Related Experiment Video
Updated: Jul 6, 2026
Identification of Disease-related Spatial Covariance Patterns using Neuroimaging Data
Published on: June 26, 2013
Extension of voxel-based lesion mapping to multidimensional neurophysiological data
Richard Hardstone1, Lauren M Ostrowski2, A Nicole Dusang2,3,4
1Center for Neurotechnology and Neurorecovery, Department of Neurology, Massachusetts General Hospital, 101 Merrimac Street, 317A, Boston, MA, 02114, USA. rhardstone@mgh.harvard.edu.
We developed Multidimensional Voxel-based Lesion Mapping (MD-VLM) to link brain lesion locations with complex neurophysiological changes, like electroencephalography (EEG) data, after stroke. This method reveals specific neuroanatomical-neurophysiological relationships, improving understanding of stroke recovery.
Area of Science:
- Neuroscience
- Neurology
- Biomedical Engineering
Background:
- Focal brain lesions induce neurophysiological alterations in neural systems.
- Linking specific lesion locations to electroencephalography (EEG) changes post-stroke remains challenging.
- Existing methods like Voxel-Based Lesion Symptom Mapping (VLSM) are limited to single-feature symptoms and cannot analyze multidimensional neurophysiological data.
Purpose of the Study:
- To introduce Multidimensional Voxel-based Lesion Mapping (MD-VLM), an extension of VLSM designed to associate lesion locations with multidimensional neurophysiological data.
- To establish a method for identifying significant neuroanatomical-neurophysiological relationships in stroke patients.
- To demonstrate the feasibility and utility of MD-VLM in stroke research.
Main Methods:
- MD-VLM was developed as a data-driven extension of VLSM, capable of integrating various lesion and neurophysiological data types.
- The method employs robust statistical approaches to detect significant associations between anatomical injury and complex neurophysiological patterns.
- MD-VLM was applied to electroencephalography (EEG) data from chronic stroke patients during a cued-movement task.
Main Results:
- MD-VLM successfully identified significant relationships between lesion locations and neurophysiological responses in stroke patients.
- Specific associations were found between frontal white-matter lesions and reduced ipsilesional parietal cue-evoked EEG responses.
- Findings are consistent with known fronto-parietal network disruptions following stroke.
Conclusions:
- MD-VLM is a novel and effective tool for linking brain lesion locations to multidimensional neurophysiological data.
- This method enhances mechanistic understanding of post-stroke neurological impairments by clarifying neuroanatomical-neurophysiological links.
- MD-VLM holds potential for guiding the development of new biomarkers for stroke recovery and rehabilitation.
