,