Individuelle Erholung von kognitiven Defiziten nach Schlaganfal

• Antragsteller: Professor Dr. Hans-Otto Karnath
• Fachliche Zuordnung: Kognitive und systemische Humanneurowissenschaften ; Kognitive, systemische und Verhaltensneurobiologie
• Förderung: Förderung von 2016 bis 2023
• Projektkennung: Deutsche Forschungsgemeinschaft (DFG) - Projektnummer 318452167

• Erstellungsjahr: 2023

Zusammenfassung der Projektergebnisse

A first aim of our project was to evaluate and further improve lesion analysis methods as well as analysis of structural disconnection of white matter fibres as preparation resp. accompanying analyses for the main objective of our project. In this context, we studied how lesion-behaviour mapping studies are affected by a multitude of factors. In simulation experiments, we found that the use of correction factors ‚lesion size’ and ‘sufficient lesion affection’ reduced misplacement markedly compared to uncorrected analyses. Further simulation experiments addressed the implementation of supervised machine learning algorithms in the field of lesion-behaviour mapping. We found that for the use of multivariate models correction for multiple comparisons is required, and that sample sizes of at least 100 to 120 subjects are required to optimally model voxel-wise lesion location. We further filled a methodological gap in statistical disconnection-symptom mapping by guiding disconnection data and the statistical mapping of their relationship to behavioural measurements using either univariate or multivariate statistical modelling. We then used these improved methods in order to investigate the complex and disconnected network structure in three neuropsychological disorders of stroke patients, namely spatial neglect, apraxia, as well as the pusher syndrome. The principal objective of our project was to conduct a prospective longitudinal study in order to investigate and ‒ if possible ‒ predict long-term recovery of a cognitive defect. This question arises from the urgency to determine for an individual patient how likely it is after stroke-onset that a particular cognitive defect will resolve in the time course of recovery, allowing more effective rehabilitation treatments targeting at individual neglect patients. As a showcase for the general principle we concentrated on the principal consequence of stroke to the right hemisphere, i.e. spatial neglect. A first result of this endevour addressed the white matter hyperintensities (WMH). Our analysis revealed that topographic WMH infact serve as a valuable clinical imaging biomarker for predicting the severity of spatial neglect and bears great potential for rehabilitation guidance of acute stroke patients. Caused by the delay due to the COVID pandemic, we are still continuing including further patients for a final analysis of the longitudinal study design. However, a preliminary analysis based on 73 patients, which we performed for the purpose of the present report to the DFG, already gives reason to assume that it will indeed be possible to make very satisfactory predictions of the further course and the probability of individual recovery of the disorder on the basis of imaging as well as other clinical variables in the acute stage of stroke: the prediction of chronic neglect in our study on the basis of lesion map components and lesion volume, age, and the acute neglect behavior explained nearly 72 % of the total variance. This extremely promising result gives rise to the expectation that our final analysis will provide excellent prediction of individual neglect recovery, allowing clinicians to predict long-term prognosis of spatial neglect for single cases based on the clinical scans in combination with further clinical variables obtained at admission.

Projektbezogene Publikationen (Auswahl)

• An empirical evaluation of multivariate lesion behaviour mapping using support vector regression. Human Brain Mapping, 40(5), 1381-1390.
  Sperber, Christoph; Wiesen, Daniel & Karnath, Hans‐Otto
  
• Mapping human brain lesions and their functional consequences. NeuroImage, 165, 180-189.
  Karnath, Hans-Otto; Sperber, Christoph & Rorden, Christopher
  
• Post‐stroke cognitive deficits rarely come alone: Handling co‐morbidity in lesion‐behaviour mapping. Human Brain Mapping, 41(6), 1387-1399.
  Sperber, Christoph; Nolingberg, Chloé & Karnath, Hans‐Otto
  
• Using machine learning-based lesion behavior mapping to identify anatomical networks of cognitive dysfunction: Spatial neglect and attention. NeuroImage, 201 (2019, 11), 116000.
  Wiesen, Daniel; Sperber, Christoph; Yourganov, Grigori; Rorden, Christopher & Karnath, Hans-Otto
  
• Inhibition between human brain areas or methodological artefact?. Brain, 143(5), e38-e38.
  Sperber, Christoph & Karnath, Hans-Otto
  
• Lying in a 3T MRI scanner induces neglect-like spatial attention bias. eLife, 10 (2021, 9, 29).
  Lindner, Axel; Wiesen, Daniel & Karnath, Hans-Otto
  
• Strategies for feature extraction from structural brain imaging in lesion‐deficit modelling. Human Brain Mapping, 42(16), 5409-5422.
  Kasties, Vanessa; Karnath, Hans‐Otto & Sperber, Christoph
  
• Disconnectomics to unravel the network underlying deficits of spatial exploration and attention. Scientific Reports, 12(1).
  Wiesen, Daniel; Bonilha, Leonardo; Rorden, Christopher & Karnath, Hans-Otto
  
• Indirect structural disconnection-symptom mapping. Brain Structure and Function, 227(9), 3129-3144.
  Sperber, Christoph; Griffis, Joseph & Kasties, Vanessa
  
• Right hemispheric white matter hyperintensities improve the prediction of spatial neglect severity in acute stroke. NeuroImage: Clinical, 36(2022), 103265.
  Röhrig, Lisa; Sperber, Christoph; Bonilha, Leonardo; Rorden, Christopher & Karnath, Hans-Otto